Glass 
 Book_ 
 
 COPYRIGHT DEPOSIT 
 
THE PSYCHOLOGY OF LEARNING 
 
The Psychology of Learning 
 
 An Advance Text in Educational Psychology 
 
 By 
 
 William Henry Pyle 
 
 University of Missouri 
 
 BALTIMORE 
 
 WARWICK & YORK. Inc. 
 
 1921 
 
l^% 
 
 Copyright, 1921, by 
 WARWICK & YORK, INC. 
 
 S)C!.A530355 
 
 SEP 24 '21 
 
 A\iQ 
 

 CONTENTS. 
 
 I The nature of learning 1 
 
 II The learning curve 16 
 
 ^_ III Economical learning 36 
 
 IV Economical learning, general factors 53 
 
 V Economical learning, special factors 73 
 
 VI Ideational learning 91 
 
 VII The retention of experience 113 
 
 VIII The nature of learning capacity . . .' 158 
 
 AX Measuring learning capacity 175 
 
 X Differences in learning capacity 193 
 
 XI Transference and interference 213 
 
 XII Fatigue and learning 245 
 
 XIII Inborn nature and learning 270 
 
 XIV Measures 283 
 
 Bibliography 294 
 
 Index 305 
 
PREFACE. 
 
 In this book I have tried to state every thing that is 
 known about learning. All the experimental work that 
 throws any light on the nature of learning has been 
 carefully examined, and in the light of the experimental 
 results I have endeavored to give the present impartial 
 verdict of educational psychology. Throughout, I have 
 tried to keep within the facts. When the evidence does 
 not warrant a definite conclusion, I say so. The few 
 cases of theory and speculation are plainly labeled as 
 such. The practical significance to the teacher of the 
 facts discussed is pointed out. 
 
 The successful use of this book will depend upon two 
 things : first, experimentation ; and second, application. 
 Educational psychology is an experimental science. All 
 its facts come from experimentation and careful obser- 
 vation. These facts, whenever possible, should be dem- 
 onstrated or illustrated by class room demonstrations 
 and experiments, or by experiments and observations 
 outside the class room. Directions for demonstrations 
 will be found throughout the book, but the teacher must 
 use his own ingenuity in adapting the experiments to 
 his needs and in devising others. And always, the facts 
 must be carried over to their practical application. The 
 students must see the source of the facts and the use of 
 the facts in teaching. The book contains many tables 
 and figures. The mastery of the book will depend upon 
 an understanding of the meaning and significance of 
 the tables and graphs. 
 
 W. H. P. 
 
Chapter I. 
 THE NATURE OF LEARNING 
 
 What is learning? What are its results? After we 
 have learned, in what way are we different from what 
 we were before? After a child has been in school for a 
 few years, he can read, write, spell, and perform vari- 
 ous other acts. He also knows that Columbus discov- 
 ered America, that George Washington was our first 
 president; he knows the name of his country, of his 
 state, and of our nation ; he knows many other facts in 
 geography, history, arithmetic, and the other branches 
 of study, and many hundreds of other facts that he has 
 not learned at school and that may not be in any organ- 
 ised branch of study. The difference in the child that 
 has come from learning, then, is that he can do things 
 and knows things that he could not do and did not know 
 before. He has knowledge and habits that he did not 
 possess before. The results of learning, therefore, are 
 knowledge and habits. And from the point of view of 
 its results, learning may be defined as the process of 
 forming habits and acquiring knowledge. This, how- 
 ever, is but a superficial statement of the facts. We 
 must know the nature of knowledge and of habits before 
 we can understand the nature of learning. 
 
 Habits. — A habit is an act following with more or 
 less regularity upon the presentation of a definite 
 stimulus, to which it has become coupled through 
 experience. Writing, reading, and spelling are such 
 
 [1] 
 
2 THE PSYCHOLOGY OF LEARNING 
 
 acts. In writing, the stimulus is the idea of the word 
 to be written. The movements of the arm and hand in 
 writing the word constitute the response. In reading, 
 the stimulus is the perception of the printed or written 
 characters. The response, in oral reading, is the move- 
 ment of the various organs of speech. In spelling, the 
 stimulus is the idea of the word to be spelled. The 
 response is the movements of the hand and arm in 
 writing the word, or in oral spelling, the movements 
 of the organs of speech as in reading. In all habits, 
 the stimulus is either a perception or an idea. The 
 response is muscular movement. When we say the 
 stimulus is a perception or idea, we mean the cortical 
 excitation which is the physiological correlate of the 
 idea or perception. Throughout this book we shall 
 speak of ideas in this v/ay without committing our- 
 selves to the doubtful theory of interactionism. We 
 believe that in a scientific psychology mental processes 
 must not be used as elements in a causal series. When 
 we so use them it is for simplicity of statement, and 
 with the understanding that the cause indicated is the 
 brain action and not the mental process. 
 
 Knowledge. — Knowledge is also a special form of 
 organised experience. Suppose, for example, I am sit- 
 ting at my desk writing and become cold. The idea 
 comes to me that coal should be put into the furnace. 
 I know that if coal is put into the furnace and the draft 
 of the furnace is opened, the room will become warm 
 and I will become warm. This knowledge is the result 
 of experience. Just as in the case of habit, a definite 
 response — action — follows upon a definite stimulus, so 
 also in the case of knowledge, an idea follows upon 
 another idea or upon a perception. In both cases, the 
 
THE NATURE OF LEARNING 3 
 
 definite sequence is due to experience, and usually to 
 repeated experience. At any rate, the repetition of the 
 experience makes the sequence more probable. We 
 may now define learning- as the process of coupling 
 responses with stimuli, and ideas with ideas or per- 
 ceptions. It is the establishing of definite sequences in 
 the passage of stimuli into responses, or in the passage 
 of perceptions or ideas to other ideas. 
 
 An Objective Consideration. — A strictly objective 
 consideration reduces both forms of learning to one. 
 If we study another person to ascertain the results of 
 learning, we find only definite forms of behavior follow- 
 ing upon definite situations. For example: if we ask 
 a child to take pencil and paper and write the word 
 America, he takes his pencil and makes the proper 
 letters in the proper sequence. If we ask him to tell 
 us the name of the largest city in the United States, 
 he says the proper word. In answer to our questions, 
 we get from the child always some form of action, as 
 movement of the hand in writing, or of the tongue 
 and mouth in speaking, or merely a nod or shake of 
 the head in assent or dissent. Briefly, we learn in an 
 obective study only about behavior. If we study our- 
 selves, however, we can see a difference in the results 
 of learning. We may sit perfectly still in our chair and 
 have the idea Napoleon followed by the idea General 
 of the French Revolution. There need be no move- 
 ment of any muscle of the body involved in the sequence. 
 It may be that the idea General of the French Revolu- 
 tion is followed sooner or later by muscular move- 
 ments, but such movements, even granting that they 
 take place, have nothing to do with the above mentioned 
 sequence. The only important thing to me who has 
 
4 THE PSYCHOLOGY OF LEARNING 
 
 these ideas is, that they are coupled together. The 
 movements that may follow are not necessary to the 
 sequence itself. The essential thing in knowledge is 
 idea. The essential thing in habit is movement. 
 
 It may of course be said that / know nothing of 
 another^s ideas except through his movements. True 
 enough, but he does. It may also be said that ideas 
 are worth nothing to the world except as they issue in 
 movement. This is also true. But my ideas may be 
 worth something to me and not issue in movement. 
 They may give quite as much pleasure and satisfaction 
 as any that issue in movement. Neither the fact that 
 the objective value of ideas depends upon their issuing 
 in movement, nor the fact that we can not know 
 another's ideas except through this other person's 
 movements has anything to do with the existence of 
 the ideas themselves and the connections or sequences 
 that are organised among them. Each of us knows for 
 himself that he has habits and organised ideas or 
 knowledge. This distinction between the two aspects 
 of the results of learning — objective and subjective — 
 is clear, and it is a helpful one to make in our study 
 of the psychology of learning. 
 
 Physiological Considerations. — It is evident that all 
 learning is connecting stimulus with response, idea 
 with idea. We have now to enquire concerning the 
 nature of this connection. We have spoken of it above 
 as a sequence, but have not explained the cause of the 
 sequence. The sequence itself is evident, in the case 
 of habit to objective obsei*vation ; in the case of knowl- 
 edge to internal observation. But why the sequence? 
 What underlies it? Of course it may be said, and 
 
THE NATURE OF LEARNING 5 
 
 truly, that the sequence is detei-mined by repetition of 
 experience. This is the cause of the sequence as one 
 sees it on the outside by objective observation of 
 behavior. But what has happened on the inside, in 
 OUT bodies, to determine the sequence? Now a stimulus, 
 in the simplest cases, is due to the excitation of some 
 sense organ, and a response is a muscular contraction. 
 The excitation passes over a neurone or series of neu- 
 rones to a muscle or group of muscles and brings about 
 a contraction. The process of learning, in habit forma- 
 tion, consists in doing something that insures, with 
 more or less certainty, that the stimulus shall go to a 
 certain muscle rather than to some other. This doing 
 something consists in reducing the resistance along 
 the nervous route that leads to the proper muscle or 
 group of muscles. Stimulation or excitation passing 
 through the nervous system is guided and directed by 
 nothing other than the various resistances offered by 
 the parts of this system. And learning consists, physi- 
 ologically, in building up and establishing preferential 
 routes in the nervous system. A preferential route 
 depends upon the fact that this route offers less resist- 
 ance to the particular form of excitation than do other 
 routes. This smaller resistance is due in most part to 
 previous passages of the same kind of excitation along 
 the same route. 
 
 It must be pointed out that preferential i^outes are 
 not fixed and static things, nor absolute and perma- 
 nently definite. We can never be quite certain as to 
 what route any stimulus will take. It is always merely 
 a matter of probability, for the reason that nervous 
 resistance is subject to great variation, is influenced 
 by many factors, such as fatigue, temporary blood sup- 
 
6 THE PSYCHOLOGY OF LEARNING 
 
 ply in the pai-ts involved, immediately preceding stimu- 
 lation, other concuiTent stimulation, the feeling accom- 
 paniments, etc. In the light of these facts, we should 
 say that learning, in habit-formation, consists in 
 increasing the probability, other factors being equal, 
 that a given stimulus will take a certain course. Learn- 
 ing can probably yievei' insure that a given stimulus 
 will, in spite of all other factors, go over into a certain 
 response. As well as we knov/ our own name, there 
 come times when we can not immediately say it upon 
 being asked. As well as we know to say "eighty one" 
 for the stimulus nine times nine, there come times when 
 we can not immediately do it. 
 
 The physiological basis of knowledge is precisely the 
 same as that of habit. Knowledge is a matter of one 
 idea following another idea. Now the reason that a 
 certain idea, as x, follows idea a, is that the cerebral 
 excitation underlying the idea a passes over and causes 
 the excitation underlying the idea x. That the excita- 
 tion takes this course rather than some other, and 
 brings about the idea x rather than some other idea, 
 as y or z, is due, as in habit, to preferential routes in 
 the central nervous system. From the point of view 
 of the nei*vous system, there is no difference whatever 
 between habit and knowledge. From this point of 
 view, all learning is the establishing of differential 
 resistances in the central nervous system. The paths 
 of smaller resistance have been called "bonds," because, 
 in a sense, they bind together stimulus and response, 
 or idea with idea. 
 
 Securing the Initial Connection.-— Learning is con- 
 necting, but we must now ask how the connection is 
 first secured. In the case of habit, how does the person 
 
THE NATURE OF LEARNING 7 
 
 who is foiTning the habit get the stimulus to pass over 
 to the desired response in the first place? Suppose we 
 wish to couple the stimulus ''five times five" with th6 
 response "25," we have only to tell the child to say- 
 that five times five equals twenty-five, and he says it. 
 He is able to say ttventy-five when he wishes, and he 
 can couple it to any stimulus of his own volition. He 
 has long since mastered this response. But suppose 
 it is a very young child and we ask him to make the 
 letter a. We can show him the letter and tell him to 
 make one like it. The perception of the letter may be 
 considered the stimulus. But the child can not imme- 
 diately make the letter. This is a response which he 
 has not yet mastered. How does he master it? With 
 pencil in hand he continues to make marks on the paper. 
 The results of the first trials may have no resemblance 
 to the letter a. The stimulus being maintained, the 
 efforts continue. Each time the response is somewhat 
 different. Slowly the marks begin to resemble the copy. 
 This form of learning has been named the trial and 
 error method. It is the only method in habit-forming 
 ivhen the movement involved has not already been 
 mastered and coupled with the idea of the movem.ent 
 as its stimidus. 
 
 During the early years of infancy the child, by the 
 trial and error m.ethod, is learning to make movements 
 with all the muscles of his body. Especially is he get- 
 ting control of legs and arms and hands and feet and 
 of the vocal organs. In these early years he learns to 
 make nearly all possible movements with his hands 
 and vocal organs, and these movements are organised 
 with the cortical substrate of their corresponding ideas 
 as their adequate stimuli. This learning is so far 
 
8 THE PSYCHOLOGY OF LEARNING 
 
 advanced when the child enters school, that he can 
 make with some degree of precision, any kind of mark 
 he is shown or told to make. He can also make any 
 sounds he hears or is told to make. These last two 
 statements are not wholly true, for there will be some 
 movements and sounds not wholly mastered. But most 
 of the movements and vocal responses have been mas- 
 tered, so that little trial and error learning is left to 
 be done. Many of the child's responses will be crude 
 and will have to be perfected through practice. The 
 method of perfecting them is the trial and error method. 
 The fact is, that when a child enters school much learn- 
 ing has already taken place. He has partially mastered 
 most of the responses that will ever be required of him. 
 What remains is to couple these responses to their 
 appropriate stimuli, and this is done by what may be 
 called the ideational method. This method is possible 
 only after the movement involved has been attained by 
 the trial and error method and connected with its idea 
 as its adequate stimulus. As previously explained, 
 when we speak of an idea as a stimulus, we mean the 
 cortical activity which is the neural correlate of the 
 idea. 
 
 In the case of knowledge, the initial connection be- 
 tween the ideas is secured by simultaneous experience. 
 If we wish two ideas to be joined together in our minds, 
 so that when one of the two ideas comes, the other fol- 
 lows, w^e must experience these two ideas together. And 
 the oftener the ideas are experienced together, the more 
 surely will one arouse the other. Suppose I wish to 
 teach a child the names of the parts of a flower and the 
 functions of the essential parts of the flower. I show 
 him a flower and tell him its name. The idea of the 
 
THE NATURE OF LEARNING 9 
 
 thing becomes associated with the name through 
 simultaneous experience. I then point out the different 
 floral organs, the calyx, corolla, stamens and pistil. As 
 the child looks at each organ, I pronounce the name of 
 the organ upon which his attention is fixed. In this 
 manner the things and their names become coupled 
 together. Then in a similar manner we couple the 
 names of the parts with the functions of the parts. In 
 knowledge-getting in all its various fields, our only 
 method of connecting ideas together is by experiencing 
 those ideas together in a state of attention. Since an 
 idea can enter into any number of such connections, 
 all of our knowledge becomes connected or related, so 
 that we are able to go from any idea to any other idea 
 through intermediary ideas. For example, horse is 
 connected with buggy through simultaneous experience, 
 similarly buggy with city, city with Athens, Athens 
 with Homer, Homer with the Iliad, Iliad with school, 
 school with a certain classmate, this classmate with 
 death, death with heaven. Similarly all the items of 
 our experience are united so that we can go either 
 directly or indirectly from any bit of experience to any 
 other, although the only method of uniting these bits 
 of experience is through simultaneous experience. 
 
 The above account of the method by which stimuli 
 are first connected wdth their responses and ideas with 
 other ideas is objective and to some extent superficial. 
 We have described merely what is seen on the outside. 
 \Vhat goes on inside is not entirely clear. Why two 
 cortical processes which underlie two ideas simultane- 
 ously experienced should be so related that when later 
 one of these processes is revived it should revive the 
 other, is not known. Probably two simultaneous brain 
 
10 THE PSYCHOLOGY OF LEARNING 
 
 processes are to some extent one process, and flow into 
 one another, so that they become more or less a unity. 
 If such is the case, then later when one of the pro- 
 cesses is revived it revives the other because it is really 
 a part of the other, or rather they both constitute but 
 one process. 
 
 In the case of habit, it may well be asked Vv^hy, when 
 a child tries for the first time to make the letter a, 
 whatever the first response, does he not continue to 
 make the same marks over and over again instead of 
 making somewhat different marks at each trial? It is 
 possible that the stimulus changes somewhat with each 
 trial. The child sees that the first result is not the 
 result desired, this very fact makes the stimulus for 
 the second attempt a somewhat different stimulus. 
 Moreover, the resistances along the various possible 
 routes are doubtless not very different so that the 
 slightest change in resistance makes the stimulus take 
 a somewhat different route producing a different result. 
 Probably the fatigue toxins are sufficient to change the 
 resistance temporarily. Possibly also the physiological 
 correlates of the feeling element involved have their 
 effects in varying the resistances and shifting the direc- 
 tion of the excitation. These suggestions are entirely 
 theoretical, and we must confess that we are ignorant 
 of the nature of the neurological processes which con- 
 stitute the causes which we are seeking. For the pres- 
 ent we must be content with the objective statement of 
 facts given above. Fortunately this objective state- 
 ment is sufficient for our practical purposes. 
 
 Function of Habits. — Habit is the basis of efficiency. 
 It insures, as much as anything can insure, that the 
 right response follows upon the appropriate stimulus. 
 
THE NATURE OF LEARNING 11 
 
 Habit enables us to perform an act with greater accu- 
 racy, greater speed, less waste of energy, and conse- 
 quently with less fatigue. Motor efficiency is impossi- 
 ble except through habit. In whatever field our work 
 may be, whether in carpentry, masonry, farming, 
 blacksmithing, engineering, or cooking, we become effi- 
 cient only through habituating the processes. There 
 is no other way. It is true that habituation resists 
 change. Through habit we become set and fixed, but 
 this is the price that we must pay for efficiency. 
 
 If we consider the function of habit from the mental 
 side, the subjective side, w^e find that habit removes 
 actions to the control of lower cortical levels, relieving 
 the higher levels which are thereby freed for other 
 processes which may go on simultaneously. In other 
 words, habit removes an act from the state of focal 
 attention, leaving the higher attention levels for othev 
 processes. For example, when we are learning to run 
 a typewriter, the operation of the machine demands 
 all of our attention. After we have become expert in 
 running the machine, the mere mechanical work 
 demands little or no attention. 
 
 Function of Knowledge. — The chief function of 
 knowledge is to put meaning into the world. The get- 
 ting of knowledge consists in learning the names of 
 things and the functions of things. It is only the fact 
 that we live a social life and need to communicate with 
 one another that makes it necessary to learn the names 
 of things. If we lived an individual, solitary life, we 
 should not need to know the names of things, therefore 
 the essential thing in knowledge is the coupling of the 
 thing with its functions. Now all things in the world 
 are related, and through experience we leam the func- 
 
12 THE PSYCHOLOGY OF LEARNING 
 
 tions of things, or in other words, their characteristics, 
 what they do, what v/e can do with them. The world 
 has unity and order in it. Through our experience, 
 we get this unity, this order ; we learn the several uses 
 to which the various things in our environment can be 
 put. For example, Vv^e have experience with a substance 
 which we name 17^071. We learn that it is hard, tough, 
 heavy, that it gets red when heated, white and malleable 
 when heated still more. We learn that we can hammer 
 it till it has a shaip edge and then v/e can cut with it; 
 that we can mould it into various shapes and make it 
 into a multitude of useful instruments and implements. 
 In a similar way we learn of all the substances and 
 organisms of our environment. We come to know the 
 world as it is. This knowledge of the world gives us 
 control of the world. The function of knowledge is, 
 then, to put meaning into the world, and through this 
 meaning to give us control of the world. We build an 
 inner, thought world, which corresponds to the 
 outer world. The closeness of the correspond- 
 ence between our inner world and the outer 
 world indicates our approach to the truth. If 
 the correspondence is not fairly complete and accurate, 
 we are in error, we can not use the outer world. We 
 can not use it because we do not really know it. We do 
 not have its true meaning. 
 
 Plasticity. — By plasticity as uesed here, we mean the 
 ease with which we can be changed, more particularly 
 the ease with which new bonds can be established in 
 our central nervous system. As we grow old we lose 
 plasticity chiefly for the reason that in our earlier 
 years definite responses are established for most of the 
 situations of life. The passage of the same excitation 
 over the same set of neurones, time after time, produc- 
 
THE NATURE OF LEARNING 13 
 
 ing the same action sets up a path which resists change. 
 By the time we have reached middle life, most of the 
 situations of life have occurred repeatedly; forms of 
 response more or less definite have been established 
 for all of them. Therefore, in familiar situations we 
 become more set with each succeeding year. However, 
 we do not lose plasticity in unfamiliar situations, as far 
 as is known, until the decadence of the latter part of 
 life sets in. But after we reach middle life there are 
 not many unfamiliar fields. There are then very few 
 situations which we do not approach with some bias 
 or prejudice. Apart from the resistance of bonds 
 already formed our ability to form new bonds increases 
 up to maturity, and probably does not decrease till after 
 middle age. This statement is based on a comparison 
 of the learning capacity of men and women in their 
 forties with that of children of high school age and 
 younger. In fields wherein men and women are not 
 incapacitated and hindered by virtue of prejudicial 
 bonds already formed they learn as readily as do boys 
 and girls, much more readily than young girls and boys. 
 Man as the Result of His Learning. — Let us now con- 
 sider the importance to our lives of habits and knowl- 
 edge. Thomdike has truthfully said that the most 
 important thing about human nature is that it can be 
 changed. Let us couple with this statement another 
 common saying, namely, that a man is the sum of his 
 tendencies. A careful consideration of these two facts 
 gives us an idea of the importance of education. What 
 we are at maturity depends upon the habits and knowl- 
 edge that we have. Acquiring these habits and get- 
 ting this knowledge constitute our education. Whether 
 w^e are a Methodist, Presbyterian or Catholic depends 
 
14 THE PSYCHOLOGY OF LEARNING 
 
 upon our education. Whether we are a repubUcan, 
 democrat or socialist depends upon our education. 
 Whatever motor performances we are capable of de- 
 pends upon our education. Our ideals of life, of work, 
 of duty depend upon our education. Briefly, all our 
 skill, all our knowledge is merely the result of the bonds 
 that have been formed in our central nervous systems. 
 In a very true sense a man is made. He is the product of 
 all the forces that act upon him, particularly of those 
 that act upon him in his earlier years. We do not mean 
 to minimise the importance of original nature. The same 
 forces do not make the same result out of two different 
 children. W^e can not make the same kind of chair out 
 of pine that we can out of oak, nor can Vvc make the 
 same kind of man out of one child that we can out of 
 another. What we are at maturity depends upon the 
 modifications that have been wrought upon original 
 nature. Bringing about these modifications constitutes 
 our education, and education is important to the extent 
 that these modifications are important. And they are 
 of tremendous importance in the life of every indi- 
 vidual. The child comes into the world with a multi- 
 tude of original tendencies and capacities. Upon these 
 the forces of the environment, of home, school, com- 
 panions, and the whole animate and inanimate world 
 work, forming and moulding and fashioning the indi- 
 vidual into the thing that he is to be, namely a being 
 with a background of feelings and passions and in- 
 stincts, to which is added a multitude of acquired ten- 
 dencies to act in definite ways to the various situations 
 of life: a being with myriad ideas all interrelated, each 
 idea coupled with others; a being with ideals, ambi- 
 tions, desires; a l^eing of hates and loves and jealousies, 
 
THE NATURE OF LEARNING 15 
 
 — all the resultant of the original tendencies provided 
 by heredity and the forces of the environment that 
 have acted upon the individual. The changes that are 
 wrought in us constitute the results of learning. The 
 investigation of their nature and of how they come 
 about is the purpose of this book. 
 
Chapter II. 
 THE LEARNING CURVE. 
 
 Progress in learning is ordinarily shown by a curve. 
 We must now see how such curves are constructed, 
 how they should be interpreted, notice the various 
 types of curve, and the laws of learning which they 
 exhibit. We must study also the various factors which 
 enter into them, and which determine their upper limit. 
 
 The Construction of a Learning Curve. — The purpose 
 of a learning curve is to show the increase in efficiency 
 due to practice. It is customary to indicate the succes- 
 sive practices on the horizontal axis, as BC in Fig. 1. 
 The scale for the scores is indicated on the vertical axis, 
 AB. The scores are shown by the dots above the num- 
 bers on BC which represent the successive practices. 
 The line DE joining these dots is the learning curve. 
 Its rise indicates the increasing efficiency due to prac- 
 tice. 
 
 Another method of constructing the curve is shown 
 in Fig. 2. The curve is constructed as follows: In- 
 stead of representing successive practices by points on 
 the base, we represent them by successive distances. 
 The scores are represented by horizontal lines, drawn at 
 the proper height and of the same length as the dist- 
 ances on the base which indicate the practices. Either 
 of these forms of curve shows the facts as well as the 
 other, but if curves are to be dra\^Ti with the same 
 axes for purposes of comparison, the form of curve 
 shown in Fig. 1 is more convenient. 
 
 [16] 
 
THE LEARNING CURVE 
 
 17 
 
 Figure 1. Learning Curve. Card-sorting, fifteen boxes, live cards 
 to the box, experiment continued five days, four sortings first day, 
 eight sortings on each succeeding day, 12 subjects. Every sixth 
 record is taken for the construction of the curve. The records are 
 in terms of cards sorted per minute. 
 
 />^z. 
 
 ^ / 2 3 4 5 6 7 C 
 
 Figure 2. Learning Curve. Data same as for Figure 1. 
 
18 THE PSYCHOLOGY OF LEARNING 
 
 In figures 1 and 2, increasing efficiency is shown by 
 ^ rise in the curves. The curves indicate the amounts 
 of work done in successive equal periods of time. An- 
 other method of constructing the curve is to show the 
 amount of time required to do successive equal amounts 
 of work. By this method, increasing efficiency is shown 
 by a fall in the curve. It may be illustrated by the same 
 card-sorting data used in constructing the graphs in 
 figures 1 and 2. In the card-sorting experiment from 
 which this data was obtained, seventy-five cards w^ere 
 sorted into fifteen pigeon holes or card trays. Each 
 tray was numbered, and the cards were numbered cor- 
 respondingly, five to each tray. At each practice, all 
 the seventy-five cards were distributed. With each 
 practice the time required for sorting the cards became 
 less and less. This is shown in Fig. 3. To convert the 
 data into the form used in figures 1 and 2, the number 
 of cards sorted per minute in each practice was com- 
 
 So 
 
 /"^^j 
 
 Figure 8. Data from same experiment as iu figures 1 and 2. 
 Records in terms of number of seconds required to sort the 75 cards. 
 Increased efficiency is shown by the fall in the curve. 
 
THE LEARNING CURVE 19 
 
 puted. Therefore figures 1 and 2 show the increasing 
 number of cards sorted per minute in the successive 
 practices, v^hile figure 3 shows the decreasing amount 
 of time required to sort the seventy-five cards in the 
 successive practices. In the discussion of learning 
 curves which follows, we shall have in mind the types 
 of cui-ve shown in figures 1 and 2, in which efficiency is 
 shown by a rise in the curve. 
 
 The Rise of the Curve. — ^Why does the curve rise? 
 To what is the increased efficiency due? Increase in 
 efficiency from practice is due in general to three 
 causes : (1) The establishing of the bond between stimu- 
 lus and response. This bond, in physiological terms, is 
 the shortest possible route between stimulus and re- 
 sponse. (2) The second factor is decreased resistance 
 in the established neural bond or route. (3) The third 
 factor is the adaptation of the muscles to the special 
 movements required. In the present discussion, we are 
 considering learning of the habit-forming type, and not 
 ideational learning in which the motor element is un- 
 important. 
 
 We shall now illustrate the various learning factors 
 from the card-sorting experiment. We sit down to the 
 experiment for the first time. We take the pack of 
 cards in our left hand and remove a card with the right 
 hand. Let us suppose the card is number seventeen. 
 We look over the box numbers until we find number 
 seventeen, and then deposit the card into this box. We 
 take another card. Suppose it is number 20. We must 
 search for box number 20, and so proceed until all the 
 seventy-five cards are distributed. When we have fin- 
 ished, we have put five cards into each box. The first 
 card sorted was number 17. After a few seconds we 
 
20 THE PSYCHOLOGY OF LEARNING 
 
 came to another card numbered 17. We may have 
 remembered the location of this box, more likely we did 
 not. But before the first experiment was over we did 
 learn the location of some of the boxes and did not have 
 to search for them. With each successive sorting, we 
 remembered more and more of the boxes, until finally 
 we knew the location of all of them and did not have to 
 search for any of them. However, it must be pointed 
 out that all learning is a matter of degree. I have said 
 that we come to know where all the boxes are. We do, 
 but at first we know it poorly. The associations come 
 slowly. While we do not have to hunt for the boxes, 
 the kinaesthetic or visual idea of their location comes 
 to us slowly, and the movements therefore follow slowly. 
 • With more practice, the movements follow their stimuli 
 more quickly, more definitely, more surely. The asso- 
 ciative processes run their course faster and faster. 
 
 In the early stages of practice, after we have learned 
 the location of all the boxes, we may at any time be 
 temporarily unable to recall the location of a box which 
 we remembered on a former trial. This is because the 
 bonds are only poorly established, and at any moment 
 may break down and fail us. With increasing practice, 
 this failure of a particular association to run its proper 
 course happens less and less, and the certainty that all 
 the associations will take place quickly becomes greater 
 and greater. But, as pointed out in Chapter I, this cer- 
 tainly is never absolute. 
 
 In the early practices, while the neuro-muscular co- 
 ordinations are weakly established, the hand often 
 starts the wrong way, but before it reaches the box, 
 we recall the right place and have to make a second 
 movement. With increasing practice, these false move- 
 
THE LEARNING CURVE 
 
 21 
 
 '• 
 
 4oo 
 
 Figure 4. Learning Curve showing the effects of long continued 
 practice. Card-sorting, 30 boxes, five cards to the box. The records 
 from which the curve is made are the reciprocals of the scores in 
 seconds. These reciprocals are made proportional to the number 
 of cards per second. The pr^jctice was first for 20 successive days, 
 five practices the first day ; seven, the second ; eight, the third and 
 fourth ; nine, on the fifth, sixth, seventh, and eighth ; and ten on 
 all succeeding days- Fourteen days later, the practice was resumed 
 for two days, and seven days latter, for four days. The records 
 attended on these latter practices and indicated by the letters 
 a, b, and c. Practice was continued at intervals for four months 
 longer, a record of 935 being reached, indicated by d on the curve. 
 
22 THE PSYCHOLOGY OF LEARNING 
 
 ments become fewer and fewer. With every -day of 
 practice, it becomes more and more certain that when 
 we take a card from the pack, the hand will go uner- 
 ringly and with no loss of time or waste of motion, to 
 the appropriate box. 
 
 It is surprising to one unacquainted with work in the 
 psychology of learning to discover how much improve- 
 ment is possible after one has learned the location of all 
 the boxes. In a few days one knows the boxes so that 
 he does not have to hunt for them, but improvement 
 will continue for months. This fact is illustrated in 
 Fig. 4. Under the author's direction. Miss Rose Ann 
 Howe sorted cards for several months. In her experi- 
 ment, she sorted 150 cards into thirty boxes, five cards 
 to the box. After a few days she knew the location of 
 all the boxes, nevertheless improvement continued for 
 several months. This fact is even better illustrated by 
 the substitution experiment, performed by the same 
 subject. In this substitution experiment, letters were 
 substituted for the nine digits. There were but "nine 
 bonds to form. After two or three five-minute prac- 
 tices, these nine bonds are established. However, im- 
 provement continues for twenty days, as shown in Fig. 
 5. The experiment consisted in transcribing columns 
 of numbers into letters according to a key furnished 
 the subject. Practice was continued for twenty-five 
 minutes at a sitting each day. The key was learned 
 during the first day's practice. On the second day the 
 average score was 280 substitutions in five minutes. On 
 the twentieth day the score was 720 substitutions in 
 five minutes. Now since the key was known on the sec- 
 ond day, why the greater score on the twentieth day? 
 The answer again is that learning is a matter of degree. 
 
THE LEARNING CURVE 23 
 
 Figure 5. Digit-symbol Substitution, 20 days of practice, five 
 practices a day, five minutes at a practice. Records, average 
 correct substitutions in five minutes for each day. The days are 
 represented on the horizontal axis, the scores on the vertical. 
 
 While the key was known by the end of the first day's 
 practice, it was known better on each succeeding day. 
 The bonds became better and better estabhshed with 
 practice; the certainty that the proper association 
 would come instantly became ever greater and greater. 
 Moreover, the muscles became adapted to their work. 
 With practice, there was less waste of energy through 
 false and awkward movements; the muscles gained 
 strength and facility in co-ordination. 
 
 Again we have the same three factors of explana- 
 tion : forming the primary bonds, strengthening these 
 bonds through use, and adapting the muscles to their 
 work. In both the card-sorting and the substitution 
 experiments, the neuro-muscular systems involved be- 
 come organised so that the stimuli run their courses 
 with the least possible obstruction and expenditure of 
 energy. To summarise : Improvement sho\\Ti by the rise 
 
24 THE PSYCHOLOGY OF LEARNING 
 
 of the curve comes from establishing the proper bonds, 
 eliminating the wrong or useless bonds, strengthening 
 the established bonds through use, and the adapting 
 of the muscles involved to the movements required in 
 the habit. 
 
 We shall now give a further illustration of the whole 
 procedure by the process of learning to add. Let us 
 take a problem of four numbers, three digits each: 
 
 987 
 789 
 654 
 456 
 suppose we know merely how to count. We take 
 six, count four more and have ten. We then count nine 
 more and have nineteen, then seven more and have 
 twenty-six. With practice we come to know that six 
 and four are ten and then do not have to count, we look 
 at six and four, and say "six and four are ten " then 
 having learned the other combinations, say ''ten and 
 nine are nineteen, nineteen and seven are twenty-six." 
 Later we shorten the process and say merely ''ten, 
 nineteen, twenty-six." Later still, the six and four be- 
 come a unity and mean ten as definitely and directly 
 as one figure by itself means "six" and the other "four." 
 So the nine and seven come to mean "sixteen," then the 
 addition becomes "ten and sixteen are twenty-six." 
 
 In the early stages, we get the sums of the second 
 and third columns by the slow process of counting, and 
 the process of carrying from the first to the second 
 and from the second to the third must be learned. With 
 practice, however, we can add two columns at once as 
 fast as formerly we could add one column. We can also 
 learn to add three columns at once with great speed 
 and accuracy. 
 
THE LEARNING CURVE 
 
 25 
 
 In adding- we first learn all the possible combinations 
 of the nine digits. These are the primary bonds, the 
 basis of all later schemes of adding. Through practice 
 we come to eliminate all useless bonds, as in saying 
 "six plus four are ten.'' We simply look at the six and 
 four and say "ten." With the ten in mind, we look at 
 the nine and say "nineteen." With the nineteen in 
 mind, we look at the seven and say "twenty-six." 
 
 After we have learned all the primary combinations, 
 practically unlimited improvement is still possible, 
 through strengthening the primary bonds, and com- 
 
 FiGURE 6. Pbactice Curve, dealing 53 playing cards into a box, 10 
 times a day for 15 days. The records shown on the vertical axis are 
 the daily averages, expressed in seconds. It will be seen that the 
 time required to deal the pack of cards was reduced to less than a 
 third of the time required on the first day. The graphs show the 
 records of two women, but of whom were used to dealing cards. 
 The amount of improvement shovra, for so simple a function, is 
 very great. 
 
26 THE PSYCHOLOGY OF LEARNING 
 
 bining into unitary processes what are at first complex 
 processes. As shown above, after we have practiced 
 a while, 6+4 are not two things but one thing and 
 mean "ten," directly and immediately. With further 
 practice, larger and larger groups of digits coalesce, be- 
 come a unitary process, and mean their sum. 
 
 All habit-formation of the motor type is essentially 
 the same. In general, the same factors are involved. 
 The illustrations given above are typical of the processes 
 in all complex habits. 
 
 The Limit of Improvement.-— In a simple habit the 
 limit is soon reached. What this limit is, is determined 
 by the subject's reaction time. Suppose that, instead of 
 sorting cards, we take a pack and put all of them into 
 the same box as quickly as possible. The neural bond 
 is soon established and we reach a limit determined by 
 the time required for the nen^ous impulse to travel 
 alonfif the necessary route. Such a curve is shown in 
 Fig. 6. In general, the more bonds involved in the 
 habit, the longer it takes to perfect them. If the bonds 
 required are many, say those involved in sorting cards 
 into one hundred boxes, improvement would continue 
 for many months. 
 
 In such a process as addition, improvement is prac- 
 tically unlimited, for higher and higher orders or habits 
 can be formed indefinitely. In such a case, improve- 
 ment comes not merely from perfecting habits, but 
 from developing more and more efficient habits. But 
 in all cases in which the habit is definite and can not 
 be supplanted by a more efficient habit, there is a limit 
 which may be called the physiological limit. This limit 
 varies with different individuals, and, as above stated, 
 depends upon their reaction times. In the card-sorting 
 
THE LEARNING CURVE 27 
 
 experiment, the correlation between reaction time and 
 efficiency becomes higher and higher with succeeding 
 practices. 
 
 Practical Limits. — As pointed out by Thorndike's few 
 people in the ordinary pursuits of life, come anywhere 
 near their possible limits. In typing, stenography, ac- 
 counting, telegraphy, and the various other perform- 
 ances of the business and industrial world, workers 
 reach a proficiency that enables them to do their work 
 reasonably well and hold their jobs. They are not 
 willing to put forth the effort that would carry them 
 to a higher plane of efficiency. Most of the workers 
 are working on an efficiency plane much below what is 
 possible for them. One in a hundred, perhaps, puts 
 forth such effort and passes up to a higher plane of 
 performance. One in a thousand, having great ability 
 and great ambition, puts forth the effort that places 
 him among the leaders in his trade or profession. 
 
 It is well that the young know this important fact, 
 for not only do laborers and clerks work on a level far 
 below their possible limit, but students, as a rule, are 
 content to continue on a level that just barely enables 
 them to "pass." In language they only poorly know 
 the declentions, the paradigms, the conjugations, and 
 the vocabularies. In mathematics, the tables, the 
 formulas, the rules and fundamental principles are not 
 sufficiently mastered. The case is similar in science. 
 The student should have pointed out to him that the 
 elements can be so mastered as to enable him to pro- 
 ceed to a higher level of achievement. Means of prac- 
 tice and drill should be devised to enable the student 
 
 'E. L. Tliorndike, Educational psychology, briefer course, p. V 
 
28 THE PSYCHOLOGY OF LEARNING 
 
 to pass to a higher level. There is no short-cut ; there 
 is no easy way; there is no substitute for drill, prac- 
 tice, repetition. Easy, confident, and efficient perform- 
 ance in any field comes only from complete habituation 
 of the elementary processes. 
 
 Different Forms of Curve. — ^The form of the learning 
 curve depends upon the nature of what is learned. Usu- 
 ally, the initial rise is rapid, with slower increments 
 later. This gives a curve convex on its upper side as in 
 Fig. 5. If learning is slow at first and faster later, the 
 resulting learning curve is concave on its upper side as 
 in Fig. 7. Whatever the form of the curve at first, it 
 eventually becomes level or nearly so. As we approach 
 
 Figure 7. Learning Curve, concave upward. Adapted from Swift 
 Practice was in ball-tossing. Scores shown on vertical axis rep- 
 resent successes. The successive days of practice are represented 
 on the horizontal axis. A concave learning cui-ve indicates increas- 
 ing increments from the effects of practice. 
 
THE LEARNING CURVE 20 
 
 our limit in any particular habit, any improvement is 
 at the cost of much greater effort than was the case 
 in an earlier stage of habituation. 
 
 In the psychological literature there has been much 
 discussion of "spurts." By a spurt is meant increased 
 efficiency as shown by an abrupt rise of the curve. With 
 some subjects and some work there is an initial spurt. 
 Sometimes there is a final spurt, particularly if the 
 subject knows that the end of the work period is near. 
 At any time during a work period a spurt may appear, 
 due to a combination of favorable circumstances. The 
 relative amounts of work done at different parts of a 
 work-i rriod depend upon several factors, particularly 
 upon the stage of habituation and how fatiguing the 
 work is. In the early stages of habituation, there are 
 considerable practice effects during the work period 
 which make the later records of, the period high. If 
 we are near the limits of practice, the practice effects 
 during a work period will be slight. In the latter case 
 any difference in the amounts of work done at different 
 parts of the period will be due to the effects of fatigue 
 and of warming up. In complicated motor perform- 
 ances, such as type-writing, the best record is not made 
 at the first part of a work-period. Only after i ^ittle 
 practice can we do our best. 
 
 In Miss Howe's card-sorting experiment, cited above, 
 for the 18th, 19th, and 20th days, the average efficien- 
 cies for each of the ten sortings made at a sitting were: 
 775, 802, 804, 824, 802, 808, 796, 802, 805, 780. Maxi- 
 mum efficiency was not reached till the fourth sorting, 
 when the speed was 6.3 per cent, better than it was 
 the first sorting. The record for the first sorting was 
 the lowest of the ten. 
 
30 
 
 THE PSYCHOLOGY OF LEARNING 
 
 Figure 8. Card-sorting. The upper graph shows the course of 
 efficiency during an hour's work in card sorting after the habits 
 involved are well fixed. It shows a period of warmiug-up leading 
 to maximal efficiency on the fourth practice. The lower curve 
 indicates the course of efficiency during an hour's practicebefore 
 the habits are well fixed. Maximal efficiency is reached on the 
 eight sorting. In tl>e latter case, the effects of practice raise 
 the latter scores. The last two scores fall because fatigue offsets 
 the practice effects. In the former case, there is little improve- 
 ment due to practice in the course of an hour. 
 
 ■-.^f^- 
 
 Figure i>. Efficiency Curve for Card-sorting. Solid line from 
 the actual data of one subject ; broken line, a smoothed curve 
 constructed from the same data, by averaging each record with the 
 adjacent records. The first and last record were doubled, added 
 to the one adjacent record, and the sum divided by three. The 
 number of cards sorted per minute is indicated on the vertical axis. 
 
: THE LEARNING CURVE 31 
 
 On the 9th, 10th and 11th days, while improvement 
 was still marked, the records for the ten practices at 
 a sitting were 685, 702, 721, 709, 728, 725, 718, 741, 
 725, 720. The highest record was made on the 8th 
 sorting, when the speed was about 8 per cent, higher 
 than on the first. These facts are shown graphically 
 in figure 8. 
 
 Smoothing the Learning Curve. — Ordinarily there 
 are fluctuations in the rise of a learning curve. It rises 
 fast, then perhaps more slowly; it may even fall, for 
 just as a combination of favorable factors makes it 
 rise, so a combination of unfavorable circumstances 
 may make it fall. There are always many independ- 
 ently variable factors that combine to produce a learn- 
 ing record. These factors are variously combined at 
 different moments of our procedure and produce a 
 curve of fluctuating rises and falls. The general ten- 
 dency of the curve is better shown if, instead of show- 
 ing the actual records, we construct a smoothed curve. 
 There are various ways of doing this. We can do it 
 roughly by indicating the actual records on our co- 
 ordinate paper by dots and then drawing a cui-ve that 
 throws as many of the dots on one side as on the other. 
 The dots show the actual records, the cui^e shows the 
 general tendency of the records. Another method is 
 to eliminate the temporary fluctuations of the curve 
 by averaging each record with the adjacent records 
 above and below. The first and final records are doubled 
 and combined with the one adjacent record, and the 
 sum divided by three. 
 
 Leai-ning Plateaus. — A plateau or level place in the 
 learning cui^ve indicates a period of no progress in 
 learning. Such plateaus nearly always appear in the 
 
32 THE PSYCHOLOGY OF LEARNING 
 
 actual work of learning, in school work as well as in 
 laboratory experiments. Various explanations for the 
 appearance of the plateau have been given. Bryan, in 
 his early work on learning, suggested that plateaus 
 were due to the fixing of lower order habits. Progress 
 due to higher order habits could not come till the lower 
 order habits were fixed. But this can not be the expla- 
 nation, for fixing of the lower order habits v/ould itself 
 give more efficiency, and should give a continuous rise 
 of the curve. Book's suggestion that the plateau is 
 due to a flagging of interest, a letting up of effort, is 
 doubtless nearer the truth for the plateau may appear 
 in simple experiments involving a single definite habit. 
 The fact seems to be that the plateau is not a necessity 
 in learning, theoretically, although in practice it usually 
 appears. As long as a learner maintains a high degree 
 of attention and puts forth maximum effort, he con- 
 tinues to improve till he reaches his limit. But main- 
 taining a high state of attention and putting forth 
 maximum effort are impossible over long periods of 
 time. The ordinary learner usually does his best for 
 a time, then because of fatigue, or lack of attention, or 
 loss of interest, he relaxes and works at a lower point 
 of efficiency. For a time, the records may actually fall 
 below previous records. This phenomenon may appear 
 during the experiment of a single sitting and be a mat- 
 ter of minutes. It may appear in a long series of ex- 
 periments and be a matter of days. 
 
 The plateau has much significance to education. When 
 it is due to loss of interest and lack of attention, prac- 
 tice is not then profitable. At such a time, practice 
 should stop or means be found to renew attention and 
 effort. Children lose interest very quickly, and soon 
 
^. THE LEARNING CURVE 83 
 
 cease to put forth eifort in continued work of the same 
 kind. Pauses in habit-formation are valuable not only 
 because they lead to a renewed interest and effort when 
 the work is taken up again, but according to some 
 writers, they lead to a weakening of inhibitions which 
 had interfered with the habit. Moreover, in habit- 
 formation, time is an element which can not be ignored. 
 Some sort of fixation goes on in the nervous system; 
 this fixation requires time and can not be unduly hur- 
 ried. As shown in a later chapter, practice beyond a 
 certain amount at one sitting is valueless. And in prac- 
 tice extending over many days, only a certain amount 
 of improvement is possible, no matter how hard we 
 work. 
 
 EXPERIMENTS AND EXERCISES. 
 
 All the principles discussed in this chapter should, 
 as far as possible, be illustrated by experiments. Any 
 learning experiment v/ill suffice, but for economy of 
 time, an experiment should be selected that will show 
 rapid improvement. Several experiments are described 
 below. The instructor can select those suited to his 
 needs, or devise others. 
 
 1. Class experiment. Material, Pyle's Digit-symbol 
 substitution test blanks. Method : Distribute about six- 
 teen test sheets to each student, placing them face down 
 before the student. Explain to the students that they 
 are to substitute letters for the nine digits, in accord- 
 ance with the key shown at the top of the test sheet, 
 and that they are to work as fast as possible. Work in 
 five-minute periods, and let the score be the number of 
 substitutions correctly made in five minutes. Eight 
 such scores can probably be made in a one-hour period. 
 The scoring should be left till the experiment is finished. 
 
34 THE PSYCHOLOGY OF LEARNING 
 
 2. From the data obtained in experiment 1, construct 
 individual learning curves similar to figures 1 and 2. 
 
 3. Construct a smoothed curve as described in the 
 chapter. 
 
 4. Construct a learning curve from the class averages 
 for each five-minute period. Note that it is smoother 
 than the individual learning curves. Why ? 
 
 5. Laboratory experiment. Card-sorting. Material, 
 ordinary playing cards. Method : Shuffle the cards, then 
 sort them into four piles according to suit, ^. e., hearts 
 together, spades together, clubs together, and diamonds 
 together. Detemiine the time for each sorting. Con- 
 struct a cun^e similar to figure 3. 
 
 6. Laboratory experiment. Material, card-sorting 
 box and cards. Method: Use only one row of boxes, 
 five cards to each box, and determine the time required 
 for each sorting. The cards must be thoroughly shuffled 
 before each sorting. Shuffling is facilitated if the cards 
 are taken up after sorting, promiscuously, one at a time. 
 Ten sortings can be finished in an hour. Construct 
 individual learning curves similar to figure 3. By com- 
 puting the number of cards sorted in a minute, curves 
 can be made similar to figures 1 and 2. 
 
 7. I;aboratory experiment. Material, mirror-writing 
 apparatus. Method: The hand of the writer is to be 
 covered from view. He is to see his hand as reflected 
 in the mirror, which is vertical before him. The sub- 
 ject is to write in script in such a way that the letters 
 appear in the mirror as they would on a sheet of paper 
 held vertically before him. Let the subjects copy the 
 material of the first paragraph in this chapter. Record 
 the number of letters written legibly in each successive 
 minute. Construct individual learning curves similar 
 to figure 1. How do the curves differ from figure 1? 
 
THE LEARNING CURVE 35 
 
 REFERENCES. 
 
 Batson, W. H., Acquistion of skill, P. Mon, No. 91, 1916. 
 
 Book, W. F., The psychology of skill ivith special reference to^ 
 its acquisition in typeioriting , 1908. 
 
 Bryan, W. L., and Harter, N., Studies in the Telegraphic lan- 
 guage; the acquisition of a hierarchy of hahits, P. R,, vi, 1899, 
 345. 
 
 Chapman, J. C, and Nolan, W. J., Initial spurt in a simple 
 Tfiental function, A. J. P., xxvii, 1916, 256. 
 
 Chapman, J. C., and Hills, M. E., Positive acceleration in im- 
 provement in a complex function, J. Exp. Psych., i, 1916, 494. 
 
 Peterson, J., Experiments in ball-tossing; the /Significance of 
 Learning Curves, J. Exp. Psych., ii, 1917, 178. 
 
 Phillips, F. M., Comparison of the work done in the successive 
 minutes of a ten-minute practice period in the fundamentals of 
 arithmetic, J. E. P., vii, 1916, 271. 
 
 Swift, E. J., Studies in the psychology and physiology of 
 learning, A. J. P., xiv, 1903, 201. 
 
 Thorndike, E. L., Educational psychology. Briefer Course-, 
 1914, Chapters xiv and xvi, a Systematic Treatment. 
 
i 
 
 Chapter III. 
 ECONOMICAL LEARNING. 
 LENGTH AND DISTRIBUTION OF PRACTICE PERIODS. 
 Learning is connecting. After the initial connection 
 is secured, we must strengthen and perpetuate it 
 through repetition or practice. The first important 
 problem in the economy of learning is concerning the 
 proper length and distribution of practice periods. By 
 proper length and distribution, we mean that length 
 and distribution which gives the most return in effi- 
 ciency for the amount of time spent in practice. Our 
 problem, then, is to find what length and what distri- 
 bution of practice periods will give the best returns. 
 Or, to state the problem in still a different way : after 
 we have secured the passage of a stimulus over into its 
 appropriate response, how many times should this pas- 
 sage be repeated before we rest? And how long an 
 interval should elapse before the processes are re- 
 peated ? We turn to the experimental evidence. 
 
 Experiments with Nonsense Syllables. — The early ex- 
 periments of Ebbinghaus and Jost showed the advan- 
 tage of short periods of practice distributed over a long 
 time as compared to longer periods of prcatice distrib- 
 uted over shorter lengths of time. All the experiments 
 that have since been performed in this field have in 
 general confirmed the earlier work and have made it 
 clear that for every kind of learning there is a proper 
 
 [36] 
 
ECONOMICAL LEARNING 37 
 
 length of practice period, and for the different stages 
 of habituation, there is also the proper distribution of 
 practice periods. 
 
 Jost, comparing ten repetitions a day for three days 
 with thirty repetitions in one day, found a saving of 
 fifteen per cent., from the shorter periods. From the 
 first it has been evident there is a certain advantage in 
 spreading practices out, temporally, rather than in con- 
 centrating them. There is an aspect of learning that 
 requires time, an aspect that resembles growth. Growth 
 is a process that can be hastened only to a limited ex- 
 tent. Such seems to be the case with habit-formation. 
 A certain amount of practice at one time is efficient in 
 fixing a habit, more practice at the same time, without 
 intennission, does no good, is time wasted as far as 
 fixing the habit is concerned. 
 
 More recently Perkins has continued the study with 
 nonsense syllables. She states her problem as follows : 
 "My own experiments were designed to ascertain how 
 much further in the distributioji of readings one might 
 go than Jost had gone before a limit would be reached 
 in learning nonsense syllables. The best arrangement 
 of readings involves two questions, — how many repe- 
 titions shoulct be used at each period, and, for each 
 number of repetitions per period, what interval be- 
 tween periods is best?" In her experiments, she used 
 seven-paired series of nonsense syllables. They were 
 presented visually at the rate of one in three seconds 
 with three seconds interval between. She compared 
 one repetition a day with two, four, and eight repeti- 
 tions a day. She tested learning by a retention test 
 given two weeks after the last presentation of a series, 
 and stated her results in terms of the per cent, correct. 
 
38 THE PSYCHOLOGY OF LEARNING 
 
 The average per cent, correct for all the series given, 
 for one, two, four, and eight repetitions a day, were, 
 respectively, 75.25, 57.75, 42, 13.25. It is seen that one 
 repetition a day was most effective, two repetitions 
 next in value, four next and eight least effective. In 
 Perkins' experiments, each series was presented six- 
 teen times. The learning was therefore spread out over 
 a period varying from two to sixteen days. 
 
 It is evident from Perkins' experiments, and this 
 point has been confirmed by extensive experiments in 
 the author's laboratory, that when a series of nonsense 
 syllables has been presented to a subject once, further 
 presentations at the same sitting have less value than 
 the first presentation. After one presentation, to get 
 the most value from, another, time must elapse. 
 
 The Substitution Experiments. — In the solution of 
 the problem under consideration, three experimenters, 
 Dearborn, Starch, and Pyle, used some form of substi- 
 tution experiment. 
 
 In some class experiments. Dearborn found ten min- 
 utes practice a day better than five minutes twice a 
 day. Starch compared four different methods of using 
 120 minutes: 10 minutes twice a day for six days, 20 
 minutes once a day for six days, 40 minutes every other 
 day for six days, and 120 minutes at one sitting. Five- 
 minute records were kept. The results are shown in 
 Fig. 10. The ten and twenty-minute practices were 
 best. There was little difference between ten and 
 twenty minute practices. The average speed of the 
 ten-minute practices was best, but the final speed of 
 the twenty-minute practices was best. Forty-minute 
 practices are not so good as ten and twenty, and one 
 hundred and twenty minutes at a sitting gave the 
 
ECONOMICAL LEARNING 
 
 39 
 
 poorest results. The experiment is inconclusive, how- 
 ever, because Starch did not measure the learning capa- 
 city of his different groups, and part of the difference 
 shown by the curves of Starch's four groups is doubt- 
 less due to differences in the learning capacity of the 
 different groups. For the first tv/o five-minute records, 
 the ten-minute and forty-minute groups are better, in- 
 dicating better learning capacity. The twenty-minute 
 
 FiorRE 10, From Starch, showing the effects of practice period of 
 different lengths. 
 
 group makes a lower record, showing smaller learning 
 capacity. In spite of this fact, the twenty-minute group 
 makes the best record for the last five or six periods. 
 The forty-minute group starts high. Forty minutes is 
 evidently too long a peiiod for best results. The fact 
 that the 120-minute group is lower for the first two 
 five-minute periods shows the group to be the poorest 
 leamei^, but the very poor scores made in the latter 
 half of the experiment are doubtless chiefly due to 
 
40 THE PSYCHOLOGY OF LEARNING 
 
 method. The legitimate inferences from Starch's ex- 
 periments are : a twenty-minute practice period is best, 
 a ten-minute period is nearly as good, forty-minute 
 period not so good, and a hundred-and-twenty-minute 
 period is poorest of all. Starch did not use a thirty- 
 minute period, which, in the author's experiments, has 
 proven better than longer or shorter periods. 
 
 In the author's attempt to solve this problem, a group 
 of students was required first to spend sixteen days in 
 a check experiment, in which all used the same length 
 of practice period and the same distribution of prac- 
 tices. The subjects then used different lengths of prac- 
 tice periods, and their learning curves were compared 
 with those made when all used the same method of 
 procedure. Fifteen, thirty, forty-five, and sixty minute 
 periods were compared. The results are shown in tables 
 1 and 2. 
 
 TABLE 1. 
 
 DIFFERENT LENGTHS OF LEARNING PERIODS COM- 
 PARED ON THE BASIS OF AVERAGjE SPEED. 
 
 Subject. Method. Improvement. 
 
 A 15 minutes 4.8 per cent, less than in check experiment. 
 
 B 30 minutes 30.3 per cent, better than in check experiment. 
 
 C 45 minutes 20.6 per cent, better than in check experiment. 
 
 D 60 minutes 22.9 per cent, better than in check experiment. 
 
 TABLE 2. 
 
 DIFFERENT LENGTHS OF LEARNING PERIODS COM- 
 PARED ON THE BASIS OF FINAL SPEED ATTAINED. 
 
 Subject. Method. Improvement. 
 
 A 15 minutes 12.7 per cent, less than in check experiment. 
 B 30 minutes 6.1 per cent, less than in check experiment. 
 
 C 45 minutes 3.3 per cent, more than in check experiment. 
 D 60 minutes 5.7 per cent, less than in check experiment. 
 
ECONOMICAL LEARNING 41 
 
 TABLE 3. 
 
 COMPARISON OF DIFFERENT LENGTHS OF PERIODS ON 
 
 THE BASIS OF AVERAGE SPEED, 
 
 SAME AMOUNT OF TIME. 
 
 Subject. Method. Improvement. 
 
 A 15 minutes 22.3 per cent, better than in check experiment. 
 B 30 minutes 36.1 per cent, better than in check experiment. 
 C 45 minutes 25 per cent, better than in check experiment. 
 D 30 minutes 14.8 per cent, better than in check experiment. 
 
 TABLE 4. 
 
 COMPARISON ON BASIS OF FINAL SPEED, 
 SAME AMOUNT OF TIME. 
 Subject. Method. Improvement. 
 
 A 15 minutes 4.9 per cent, less than in check experiment. 
 B 30 minutes 18.1 per cent, more than in check experiment. 
 C 45 minutes 5.4 per cent, less than in check experiment. 
 D 60 minutes 45.5 per cent, less than in check experiment. 
 
 In table 1 the comparison is made on the basis of the 
 average speed of the whole experiment. In table 2 the 
 comparison is based on the final speed attained. By 
 the latter method of comparison, the forty-five-minute 
 practice proves a little better than the thirty-minute 
 practice, and an hour gives almost the same return as a 
 half hour. In both tables 1 and 2, the subjects are com- 
 pared on the basis of the same number of practices, but 
 the extreme difference in the total amount of time put 
 in is as one to four. It is clear that very little return 
 comes after the first thirty minu-tes of practice. The 
 fatigue of the latter part of an hour's practice makes 
 the record show up poorer than that of a forty-five 
 minute period, and but a trifle better than that of 
 thirty minutes. 
 
 In tables 3 and 4 we compare the four different meth- 
 ods on the basis of the same actual amount of tim^e, and 
 of course, different numbers of practice periods. A 
 
42 THE PSYCHOLOGY OF LEARNING 
 
 study of these tables shows that thirty minutes is the 
 best length of practice period. The second fifteen min- 
 utes of practice continues to give a good return, but 
 the third fifteen minutes shows a decreasing return. 
 Fifteen minutes is too short a period in the kind of 
 work done in the experiment to give the best return, 
 and forty-five minutes makes too long a period. The 
 hour period shows up poorly partly because the fatigue 
 of the latter part of the hour cuts down the score and 
 obscures the effect of the work in the early part of the 
 hour. Of course it is possible that the poor work dur- 
 ing the latter part of the hour actually decreases the 
 effect of the work of the first part of the hour. 
 
 Experiments in Archery. — Murphy's experiments in 
 javelin throvdng were chiefly concerned with the in- 
 terval between practices. He did, however, compare 
 ten throws once a day with five throws twice a day and 
 found ten throws once a day better. Muiphy does not 
 compare ten throws at one practice with more than ten 
 throws at one practice, or fewer than ten throw^s. 
 
 Experiments in Adding. — Kirby, Hahn, and Thorn- 
 dike have compared various lengths of peiiods in learn- 
 ing to add. Kirby and Hahn w^orked with children, 
 Thorndike v/ith university students. Kirby and Hahn 
 compared very short working periods. 
 
 Kirby gave a fifteen-minute initial and a fifteen-min- 
 ute final test and forty-five minutes intervening prac- 
 tice. These forty-five minutes were divided up into 
 practices of 221/2, 15, 6, and 2 minutes. The median 
 gain per cent, for these groups in order were 45, 43, 
 42, and 56. As between the three longer groups there 
 is very little difference, with the advantage in favor 
 of the longer period. The two-minute period gives a 
 
ECONOMICAL LEARNING 43 
 
 better return than either of the longer periods. The 
 practice with the 22V2-J^inute group covered a period of 
 only two days, while practice with the two-minute 
 group covered a period ten times as long, and in the 
 latter case, the regular school w^ork and the work out- 
 side of school could have a much greater effect on the 
 score than in the longer periods. 
 
 Hahn does not find greater value in the short periods. 
 Nor does Thorndike. His experiments consisted in 
 practical adding by means of a table. There vv^ere 80 
 operations on a sheet. Eight sheets at a sitting proved 
 more valuable than two sheets at a sitting. He reports 
 that he finds ''little or no advantage in very short 
 periods of learning." 
 
 Distribution of Practice. — After we have practiced at 
 one time as long as practice is profitable, how soon can 
 we practice again v/ith profit? In general, the experi- 
 ments have shown that the best prcatice interval is one 
 day, but twice-a-day practice and alternate-day prac- 
 tice is nearly as profitable as daily practice. However, 
 much depends upon the nature of the learning and the 
 stage of habituation. We shall very briefly give the 
 results of the most important experiments. 
 
 Murphy compared the results of throwing ten javelins 
 at a time daily with throwing five at a time twice 
 daily, and found the once daily method best. As for 
 the other forms of distributing the practices he reports 
 that "learning periods can be distributed by giving 
 alternate days practice, and even weekly practice with- 
 out any loss in learning." 
 
 In a typewriting experiment covering a period of 
 ninety half-hour practices, the author compared two 
 half-hours a day with ten-half-hours a day. In the 
 
 fl:- 
 
44 THE PSYCHOLOGY OF LEARNING 
 
 former case, the half-hours were one in the forenoon 
 and the other in the afternoon. In the latter case, the 
 ten practices had half-hour intervals of rest between. 
 The results of the experiment may be brieflly stated as 
 follows : Concentrated effort is efficient for about five 
 practices, but the speed improved very little on the 
 first day after the fifth practice. The effect of the 
 early practices is evidently much greater if soon re- 
 peated. From the fifth practice to the fortieth, the 
 group practicing tv/ice a day steadily gained on the 
 other group. From the fortieth to the sixtieth prac- 
 tice, the difference between the groups dimin- 
 ished. To summarise: in such work as type- 
 writing, distributed practice is best, but con- 
 centrated practice brings in good returns, and 
 if one is in a hurry to acquire skill, the diminished 
 returns from concentrated practice need not be con- 
 sidered prohibitive. In the above comparisons we have 
 had in mind the same number of hours of practice. 
 The relative value of one-hour-a-day practice as com- 
 pared to five hours a day, is shown graphically in figure 
 11. If we consider merely the number of days spent in 
 practice and disregard the total number of hours actu- 
 ally spent in practice, we can say that ten half-hours 
 a day is better than two half -hours a day. The group 
 practicing ten times a day has a much faster speed 
 at the end of nine days than does the group practicing 
 only twice a day. But if we consider the amount of 
 time actually spent in practice, we get more return for 
 each hour of practice if the practices are spread out 
 over 45 days. The middle line in the figure represents 
 the results of ten half -hour practices a day for nine 
 days. The lower curve shows the results of nine days 
 
ECONOMICAL LEARNING 
 
 45 
 
 of practice, two practices a day. The upper curve 
 shows the results of the same amount of practice as 
 shown by the middle curve, but spread out over 45 
 days instead of nine days. Five hours a day with in- 
 tervals between the practice periods are more effective 
 in fixing the typewriting habits than one hour a day. 
 Just where decreasing returns begin, in repeating prac- 
 tices on the same day, can not be determined from the 
 experiments. In the early part of the experiment it 
 is at the end of the fifth practice. Later in the experi- 
 ment it is probably earlier, perhaps at the end of two 
 or three practices. 
 
 Figure 1.1. B shows the effects of practicing ten half-hours a day 
 for nine days. C shows the effects of practicing two half-hours a 
 day for nine days. A shows the effects of the same amount of 
 practice as B, but at the rate of two half-hours a day for 45 days. 
 In A and B the graphs are constructed from the averages of suc- 
 cessive practices in groups of 10. In B the 10 practices were done 
 in one day ; in A, they vrei:<? done in five days. 
 
46 THE PSYCHOLOGY OF LEARNING 
 
 In the author's substitution experiments, three differ- 
 ent forms of distribution were compared, twice-a-day 
 daily, and alternate days. If our basis of comparisor 
 is the total number of hours of practice, the most eco- 
 nomical procedure is found to be daily practice. A sec- 
 ond practice on the afternoon of the same day give? 
 nearly as much return as the morning practice, and ir 
 the later stages of habituation, alternate days practice 
 is probably as effective as daily. As between daily anc 
 twice-a-day practice, the latter is much more effective 
 if we consider the time elapsing since practice began 
 But, of course, there are twice as many actual hours oi 
 practice. 
 
 In the case of concentrated practice, the question maj 
 well be raised whether the later practices on the same 
 day, even though no higher score is made, are not 
 effective in fixing the habit. This point the authoi 
 tested in the following way. A group of four subjects 
 began in the morning to work all day at a substitutior 
 experiment. At the end of the fourth practice period 
 one of the subjects stopped, the others continuing all 
 day. Two and three days later the effects of the prac- 
 tice were determined. The subject who had worked 
 only four periods showed up as w^ell in the retention 
 tests as those who had worked all day. It is clear that 
 beyond a certain point further practice on the same 
 day is profitless, not only in raising the score but in 
 fixing the habit. 
 
 In Fig. 12 are shown the results of an all-day substi- 
 tution experiment performed by the author. The ex^ 
 periment was as follows: Using a symbol alphabet, 1 
 practiced transcribing for fourteen half-hour periods 
 with half-hour rests between, continuously throughout 
 
 
ECONOMICAL LEARNING 47 
 
 ^le day from eight in the morning till ten at night. The 
 suits are compared with fourteen daily practices of 
 m subject of the same learning capacity as the author. 
 a the concentrated practice in this experiment there 
 J no improvement after the third practice. In the dis- 
 [ributed practice, improvement continues throughout. 
 
 Figure 12. The lower graph shows the results of 14 half-hour prac- 
 tices on the same day in a letter-symbol substitution experiment. 
 The upper graph shows the results of the same number of practices, 
 one a day for fourteen days. 
 
 We need not report in detail the v/ork of other experi- 
 menters bearing on the distribution of practices. In 
 general the experiments of Murphy, Perkins, Chapman, 
 Strong and others confirm the results as stated above. 
 Three facts seem to be fairly certain: (1) In estab- 
 lishing a complicated set of habits as in typewriting, 
 frequently repeated practices at the start are profitable, 
 perhaps as many as three or four half-hour periods a 
 day separated by rests of a half -hour. (2) after habitu- 
 
48 THE PSYCHOLOGY OF LEARNING 
 
 ation has proceeded further, frequent practices seem 
 not to advance habit-fixation faster, or at least not 
 much faster, than less frequent practices. (3) In 
 processes involving complex motor co-ordination as in 
 javelin-throwing and ball-tossing long intervals seem 
 more profitable. Perhaps different laws are involved 
 in situations in which the goal desired is perfecting a 
 series of complicated muscular movements through the 
 trial and error method. In such work as card sorting, 
 the problem is not to get the hand to make a certain 
 kind of movem.ent. The hand can make all the move- 
 ments easily enough. The problem is to remember 
 what movement the hand is to make for each number 
 of card. In such an experiment as ball-tossing, the 
 problem is to secure the proper co-ordinated movement. 
 We know ivhat we are to do, we are not able always 
 to do it. 
 
 As far as it is possible to judge from the experiments, 
 brief practices with relatively long intervals are more 
 profitable in cases involving the securing and master- 
 ing of unperfected motor movements. More concen- 
 trated practice seems profitable in cases where the 
 actual movement can easily be made, and where the 
 problem is the connecting this movement with a par- 
 ticular stimulus. 
 
 What practical advice should be given to teachers 
 who must determine and direct the practice of children ? 
 The practice periods of children should be short, and 
 in most cases, once or twice a day. After habituation 
 has proceeded to a considerable degree of fixation, the 
 practices may be less frequent. By "short," we mean 
 five, ten, fifteen minutes, depending upon the nature 
 of the learning and how fatiguing it is. In general, 
 
ECONOMICAL LEARNING 49 
 
 practice gives reasonable returns until fatigue has set 
 in. Even though shorter periods and longer intei'vals 
 would, in some cases, give better returns, it is not 
 usually expedient to use them, because in most cases 
 we want efficiency at the earliest possible moment, and 
 can afford to accept a decreasing return for longer prac- 
 tices and more frequent intervals. Even if a second 
 practice on the same day with the typewriter does not 
 give me quite so much return as if I should wait till 
 the next day for the next practice, I should take the 
 second period on the same day if I have the time and 
 am in a hurry to acquire speed on the machine. On 
 the other hand, if I am in no hurry for the speed, and 
 I can use my time profitably othenvise, it would be 
 wise to delay practice till the following day. But if I 
 am in a hurry to acquire the speed, and have nothing 
 else to do with my time, I should practice for about 
 five half-hours a day, at least for a while. In deter- 
 mining the length and number of practice periods, a 
 teacher should consider all the factors. In school, usu- 
 ally there is no special hurry, and there are many 
 things demanding attention, therefore economy of time 
 should have full consideration. This means that one 
 or two vigorous practices a day, daily, is best. The 
 child may, in most cases, continue practice at one time 
 as long as he shows a high degree of efficiency, or in 
 other words, till fatigue effects begin to show them- 
 selves. It is certain that in such processes as the ele- 
 mentary operations of arithmetic, practice periods of a 
 very few minutes daily give very high returns. With 
 young children no kind of practice period should be 
 long. For children decreasing returns set in very 
 early. 
 
50 THE PSYCHOLOGY OF LEARNING 
 
 EXPERIMENTS 
 
 The comparison of different lengths of intervals is a 
 work too difficult and extended for a class experiment, 
 but a comparison of different lengths of practice periods 
 may well be undertaken as a demonstration experiment. 
 Two experiments will be described : 
 
 1. Card-sorting'. Material, card-sorting box and cards 
 for one row of boxes for the check experiment and 
 cards for a different row of boxes for the second experi- 
 ment. 
 
 Method : Have all the members of the class sort cards 
 into one row of boxes for ten times at one sitting. On 
 the basis of the records, divide the class into four groups 
 having equal learning ability, then take a different row 
 of boxes for the second experiment. Have one group 
 sort twice a day for ten days, another group sort five 
 times a day for four days, another group sort ten 
 times a day for two days, and the other group sort 
 twenty times at one sitting. With the same scale and 
 on the same base, plot learning cui'ves for each of the 
 four groups. 
 
 2. Substitution experiment. Material, the same as in 
 experiment 1, Chapter H, and an additional key. Method: 
 On the basis of experiment 1 in Chapter H, divide the 
 class into four groups of approximately equal learning 
 ability. Use the same digit-symbol test sheets, but 
 prepare a different key by taking different letters of 
 the alphabet. This key can be supplied to the students, 
 or they can copy it from the blackboard. Work in five- 
 minute periods. One group can work ten minutes a 
 day for six days; another, twenty minutes a day for 
 three days ; another, thirty minutes a day for two days ; 
 
ECONOMICAL LEARNING 51 
 
 and the other group, sixty minutes at one sitting. In 
 all cases, keep the individual records for each five-min- 
 ute period. Plot the learning curves for the different 
 groups as in experiment 1 above. 
 
 3. One or the other of the above experiments is per- 
 haps all that can be undertaken in an ordinary lecture 
 course, but if time and opportunity permit, an experi- 
 ment showing the poor results of extended practice on 
 the same day would be very instructive and impressive 
 to the students. Any sort of learning experiment would 
 suffice. An experiment with the same material as in 
 1 or 2 above is recommended. To illustrate the pro- 
 cedure, we shall take the card-sorting experiment. Use 
 three rows of boxes that have not been used in an 
 experiment before. Make five sortings at a sitting, 
 with half-hour rests between sittings. The class can 
 be divided into four groups of equal ability as deter- 
 mined by previous experiment. One group can sort all 
 day; another group, three-fourths of a day; another, 
 one-half of a day; and the other group, one-fourth of 
 a day. The division of work might very well be two 
 sittings, four sittings, six sittings, and eight sittings. 
 The results of the experiment should be tested a day or 
 two later, by having all the subjects sort the cards 
 tv/ice, and taking the average of the two sortings as 
 an indication of the efficiencj^ attained. 
 
 REFERENCES. 
 
 Bradford, C. J., An experiment in typewriting, Ped. Sem., 
 xxii, 1915, 445. 
 
 Cummins, R. A., Improvement and the distribution of prac- 
 tice, T. C. Cont. to Ed., No. 97, 1919. 
 
 Leuba, J. H., and Hyde, W., An experiment in learning to make 
 hand movemc7its, P. R. xii, 1905, 351. 
 
 Murphy, H. H., Distribution of practice periods in learning. 
 J. E. P. vii, 1916, 150. 
 
52 THE PSYCHOLOGY OF LEARNING 
 
 Perkins, N. L., The value of distributed repetitions in rote 
 learning, B. J. P., vii, 1914, 253. 
 
 Pyle, W. H., Economical learning, J. E. P., iv., 1913, 148; Con- 
 centrated vs. distributed practice, J. E. P., v., 1914, 247. 
 
 Starch, D., Periods of work in learning, J, E. P., iii, 1912, 209. 
 
 Strong, E. K., Two factors which influence economical learn- 
 ing, J. of Phil., etc., No. 5, 1914. 
 
 Thorndike, E. L., Notes on practice, improvability and the 
 curve of ivork, A. J. P., xxvii, 1916, 550. 
 
 Washburn, M. F., and others. The effect of the interval beticeen 
 repetitions in the speed of learning a series of movements, 
 A. J. P., xxiv, 1913, 580. 
 
Chapter IV. 
 
 Economical Learning. 
 
 general factors. 
 
 Attention and Learning*. — In the preceding chapter 
 we have discussed repetitions as if their vaUie were a 
 constant thing. In all the experiments on which our 
 discussions and conclusions have been based, we have 
 assumed that the work was done under favorable psy- 
 chological and physiological conditions. But the fact 
 is that repetitions are not of equal value, and it is 
 now our problem to enquire into the general factors 
 that affect their value. The first and most important 
 of these factors is attention. It will be necessary for 
 us to explain the precise sense in which the term atten- 
 tion is here used. Defined in tern>s of consciousness, 
 attention means sensory cleamesSji To say that we 
 are attending to a process is to saj/ that the process is 
 clear in consciousness. It stands out from the other 
 simultaneous processes; it is focal. Physiologically, 
 attentive processes are those that usurp or monopolise 
 the cerebral activity at the time. As a result of evolu- 
 tion, our central nervous systems are so constituted 
 that there is always a synthesis or unification of the 
 various stimulations of the moment for the initiation of 
 action. This synthesis or unification, physiologically, 
 is attention. It is as if we had not one brain but many ; 
 but to control action, one of these many brains must 
 
 [53] 
 
54 THE PSYCHOLOGY OF LEARNING 
 
 momentarily usurp the function of initiating movement. 
 We have many sense organs. Their constant stimula- 
 tions are in eternal conflict for the control of action. 
 Now, in learning, the processes which we wish to couple 
 together must be among those processes that for the 
 time are monopolising the cortical energy or, in other 
 words, the processes involved in learning must be in 
 the focus of attention. 
 
 It is not so much a matter of definite experiment as 
 a matter of general observation in learning experiments 
 and in life in general that enables us to say that if we 
 wish to learn, we must attend. The processes which 
 we wish to unite through learning must be processes 
 that for the time are monopolising the available cere- 
 bral energy. The primary, fundamental law of asso- 
 ciation is that ideas to be bound together must be 
 experienced together in a state of attention. In the 
 sphere of habit-formation, the law seems to have at 
 least wide application, if it is not indeed universal in 
 its applicability. 
 
 In all the learning experiments performed in the 
 author's laboratory it has been observed that those 
 learners who carried on the practice with the highest 
 degree of concentration, other factors being equal, 
 made the fastest progress. Now, by a high degree of 
 attention we mean no more than that the processes in- 
 volved in the learning monopolised the person's activity 
 for the time. The fast learner is tense, the whole body 
 seems devoted to the learning processes, no other pro- 
 cess can participate to any considerable degree in the 
 organism's activities while these particular learning 
 processes are in progress. In attentive learning the 
 doors seem to be shut against all other processes. 
 
ECONOMICAL LEARNING — GENERAL FACTORS 55 
 
 In inattentive learning, which is also ineffectual 
 learning, the leaniing processes have successful com- 
 petitors; they do not usurp the organism's central 
 neural activities. In such case, the practice or repeti- 
 tions are largely ineffectual in strengthening the neural 
 bonds between stimuli and responses. 
 
 In the card-sorting experiments, the fast learners 
 gave themselves over completely to the work. For the 
 time being the world was to them a card-sorting world ; 
 nothing else existed; nothing else was for the time of 
 any consequence. The card-sorting completely occu- 
 pied the central neural activities. Their bodies were 
 rigid and tense, they leaned forward to their tasks, 
 they whispered to themselves the numbers of the cards. 
 
 With the slow or poor learners, all was different. 
 Their bodies were relaxed, and many other activities 
 shared the central field of neural activity with the card- 
 sorting performance. They often gazed about the 
 . room, listened to the various noises, and watched their 
 associates in the laboratory. 
 
 When the fast learner found a box, he deposited the 
 card and then djvelt upon the location of the box. He 
 allowed the idea of the location of the box to come into 
 full and focal consciousness, and have a noticeable tem- 
 poral existence. He might use some device to aid. He 
 might, for example, say ''Now I must remember num- 
 ber 14, it is there near the corner just under 19". The 
 idea of location would thereby have more than a fleet- 
 ing existence, and would also likely acquire some 
 helpful associations. 
 
 When the poor learner finds a box the card is dropped 
 into it and before there is time for the idea of location 
 to develop and helpful associative ideas to form, the 
 
56 THE PSYCHOLOGY OF LEARNING 
 
 subject passes on to the next card, and the whole ex- 
 IDerience with the card just deposited becomes almost 
 as if it had not been. And when he, a little later, comes 
 to a card of the same number, he must hunt for the box 
 again. 
 
 In the early stages of learning to sort cards, sensa- 
 tional and ideational disturbances are fatal. The sen- 
 sational, perceptual disturbers are such as others sort- 
 ing cards in the same room, the noise of their perform- 
 ance, their whispering to themselves, the noise of an 
 opening door. Any sensational disturbance tends to 
 prevent the formation of the bonds. Any sensational 
 competitor for focal consciousness, driving out the 
 box-location idea, obliterates the bond between stimu- 
 lus and response. 
 
 Ideational disturbances are equally effective in 
 obliterating or obstructing the bonds which we are 
 trying to form and strengthen. With proper care and 
 precaution we can guard against most of the sensa- i 
 tional disturbances but we are powerless to prevent 
 ideational ones. Any ideas foreign to the learning 
 process will disturb. For example, if a subject is 
 doing well, and the idea comes, "now I am doing fine,'* 
 this idea interferes, he forgets the box-locations and 
 the score is lowered. 
 
 Sensational disturbers or distractions can lose their 
 force. We can become adapted to them. As a rule, 
 constant environmental factors become as if .they did 
 not exist. It is therefore necessary in a comparative 
 learning experiment to keep all external conditions con- 
 stant. We must have the same persons sit in the same 
 places, and all persons and things in the room must 
 maintain the same relative positions in successive ex- 
 
ECONOMICAL LEARNING—GENERAL FACTORS 57 
 
 periments. If a person is used to working with a 
 group, then not to have the group present is a distrac- 
 tion. On the other hand, if a subject is used to working 
 alone, the group will disturb. Great importance often 
 attaches to apparently insignificant factors. We can 
 not ignore the details of time and place, the method of 
 starting the experiment, the arrangement of the 
 apparatus, the place in the room or about the table 
 where the person works. Even the general attitude of 
 the experimenter and his tone of voice are important 
 factors and must be kept constant. 
 
 In Geissler's important attention experiments it was 
 found that the best adding records were made under 
 conditions of distraction. This seems paradoxical, but 
 is really in haiTnony with the principles laid down 
 above. The subjects in Geissler's experiments had dis- 
 traction series and no-distraction series. In a no- 
 distraction series, the subject said, *'0h, this is easy, I 
 can take my time". He would consequently relax and 
 leisurely perform the adding, making only a moderately 
 fast record. But if it was a distraction series, the sub- 
 ject took a different attitude. He said, "Now, this is 
 difficult ; this will take all there is of me." He became 
 rigid, tense, used every known device to give the adding 
 processes the right-of-way and keep the distracting 
 stimuli out. As a rule, he was successful. The dis- 
 tracters did not really distract; they were not able to 
 become focal ; they beat in vain at the gates of central 
 neural activity. 
 
 Attitude and Learning. — The relation of attention to 
 learning is fundamental. Several other factors that 
 affect the rate of learning, probably produce this effect, 
 in most cases at least, through their effect on attention. 
 
58 THE PSYCHOLOGY OF LEARNING 
 
 One such factor is attitude. The subject's attitude 
 toward the work of a learning- experiment is one of the 
 most important factors determining progress. If for 
 any reason the subject disHkes the work, progress is 
 slow; little effort is put forth; the subject works in a 
 poor state of attention to the work. There is no ambi- 
 tion to do well. The subject is content to make poor 
 records, for in his opinion the work is of no consequence, 
 and it therefore is no disgrace to do poorly in it. It 
 often happens that poor records at the beginning are 
 the cause of the dislike. The dislike in turn becomes 
 the cause of more poor records. Poor work and bad 
 attitude reinforce and perpetuate each other. 
 
 A bad attitude retards learning, but just as surely a 
 favorable attitude accelerates learning. The subject 
 who likes the work seems able to give himself over 
 more fully and completely to the work than one who 
 dislikes it. Good records at the beginning of an experi- 
 ment often are the cause of a favorable attitude. The 
 person does well in the work and therefore likes the 
 work. The favorable attitude caused by good records 
 becomes itself the cause of more good records. It 
 therefore turns out that the good become better and the 
 poor become worse. In learning we have an example 
 of the law that "To him that hath, shall be given and 
 from him that hath not shall be taken even that which 
 he hath". 
 
 Peterson reports a class experiment which shows 
 the effect of attitude. A list of words was put on the 
 blackboard and then copied by the students of the class. 
 An immediate reproduction was called for and also a 
 ^■elayed reproduction. The experiment was repeated 
 in the same way except that the students were told 
 
ECONOMICAL LEARNING — GENERAL FACTORS 59 
 
 that a reproduction would be called for. In the ex- 
 periments in which the students knew that a re- 
 production would be called for, the immediate 
 reproduction was 14.8% better for one section of 
 students and 30% better for the other section. The 
 delayed reproduction was 48.4% better for one section 
 and 51% better for the other. The difference in atti- 
 tude made a very great difference in the results. When 
 the students knew that a reproduction would be called 
 for, they paid more attention to the words, kept them 
 in the focus of attention longer, said them over to 
 themselves. When one knows what is to be expected 
 of him, he directs and controls his attention and obser- 
 vation accordingly. 
 
 A somewhat similar result is shown by the experi- 
 ments of Ordahl. Her experiments showed that if ob- 
 jects appeared in the field of vision but were not at- 
 tended to, they were no more easily learned later than 
 if they had never been seen. It is evident that in idea- 
 tional learning, objects must be in the focus of atten- 
 tion. Attitude has much to do in determining whether 
 a process shall be focal or not. We usually see what 
 we look for. In learning experiments I have often read 
 the same passage over and over again to different sub- 
 jects. The subjects, one after another, would learn the 
 passage, while I would not, although I had read it to 
 many subjects in succession, each of whom had learned 
 it. The reason I had not learned it was because of atti- 
 tude. I had not tried to learn it. 
 
 The importance of the attitude of school children is 
 certainly very great. If they are to learn effectively, 
 they should be favorably disposed toward the teacher, 
 toward the school, toward the various school subjects 
 
60 THE PSYCHOLOGY OF LEARNING 
 
 and toward the specific task or lesson. If the child likes 
 the school, the teacher, and the particular work, fast 
 progress is assured if other factors are favorable. Dis- 
 like of the school, the teacher or the task works against 
 progress. When a favorable attitude on the part of 
 the pupil is secured toward the various aspects of school 
 life, successful learning is assured, as far as the pupil 
 has ability to learn. On the other hand a pupil that 
 has a dislike to the school, the teacher, or the particu- 
 lar work, proceeds under a very great handicap. A 
 year of school is wasted for many a child because of 
 such dislike. One of the teacher's greatest problems 
 and duties is to secure a favorable attitude on the part 
 of those to be taught. 
 
 Life Bents or Dispositions. — Attitudes may be tem- 
 porary or permanent. In the latter case they may be 
 called life bents or dispositions. We can become favor- 
 ably disposed toward a certain study as history, a cer- 
 tain principle as evolution or democracy, a certain 
 method as that of analysis. These favorable disposi- 
 tions foster their own growth and perpetuate them- 
 selves, becoming permanent dispositions. They make 
 learning easy in certain narrow directions, and difficult 
 in others. 
 
 Definite attitudes are often formed early in life. 
 These attitudes then may affect all future learning. A 
 child because of having a poor teacher, because of get- 
 ting started in a wrong way, because of starting too 
 early, or for some other reason, may have a dislike for 
 arithmetic. Because of this dislike, he studies it as 
 little as possible, putting his time on studies that are 
 liked. He therefore makes little progress, and this 
 rr? ^kes mastery ever more difficult and increases the 
 
ECONOMICAL LEARNING — GENERAL FACTORS 61 
 
 dislike. Progress in arithmetic continues to be slow 
 because of the unfavorable attitude and because 
 of the poor work done in it before. In a similar 
 way a pupil may acquire an unfavorable attitude toward 
 any of the school studies. These facts make clear the 
 importance of beginnings, the first day in school, the 
 first experiment with the teacher, the first lesson in any 
 subject. School work is hard at best and should not 
 be made harder by means of unfavorable experiences. 
 
 The learning of lessons or any other assigned tasks 
 should never be used as forms of punishment. Chil- 
 dren should never be kept after school as a form of 
 punishment. The reason that such practices are wrong 
 is because such procedures create dislike for things 
 that should be liked. They create unfavorable dispo- 
 sitions and attitudes that make learning difficult. 
 Failure to take account of these simple principles is 
 the cause of many a failure in the school-room, and the 
 cause of many a wasted life. 
 
 Incentives to Practice. — Attention is the fundamental 
 condition necessary to learning, and attitude may be 
 considered a permanent state of attention, or readiness 
 to attend to a certain type of process. The practical 
 problem is how to secure attention and permanent dis- 
 positions. There are various factors or elements which 
 we may discuss under the head of incentives. 
 
 (1) Knowledge of the End Sought. — If the learner 
 can be made to see the end sought by practice, learning 
 is usually facilitated. This facilitation comes through a 
 better attention and more favorable disposition. Few 
 people are willing to work blindly. Exactly what is to 
 be gained by a habit and exactly what the habit is, 
 should be made clear to the learner. In addition, for 
 
62 . THE PSYCHOLOGY OF LEARNING 
 
 example, the immediate end sought is to be able to pro- 
 nounce the sum immediately upon seeing the numbers. 
 The skill when attained enables us to add the scores in 
 a game, our money, our account, etc. 
 
 (2) Knowledge of the Score. — Knowledge of one's 
 record is a material factor in progress. If a learner 
 keeps a close and accurate record of his progress, it 
 incites him to practice at the highest efficiency. With- 
 out a knowledge of results, one usually practices on a 
 level much below the maximum. Teachers should find 
 accurate means of measuring the progress of the pupils. 
 The improvement from day to day and from month 
 to month should be shown to the children. They should 
 be taught to make and interpret learning curves, and 
 should plot their own curves. 
 
 Arps furnishes some experimental evidence of the 
 effects of a knowledge of the results of practice. Using 
 the Bergstrom ergograph, he took series of records in 
 which the subjects were shown their records. He took 
 other records and gave the subjects no knowledge of 
 their records. In the case of work with knowledge, 
 the results were 18 per cent, better than in work with- 
 out knowledge. 
 
 Wright, also working with the ergograph, used as 
 an incentive the ambition to reach a certain standard 
 of work. The students were shown their records. More 
 work was accomplished with the incentive than without 
 it. If an impossible limit were set, the total amount 
 of work was decreased. Fatigue was found to be less 
 when the subject worked with the incentive. 
 
 In ordinary school work, the pupils work on blindly, 
 not knowing — or caring — what they are aiming at, nor 
 what it will be when attained ; they do not know what 
 
ECONOMICAL LEARNING — GENERAL FACTORS 63 
 
 progress they are making; they do not know what 
 progress is possible nor what attainment is finally pos- 
 sible. When they go on the playground to play, the 
 situation is entirely changed. There is something defi- 
 nite to be attained. They engage in running or jump- 
 ing matches, let us say. They know how fast they 
 can run and how far they can jump. As a rule they 
 know it very accurately. They know who among them 
 is best and how much he is best. They know the estab- 
 lished records in the various sports. They know how 
 much they themselves have improved since the year 
 before. The performances are definite, the methods are 
 definite, the results are definite and definitely known. 
 Inside the school-room, everything is vague and indefi- 
 nite. Everything is imposed from without. There is 
 little inward impulsion or desire as there is on the 
 playground. Definite knowledge of progress made by 
 the pupils is one means of improving the work inside 
 the school-room. 
 
 (3) Knowledge of Errors, — Not only should children 
 and all learners know the results attained, they should 
 also have definite knowledge concerning their mistakes. 
 In that type of learning in which errors are possible, 
 unless one knows his errors he may practice on indefi- 
 nitely with little or no improvement. Experiments have 
 shown that in such tasks as mirror writing and pris- 
 matic writing, knowledge of the erroneous movements 
 are necessary to progress. In spoken and written lan- 
 guage, as Thorndike has pointed out, one may go on 
 practicing indefinitely without improvement. 
 
 (4) Knowledge of When the Material Learned Is to 
 Be Reproduced.— li one knows when learning, that the 
 matter learned is to be reproduced, immediately or 
 
64 THE PSYCHOLOGY OF LEARNING 
 
 later, he proceeds differently in his learning. He takes 
 a different attitude, and the learning is more efficient. 
 
 These four points may be summarised as follows : 
 Learning can be improved by giving the learner a defi- 
 nite idea of the nature of the habit or the end sought 
 by learning and the advantage that is to come to him 
 from possessing the knowledge or habit, by a definite 
 knowledge of the progress made and of the mistakes, 
 and by a knowledge of the use — ^immediate or delayed — 
 that is to be made of the results of learning. 
 
 Definiteness of the Bond. — In some kinds of learning 
 it is difficult to use the principles set forth above be- 
 cause of the indefiniteness of the nature of the bond to 
 be formed. Progress is always easier when the nature 
 of this bond can be definitely and clearly known. In 
 mathematics, for example, the nature of the bond is 
 always clear. In all the fundamental operations of 
 arithmetic, the bonds can be definitely known : 9-f 8=17 ; 
 6X7=»42; V81—9; (13)2=169. All this is definite and 
 practice can be to the point. The end sought is known, 
 it is definite; mistakes can be accurately pointed out, 
 there is never any doubt about the matter; progress 
 can be accurately and definitely measured. But it is 
 not so in writing, drawing, or English composition. In 
 writing, it is true, a copy can be set, and the child can 
 know what sort of result is expected, but just what he 
 is to do to achieve that result, he does not know and 
 no one can tell him. All he can do is to keep on trying. 
 If his product is not like the copy, it is often difficult 
 to say what the difference is. Sometimes, we can get 
 at it approximately. We can say a letter is too high, or 
 too low, or too nan'ow, or too wide. In drawing we 
 have precisely the same difficulties, so also in learn- 
 
ECONOMICAL LEARNING — GENERAL FACTORS 65 
 
 ing to write good English. When a pupil 
 writes an English composition, it is difficult 
 for the teacher to point out exactly what its 
 defects are. But progress by the pupil is pos- 
 sible only to the extent that the defects can be pointed 
 out. To illustrate: A particular sentence may not be 
 clear, it may be ambiguous. Such defects can be pointed 
 out, and just why the sentence is not clear must be 
 shown, just wherein it is ambiguous must be pointed 
 out. 
 
 In all subjects in which the nature of the bonds to 
 be formed is more or less indefinite, teaching is likely 
 to be loose and vague, and the results uncertain. When 
 we can not know exactly what we are seeking, it is im- 
 possible to measure our approach to the thing sought. 
 When an English theme is handed in we cannot indi- 
 cate either its excellencies or its deficiencies as we can 
 when the solution to a problem in mathematics is 
 handed in. In English, therefore, and in all subjects 
 in which the same situation exists, teaching can never 
 have the definiteness that it does in science and mathe- 
 matics. English teaching is vague still in another sense. 
 Different teachers have different ideas as to what is to 
 be attained and put different estimates on the value of 
 a literary product. There can be little difference of 
 opinion as to whether a problem is solved correctly. 
 There can be much difference of opinion as to the merit 
 of a composition. 
 
 Franklin gave us two good examples of definiteness 
 in practice. One was in the field of ethics and the other 
 in the field of English. His clear insight and farsighted- 
 ness enabled him to see that one must go about im- 
 provement in any field in a definite, systematic and 
 
66 THE PSYCHOLOGY OF LEARNING 
 
 methodic way. He knew that character was dependent 
 upon habits. He knew that he could not merely resolve 
 to be good and immediately make improvement in all 
 aspects of his character. So he made a list of the 
 desirable virtues and practiced on one for a time, then 
 added another, then another, and so on till he had in- 
 cluded all the virtues. Then he went all over the whole 
 scheme again. To improve his English, he memorised 
 certain material in the Spectator. After he had for- 
 gotten the words, he wrote out the thought in his own 
 words. He then compared his expression of the thought 
 with the expression in the Spectator. Here was some- 
 thing definite. On the one hand, he had his own ex- 
 pression of a thought ; on the other, he had the expres- 
 sion of the same thought by one of the best writers of 
 the time. He could compare them. He could see 
 wherein his writing was poor. 
 
 By being more definite and specific there is possi- 
 bility of improvement in our teaching, whatever we may 
 be teaching. We should make careful analysis of the 
 material which we are teaching, and come to a clear 
 understanding of it ourselves and then make it clear 
 to those taught. We must have a definite goal, a defi- 
 nite route to travel to reach it, and a definite means of 
 knowing when we have arrived. Learning must cease 
 to be a travelling by an unknown route to an unknown 
 place. 
 
 Feeling and Learning. — The relation of feeling to 
 learning may be discussed from two points of view. 
 Feeling is important in securing practice, and in mak- 
 ing practice effective through attention. Feeling and 
 attention are different aspects of the same thing. When 
 processes are pleasant, we wish to experience them. To 
 
ECONOMICAL LEARNING — GENERAL FACTORS 67 
 
 say that we like a thing is about the same as saying 
 that we shall attend to it. Attention is necessaiy to 
 effective learning. Feeling is necessary to attention. 
 Most of the importance of attitude previously dis- 
 cussed, depends upon favorable disposition, and favor- 
 able disposition depends upon pleasurable experience. 
 If a child begins to form a habit and for any reason 
 the processes are accompanied or followed by pain, then 
 the child does not want to practice again. Feeling is 
 therefore an important element in learning because 
 without accompanying or resulting pleasure we do not 
 like to practice, and when we do practice, do not throw 
 our whole selves into it. Pleasure accompanies experi- 
 ences of which we wish more, experiences that we seek. 
 Pain is connected with experiences which we wish to 
 avoid. Briefly, pleasure leads to practice and makes ^ 
 practice more profitable. It leads to practice because/ 
 we like to do things which give us joy. It makes prac- 
 tice more profitable because it insures a higher degree 
 of attention. From the point of view of securing 
 effective practice we must take great pains to make 
 conditions such that practice will be pleasurable. 
 
 Thomdike claims that pleasure has still another re- 
 lation to learning, namely, that it stamps in the process. ^ 
 If the passage of a stimulus over to its response is ac- 
 companied or followed by pleasu^^e, this pleasure has an 
 effect upon the nervous path which results in strength- 
 ening the bond, — so the argument runs. The recent 
 work of Kline gives some warrant for the contention. 
 Kline found that the pleasure accompanying movements 
 helped to fix the bonds. 
 
 Snoddy, however, in a recent ^'Experimental analysis 
 of a case of trial and error learning in the human sub- 
 
68 THE PSYCHOLOGY OF LEARNING 
 
 ject," claims that the pleasure resulting from a move- 
 ment has nothing to do with stamping it in, but he 
 gives no conclusive evidence. He points out that in his 
 experiment — the mirror tracing of a star — improve- 
 ment followed a recess or rest period. *Trom such evi- 
 dence," he says, "it is readily seen that no selecting 
 agent, such as 'satisf yingness of a response' is operative 
 to 'select out' the successful responses made in a series 
 of random movements — the basis of Thorndike's view, 
 since the successful responses did not occur in the 
 tracings before recess [rest] period." But Thorndike 
 never claimed that pleasure could stamp the bond in 
 before its formation. After the rest periods, successful 
 movements or tracings were made. Thorndike's view 
 is that the pleasure resulting from the successful 
 tracings is a causal element in fixing these movements 
 for the future. And Snoddy gives no evidence to prove 
 that satisfaction had nothing to do with stamping 
 them in. 
 
 The fact is that the experimental evidence available 
 does not enable us to say whether the physiological 
 correlate of pleasantness is a causal element in fixing 
 a bond once secured. It certainly does not enable us 
 to say that it is not, and the physiological facts at 
 present known — facts concerning the nature of pleas- 
 antless and unpleasantness — rather support the Thorn- 
 dike view that pleasure is causally effective. 
 
 Recent psychological and physiological investigations 
 into the nature of feeling and emotion make it clear 
 that pleasantness and unpleasantness have far-reaching 
 and fundamental effects upon the body which must pro- 
 foundly affect learning, in ways other than merely se- 
 curing attention or stamping in the process. Pleasant- 
 
ECONOMICAL LEARNING — GENERAL FACTORS 69 
 
 ness has important positive relations to general well-.- 
 being, and specifically to digestion and nutrition, also 
 to cardiac activity and to neural activity. Unpleasant- 
 ness produces in general the opposite effects. Pleas- 
 antness heightens and facilitates all positive, health- 
 ful, life-giving functions of the body. Unpleasantness 
 is negative. It retards, inhibits, constricts. Pleasant- 
 ness makes for more of life ; unpleasantness makes for 
 less of life. It is therefore clear that pleasantness is 
 a necessary condition of learning, and that unpleasant- 
 ness retards learning. 
 
 It will be wise for teachers to bear these facts in 
 mind, and use all available means to make the learning 
 processes pleasant. The age-old practice of rewarding 
 the child when it does what we think it ought, and of 
 punishing it when it does what we think it ought not, 
 has sound scientific justification. The Montessori prin- 
 ciple of using great care to prevent unpleasantness from 
 becoming attached to learning, is also sound. So im- 
 portant is the matter of attitude and feeling in learning 
 that we may say that attention, favorable attitude and 
 pleasantness are the absolute essentials of economical 
 learning. 
 
 But while pleasantness facilitates learning, any in- 
 tense emotion of whatever kind is unfavorable to learn- 
 ing. This statement is based upon general observation 
 of the effects of intensive emotions on a learning sub- 
 ject and also upon the results of physiological studies. 
 Cannon says : **Any high degree of excitement in the 
 central nervous system, whether felt as anger, terror, 
 pain, anxiety, joy, grief or deep disgust, is likely to 
 break over the threshold of the sympathetic division 
 and disturb the functions of all the org:ans which that 
 division innervates." 
 
70 THE PSYCHOLOGY OF LEARNING 
 
 Violent emotions prepare the body for intensive mus- 
 cular activity. To quote Cannon again: "Muscular 
 action is made more efficient because of emotional dis- 
 turbances of the viscera. The cessation of processes 
 in the alimentary canal ; the shifting of blood from the 
 abdominal organs; the increased vigor of contraction 
 of the heart ; the quick abolition of the effects of mus- 
 cular fatigue ; the mobilizing of energy-giving sugar in 
 the circulation — every one of these visceral changes is 
 directly serviceable in making the organism more 
 effective in the violent display of energy which fear or 
 rage or pain may involve." The violent emotions pre- 
 pare us for violent muscular activity but not for learn- 
 ing. The conscious processes accompanying the various 
 physiological changes listed by Cannon would always 
 be focal in consciousness to the exclusion of learning 
 processes. 
 
 EXPERIMENTS AND EXERCISES. 
 
 1. Object, to study the effects of distracting the at- 
 tention of a learning subject. Material, digit-symbol 
 test sheets, two different keys. Method: Divide the 
 class into two equal groups. Let one group do a five- 
 minute practice under distraction, then a five-minute 
 practice without distraction. Let the other group do 
 first a non-distraction practice and then a practice 
 with distraction, five minutes each time. The keys 
 should be used in the same order by the two groups. 
 If the groups are small and are not known to have the 
 same learning capacity, they should do a substitution 
 experiment with a different key under the same condi- 
 tions. Find the average performance of the two groups 
 under distraction and without distraction. The in- 
 
ECONOMICAL LEARNING — GENERAL FACTORS 71 
 
 structor can determine the form of distraction to be 
 used. Some form of auditory distraction will prob- 
 ably be found most convenient ; for example, a continu- 
 ously ringing electric bell, or a distraction suddenly 
 introduced. 
 
 2. Object, to study the effects of directing the atten- 
 tion of a learning subject. Material, ten cards with a 
 word printed in the middle of each, with a color at 
 the top and a number at the bottom. The colors should 
 all be different and the numbers different. Method: 
 Say to the subject that you will show him in succession, 
 ten cards with a word printed in the middle of each, 
 and that he is to write down the words after the ten 
 have been exposed. Expose the cards very quickly, 
 giving just time for the words to be clearly seen. 
 After the exposure, ask the subject to reproduce the 
 colors and numbers instead of the words, and compare 
 the results with the reproduction of the colors and 
 numbers when the attention is directed to them. This 
 experiment can be given only to subjects who are un- 
 acquainted with its purpose. 
 
 3. Further experiments on the effects of attitude can 
 be devised by the instructor or students. For example : 
 Select a short poem and read it aloud to a learning sub- 
 ject. Determine the number of repetitions required 
 for the poem to be learned. Then read the poem to 
 other learners till each has learned it. If you have not 
 tried to learn it, you probably can not reproduce it 
 although several persons have learned it while you have 
 been reading it. Repeat the experiment with a dif- 
 ferent poem, and try to learn it while your subjects are 
 learning it. The effects of attitude will be evident. 
 
72 THE PSYCHOLOGY OF LEARNING 
 
 4. The effects of knowledge of errors can be studied 
 by means of an experiment in which some form of 
 concealed hand movement is required. Compare the 
 results when the subject is ignorant of his errors with 
 the results of other subjects who are shown their 
 errors. The movement can be such as placing a pencil 
 at a certain angle with the perpendicular. The details 
 should be planned by the instructor. 
 
 5. The effects of attitude and consciousness on learn- 
 ing can be further studied by means of experiments 
 modeled after those of Ordahl and Peterson. See the 
 references. 
 
 REFERENCES. 
 
 Arps, G. F., A preUminaru report on work with Ixnowledge vs. 
 iDork without knowledge of results, P. R., xxiv, 1917, 449. 
 
 Boswell, F. P., and Foster, W. S., On memorising with the 
 intention permanently to retain. A. J. P., xvii, 1916, 420. 
 
 Chapman, J. C, and Feder, R. B., The effect of external incen- 
 tive on improvement tested in addition, J. E. P., viii, 1917, 469. 
 
 Conrad, H. E., and Arps, G. F., An experimental study of eco- 
 nomical learning, A. J. P., xxvii, 1916, 507. 
 
 Dallenbach, K. M., The effect of practice upon visual appre- 
 hension in school children, J. E. P., v, 1914, 321 and 387; The 
 effect of practice upon visual apprehension in the feeble-minded, 
 J. E. P., X, 1919, 61. 
 
 Ordahl, L. E., Consciousness in relation to learning, A. J. P., 
 xxii, 1911, 158. 
 
 Peterson, J., The effect of attitude on immediate and delayed 
 reproduction, J. E. P., vii, 1916, 523. 
 
 Rich, G. J., Directed attention and learning, J, E. P., viii, 
 1917, 239. 
 
 Smith, F. 0., The effect of training in pitch discrimination, 
 P. R. Mon., No. 69, 67. 
 
 Snoddy, G. S., An experimental analysis of a case of trial and 
 error learning in the human su1)ject, P. Mon. No. 124, 1920. 
 
 Wells, F. L., Practice effects in free association, A. J. P., xxii, 
 1911, 1. 
 
 Whipple, G. M., and Curtis, J. N., Preliminary investigation 
 of skimming in reading, J. E. P., viii, 1917, 333. 
 
 Wright, W. R., Some effects of incentives on work and fatigue, 
 P. R., xiii, 1906, 23. 
 
Chapter V. 
 
 ECONOMIC LEARNING. 
 
 SPECIAL FACTORS. 
 
 In the preceding chapter we considered the funda- 
 mental factors that influence learning. We shall now 
 consider a number of specific factors that have a bear- 
 ing on economical learning. The ultimate explanation 
 of the specific factors is to be found, in most cases, in 
 the more general and fundamental principles already 
 discussed. 
 
 School Room Drill. — Habits can be established only 
 through repetition. Knowledge can be organised and 
 fixed for permanent and practical use only through 
 thinking it over, by bringing the ideas to conscious- 
 ness again and again. The school undertakes to estab- 
 lish a great number of essential habits in the pupils, 
 and to help them acquire and organise a great body of 
 knowledge. There are writing habits to be formed, 
 spelling habits, reading habits, language habits, arith- 
 metic habits and a multitude of social, moral and con- 
 ventional habits. This work takes most of the time of 
 the school. The public school in this country originated 
 as a place where the education of the home was supple- 
 mented by drill in the so called three R's. As the years 
 have gone by, the school has enlarged its function and 
 taken over nearly the whole work of the education of 
 the child. It is not our purpose here to go into the 
 general question of the socialisation of the school, and 
 its other modern developments. We shall consider only 
 one aspect of this development. 
 
 [73] 
 
74 THE PSYCHOLOGY OF LEARNING 
 
 Certain modern educational reformers have under- 
 taken to dispense with formal drill altogether. Their 
 notion is that the child will incidentally get enough 
 repetition to fix the fundamental habits. In such a 
 scheme, spelling, writing, arithmetic and even reading 
 are not taught as such specifically. The child writes 
 letters and essays and in writing them gets practice in 
 writing and spelling. The child is never drilled in the 
 fundamentals of arithmetic, but he solves concrete 
 problems and in doing so gets practice in the use of 
 the fundamentals. 
 
 The problem of formal versus incidental drill is really 
 an experimental one. It can not be settled by argu- 
 ment but only by an appeal to the facts. One of the 
 primary principles of learning is, there must be repeti- 
 tion. For learning to be economical, the repetitions 
 must be under the general conditions of economical 
 learning already discussed. Any procedure that se- 
 cures practice under these principles is psychologically 
 sound. We have therefore to compare the results of 
 repetition that is merely incidental with the results of 
 repetition that is specific or formal. There are several 
 studies that touch upon the problem. 
 
 Drill in Arithmetic. — In 1911 J. C. Brown reported 
 an experiment in which he had undertaken to deter- 
 mine the effect of drill in arithmetic. He worked with 
 children whose average age was thirteen and a half 
 years. He divided them into two groups of equal 
 average ability in arithmetical operations. One group 
 of twenty-five pupils was given five minutes of daily 
 drill in the fundamental operations of arithmetic for 
 thiiiy days. The other group of twenty-six pupils may 
 be called the control group. It did the same work in 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 75 
 
 arithmetic as the drill group except that it had no 
 formal drill in fundamentals. The only practice it had 
 in the fundamental operations of arithmetic was what 
 it got in solving the regular problems. At the end of 
 thirty days the groups were tested and the drill group 
 was found to be 21.2 per cent, better in the number of 
 problems solved, while the non-drill or control group 
 was only 9.8 per cent, better. After a twelve weeks 
 vacation, the groups were again tested and it was found 
 that the drill group maintained its superiority. 
 
 Brown later repeated his experiment using 222 pupils 
 whose average age was 12.2 years. The drill was in 
 arithmetical fundamentals as before and continued five 
 minutes a day for twenty days. The drill group im- 
 proved 16.9 per cent, while the non-drill group improved 
 only 6.4 per cent. This second more extensive study 
 therefore corroborated the results of his first study. 
 
 Thorndike in 1910 and Donovan and Thomdike in 
 1913 reported some experiments which show the value 
 of school room drill in arithmetic. In one experiment, 
 twenty-nine fourth grade children were drilled two 
 minutes twice a day for fifteen days — sixty minutes 
 practice in all. They improved from two and three- 
 foui-ths examples per minute to four and a half a 
 minute, showing the great effects of practice. 
 
 Phillips reports (1913) an experiment with sixth, 
 seventh, and eighth grade children. He divided them 
 into a drill and a control group. The drill group had 
 ten minutes a day practice for two months in arithmeti- 
 cal fundamentals and in reasoning. The drill group 
 gained 15% more in fundamentals and 50% more in 
 reasoning than did the control group. 
 
76 THE PSYCHOLOGY OF LEARNING 
 
 Winch, while studying the problem of the transfer 
 of training, obtained evidence of the great effect of 
 drill in arithmetical fundamentals. He made several 
 extensive studies, in every case dividing the pupils 
 studied into a drill group and a control group. I give 
 the results of two experiments which may be taken as 
 typical. 
 
 (1) Thirty-three pupils, ten years old, were drilled 
 in arithmetical fundamentals thirty minutes a day for 
 ten days. As a result of the drill they showed an 
 improvement of 21% in numerical accuracy. The 
 basis of the comparison was the improvement of the 
 last two practices over the first two. 
 
 (2) Thirty-two boys, ten years and three months 
 old, were given ten daily practices of thirty minutes 
 each, and as a result showed an improvement of 45.5% 
 in numerical accuracy. 
 
 In these experiments, the control group did no 
 mathematical work whatever. While the drill group 
 was practicing on arithmetical fundamentals, the con- 
 trol group studied English in one case and drawing in 
 the other. The drill group did no mathematical work 
 except the drill of the experiment. Winch was inter- 
 ested only in ascertaining whether the drill made the 
 pupils better in arithmetical reasoning. He was not, 
 therefore, able to compare formal drill with incidental 
 drill, since the control group did not have any drill of 
 any kind during the experiment. The experiments do 
 show, however, the very remarkable effects of a few 
 hours of drill. 
 
 The experiments leave no doubt of the great effect 
 of even very short drills daily in the fundamentals of 
 arithmetic. It seems clear that besides the regular 
 
ECONOMICAL LEAKNING — SPECIAL FACTORS 77 
 
 work in arithmetic, that of learning the principles and 
 solving problems, pupils should be given short, vigorous 
 drills on fundamentals and perhaps also in solving 
 easy problems mentally. The returns from such drills 
 are enormous. Not only do drills in fundamentals give 
 ease, facility and confidence in their use but make the 
 solving of concrete problems easier because, since the 
 pupils have mastered the fundamentals through drills, 
 their energies are all available for the solution of the 
 problems. The earlier experiments of Winch gave 
 some experimental evidence for the above statement, 
 but his later experiments did not. However in his 
 experiments, the pupils did not actually solve the 
 problems, they only indicated the solution. If they 
 had actually perfoiTned the operations, the effects of 
 their previous drill in fundamentals would doubtless 
 have been evident. 
 
 Drill in Other Subjects. — Wallin's studies in the 
 Cleveland schools showed the value of spelling drills. 
 He says, *Tt is by no means evident that modern peda- 
 gogy demands the substitution of incidental spelling 
 instruction for the spelling drill. It is stOl less evident 
 that the schools have outgrown the drill in the other 
 formal branches of the curriculum." 
 
 Peters made a study of the influence of speed drills 
 upon the rate and effectiveness of silent reading. He 
 gave speed drills for a period of seven months to grades 
 three, four, five and six. Grade three is omitted from 
 his results. Two hundred and seven pupils took part 
 in the experiment. In each grade, there w^ere two drill 
 groups and one control group. The drill groups took 
 five to ten minutes of the regular reading period for 
 drill in rapid silent reading. The drill groUP at the 
 
78 THE PSYCHOLOGY OF LEARNING 
 
 beginning of the experiment read 83.8% as fast as the 
 non-drill group; at the end of the experiment the drill 
 group read 107.5% as fast as the non-drill group. The 
 pupils that were trained in rapid reading therefore 
 improved 18.7% more than did the control group, and 
 this improvement was not at the expense of quality, 
 which, according to Peters, was not materially affected. 
 The experimenter says his result "strongly suggests 
 the advisability of giving speed drills as a part of the 
 teaching of reading", and says further: "It is also 
 probable that it is worth while to teach children to 
 skim." 
 
 Whipple and Curtis made a specific investigation of 
 skimming in reading. They seemed not to be concerned 
 with improvement in skimming through practice, but 
 rather in a comparison of skimming with other methods 
 of reading. It is impossible to determine from the 
 published report of their study how much their sub- 
 jects improved by practice. The experimenters con- 
 clude, however, that "It seems probable that practice 
 in skimming might profitably be given in the public 
 school." 
 
 Thomdike reports three experiments showing the 
 great improvement of adults resulting from a relatively 
 short period of practice. I give the results of one of 
 these experiments. Ten hospital nurses, 21 to 35 
 years old, were given 2 hours and 25 minutes of 
 practice, five minutes daily except Sunday, in adding 
 one place numbers. The results are shown in terms 
 of the number of one-place numbers added in five 
 minutes. The first number of each pair represents the 
 initial ability and the second number represents the 
 ability after practice. 180-230, 200-430, ^^5-868, 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 79 
 
 225-460, 290-540, 150-280, 220-380, 235-570, 250-440, 
 260-540. This makes an average improvement of 
 88.7%. 
 
 Some Miscellaneous Factors Influencing Drill. — 
 
 Conard and Arps compared two methods of drill in 
 arithmetic. They divided 76 high school pupils into 
 two equal groups by the Courtis tests. One group was 
 drilled for eight periods on arithmetical fundamentals, 
 by the traditional method ; the other group was drilled 
 for the same time, but their method was to name only 
 the results. For example, instead of saying "six plus 
 four are ten", they simply looked at six and four and 
 said "ten". The latter group showed a great saving 
 of time. In every school subject, the best and most 
 economical procedures or methods for doing the various 
 processes should be experimentally determined, then 
 these processes and procedures should be followed. 
 
 Kirkpatrick compared two methods of learning to 
 multiply. He had two groups learn to multiply by 
 using a table. One group memorised the table, the 
 other group used the table from the beginning. The 
 latter method proved the most advantageous. Kirk- 
 patrick draws the following conclusion from his study : 
 "The results indicate that in many lines of teaching 
 there has been a tremendous waste of time, energy, 
 and interest in first memorising, then later practicing 
 the use of what has been learned." It is not economical 
 to acquire skills too far removed from their practical 
 use in life. I have performed an experiment in card- 
 soi-ting which gives results similar to those obtained 
 by Kirkpatrick in multiplication. The experiment was 
 as follows: Two subjects worked an hour a day for 
 two days sorting cards into five boxes. On the first 
 
80 THE PSYCHOLOGY OF LEARNING 
 
 day, one subject sorted cards for the whole hour, the 
 other subject, instead of sorting, studied the boxes 
 trying to learn their locations without sorting cards 
 into them. On the second day, both sorted the cards 
 into the boxes. By repeating the experiment several 
 times, having first one subject study instead of sort, 
 and then the other, it became quite evident that the 
 most economical way to learn to sort cards is to sort 
 them. Studying the boxes for an hour made learning 
 progress on the next day faster than it would have been 
 without study but not so fast as when the first hour 
 had been spent in sorting. 
 
 Chapman studied the effects of various forms of 
 external incentives on the drill effects in adding, can- 
 celling out and in the digit-symbol test. The incentives 
 used were knowledge of previous records, the learning 
 graph, and credit rewards. The motivated group did 
 much better except in cancelling out. In the latter no 
 incentive was needed. 
 
 Practice in Fundamental Mental Functions. — We 
 shall now turn from the question of strengthening 
 bonds by practice to the question of the general im- 
 provement of a function by practice. This is a wholly 
 different problem from those we have been considering. 
 We take up the question here because certain mental 
 functions or capacities may be considered the tools or 
 machinery of learning. The various forms of sensory 
 discrimination, for example, are necessary elements in 
 many forms of learning. The question of whether 
 these functions, important as elements of learning, can 
 be improved by practice is therefore a vital one. If 
 there are mental functions which are important in all 
 learning, and these functions can be improved by 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 81 
 
 practice, then we can improve a person^s general learn- 
 ing capacity. We turn to the experiments. 
 
 Training in Pitch Discrimination. — Studies in pitch 
 discrimination by F. 0. Smith in the University of 
 Iowa indicate that there is no improvement from prac- 
 tice. He studied children of all ages as well as adults. 
 As a result of his extensive investigations he makes 
 the follov/ing positive statement: "The sensitiveness 
 of the ear to pitch differences can not be improved 
 appreciably by practice. There is no evidence of any 
 improvement in sensitiveness to pitch as a result of 
 practice." It looks as if functions as simple as pitch 
 discrimination are little improved by practice. They 
 depend upon inherited nervous orgnisation. More com- 
 plex functions seem to be improved by practice at least 
 early in life. 
 
 Whipple, by means of a specially devised tachisto- 
 scope, gave several adults practice in the range of 
 visual attention and in visual assimilation. He found 
 no improvement after the first few days of work. This 
 early improvement was due to the subjects becoming 
 adapted to the conditions and methods of the experi- 
 ment. 
 
 Dallenbach, following Whipple, made similar experi- 
 ments, using school children as subjects. Whipple had 
 found that adults made practically no improvement. 
 Dallenbach found that children made considerable im- 
 provement. For bright and normal children, the im- 
 provement was rapid at first and then slow. For dull 
 and feeble-minded children, the improvement was very 
 slow at first, but continued throughout the experiment. 
 At the end of the experiment some of the dull and 
 feeble-minded children were as good as the average 
 
82 THE PSYCHOLOGY OF LEARNING 
 
 normal child. Two conclusions seem warranted. (1) 
 In fairly simple mental functions, improvement is pos- 
 sible in the case of children when it is not possible for 
 adults. (2) It sometimes happens that the ordinary 
 situations of school and of life fail to develop 
 even the primary mental functions. In such 
 cases definite, special drill seems to develop 
 these backward functions. Such being the case, 
 although we cannot speak from definite experiments on 
 the subject, it seems likely that the experiences of 
 early life can at least in some measure affect the de- 
 velopment of important mental functions. It seems 
 obvious that these functions would not adequately de- 
 velop without some external stimulation; then surely 
 the character and amount of such stimulation must be 
 of consequence. When the character and amount of 
 stimulation have not been adequate to develop a child 
 to the limits of his natural capacities, then specific 
 drill is highly effective and valuable. When the char- 
 acter and amount of stimulation have been sufficient to 
 bring about the natural growth and development of the 
 various aspects of mental capacity then specific practice 
 with a view to develop mental capacity is of little value. 
 It is of little value because the development has already 
 taken place. 
 
 Practical Inferences. — What practical advice shall 
 psychology give the teacher on the question of drill? 
 The experiments leave no doubt of the great value of 
 specific drill, of direct practice. By far the larger part of 
 elementary school work is concerned with habit-forma- 
 tion. The formation of these habits should be faced 
 squarely and directly. If a boy wishes to learn how to 
 pitch balls, he pitches balls, not once or twice, but in- 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 83 
 
 cessantly, day after day, month after month, year after 
 year. We should follow a similar procedure in the case 
 of spelling habits, reading habits, and all other habits to 
 be acquired in school. Is there a ceitain skill that it is 
 desirable I should possess? Then I must practice, 
 practice, practice. By and by I shall have the skill. 
 There is no reason why we should beat about the bush, 
 evade, or come at it indirectly. I must know exactly 
 what the skill is, have some good reason for desiring it, 
 then I should practice it vigorously, regularly, directly. 
 The school room, for the early years of child life, should 
 be a drill room. The drill periods should be short ; the 
 drills vigorous. During the drill, the attention of the 
 children should be of the highest order. They should 
 be working at the highest possible point of efficiency. 
 Usually the drills should be of only a few minutes in 
 length, not long enough to tire the children. Com- 
 petition, both individual and group, can be used, if used 
 wisely, to advantage. Every known legitimate device 
 should be used to make the practice effective and profit- 
 able, particularly should we have the pupil keep his 
 record and plot his learning curve. We should arouse 
 in him an ambition to make the curve rise a little every 
 day. We should show him definitely the skill that is 
 possible of attainment. He should know about how 
 long it will take him to acquire that skill. 
 
 Children delight to practice a newly acquired skill, 
 delight in pure practice apart from the use of the skill 
 in any concrete problem. When a child has, for ex- 
 ample, learned how to do long division, he will ask for 
 problems in long division merely for the pleasure of 
 exercising the new skill. It is a fundamental principle 
 of human nature that we like to do what we can do. 
 
84 THE PSYCHOLOGY OF LEARNING 
 
 particularly what we have just learned to do. This 
 delight in mere exercise can be maintained provided 
 the drills are short as well as vigorous. 
 
 Practice in school subjects should not, however, be 
 wholly formal. Every skill is only a means to some 
 end. The skill has no real and permanent value in 
 itself. It should therefore be used in its proper setting. 
 Spelling, reading, writing, arithmetical fundamentals 
 have no value in themselves. They are all merely 
 means to ends that are intrinsically valuable. We must 
 be able to spell and v/rite if we wish to communicate 
 with friends at a distance. We must be able to read 
 to get the pleasure and information that is possible 
 from reading. We must be able to add, subtract, mul- 
 tiply nad divide in order to be able to carry on the actual 
 business of our life. In acquiring these skills, we 
 should certainly use them in their real life connections 
 and settings, but not merely so. If a child writes and 
 spells only when writing a letter, it will probably never 
 write a letter. If a base ball pitcher pitched only in 
 games, he would never win many games. Much prac- 
 tice must be preparatory, anticipatory. Early life is 
 preparatory. Nothing can change that fact. Never- 
 theless, there should be much incidental drill. Situa- 
 tions should be devised in school and in the home that 
 duplicate or mimic life situations, such as keeping 
 store with its buying and selling and calculations and 
 computations. There should also be much letter writ- 
 ing, actual as well as fictitious. The child should have 
 much practice in using his newly acquired skills in their 
 proper settings. But he should have additional prac- 
 tice of a direct and formal sort also. The two pro- 
 cedures combined will give the efficiency v/hich the 
 situations of life demand. 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 85 
 
 Definite Procedure. — The thing* about habit-forma- 
 tion that should be made clear is this: The teacher 
 should understand definitely and clearly what she is 
 trying to do, the knowledge she is to help the pupils to 
 acquire and the habits she is to help them to form. 
 These facts should be clear and definite to her. And 
 she should make definite and systematic plans for the 
 work. The school work of the past has been too indefi- 
 nite, too hazy, too nebulous. Too often neither pupils 
 nor teachers knew where they were going. Objectives 
 must be known and clear. They must also be as close 
 as possible. There must be constant realisation of 
 aims. We can not put off all realisations, all fruits, 
 all rewards to an uncertain future date. We must set 
 a certain skill to be attained this week, perhaps even 
 in this single practice. Not only must ends be definite 
 and clearly known, but the time and place and manner 
 of practice. By— means of the standard tests she 
 done at any time is too often done at no time. The 
 teacher should therefore supervise in great detail the 
 whole work of practice on the part of the pupils, what 
 work is to be done at school, what at home, the exact 
 times of practice, the lengths of periods, the manner 
 of practice. By means of the standard tests she 
 should measure the progress of the pupils. Not only 
 should all these things be definite and clear to the 
 teacher, but they should be clear to the pupils as well. 
 The pupils should know what they are doing and why 
 they are doing it. They should know the precise 
 nature of the habit they are to form, what it will be 
 worth to them when they have it, how they are to pro- 
 ceed to acquire it. And, as already pointed out, they 
 should plot the course of the voyage they make in the 
 
86 THE PSYCHOLOGY OF LEARNING 
 
 process of acquirement. Not only should teacher and 
 pupils have a clear understanding of their aims and 
 their means of attaining them, but parents should have 
 this knowledge, too. Teacher, parent, and pupil are all 
 working together for the accomplishing of the one defi- 
 nite, clear-cut purpose. Too often the parents do not 
 know, more often still, do not understand the purposes 
 of the school. Under such circumstances they can not 
 properly co-operate. It must not be forgotten that the 
 schools are maintained by the parents for their chil- 
 dren. It sometimes looks as if schools were maintained 
 by teachers for teachers, that pupils are important only 
 because it is hard to have school without them, arid 
 that parents have no importance at all. Their wishes 
 are not to be considered. So easy it is to lose proper 
 perspective and to forget the proper relations of things. 
 
 Developing General Ability. — It does not seem that 
 the school can do much toward developing general capa- 
 city. There need be no training in the use of the senses. 
 Nature provides for this training in the ordinary course 
 of life. Probably most of the simple, elemental forms 
 of mental activity need no specific training. It may be 
 that in some cases they do, in case of exceptional chil- 
 dren; generally they do not. The various aspects of 
 association, learning, remembering, attending get suffi- 
 cient practice for their development in the process of 
 foiTning the needed habits, and acquiring the necessary 
 knowledge. Briefly, we do not need to have studies or 
 practices whose purpose is the development of the mind. 
 The life that we must necessarily live develops about 
 as far as development is possible. 
 
 We do not mean to say, however, that mental func- 
 tions develop without reference to the stimulation of 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 87 
 
 the environment. On the contrary, the development of 
 mental functions is absolutely dependent upon environ- 
 mental stimulation. What we do mean to say is that 
 the ordinary stimuli which are fairly constant to all 
 human environments are sufficient to bring about nor- 
 mal mental human development. No doubt extreme 
 differences in this environment can measurably influ- 
 ence human development. The evidence of the neurolo- 
 gist supports this view. Bumham*, for example, says, 
 *'We see from the genesis of the nervous system that 
 the one condition necessary for normal development is 
 a rich environment giving plenty of stimulation and 
 freedom for the nervous mechanism to develop in its 
 own way. This seem^s especially important for the 
 cerebral cortex and its dependent structures. The same 
 thing is emphasized, too, by all the cases of defect where 
 normal stimuli are shut off. In such cases, there is 
 always imperfect or arrested development." 
 
 What Amount of Skill Is Desirable? — A legitimate 
 question and a very practical one is this : V/hat degree 
 of skill should we attain in the various school subjects? 
 It is certainly not profitable to acquire great skill in 
 functions that are not to be used for a long time, for 
 unless practice is kept up in the interval, the skill will 
 deteriorate before the time comes for its use. If we 
 are always to consider economy, a skill should not be 
 acquired until near the time for its use. It is certain 
 that m.uch time is wasted in school because we do not 
 take this principle into account. We have the child 
 acquire numerous skills in the hope that some time, 
 somewhere he will have occasion to use them. Many 
 
 *W. H. Burnham, — The significance of stimulation in ihe develop- 
 ment of the nervous system. A. J. P. 28, 1917, p. 38. 
 
88 THE PSYCHOLOGY OF LEARNING 
 
 of these skills he will never use, and many others will 
 deteriorate before the time comes for their use. In all 
 of the school subjects we should make a careful study 
 of the skills which they give, and at the same time we 
 should make a study of the present and future need of 
 the child. The skills should be acquired with some 
 reference to these present and future needs. It 
 is certain that in many subjects, the amount that we 
 teach should be cut down, stripped of unessentials. We 
 should make a careful estimate of future needs, and 
 consider the relative nearness of these needs. In mathe- 
 matics, for example, the general needs are very few, the 
 fundamental operations. These are needed constantly, 
 or at least we can say that in any normal sort of life 
 they should be used constantly. They should be mas- 
 tered, and carried to a point where they can be used 
 with speed, certainty and facility. The same thing is 
 true of reading to get thought, of spelling, and of writ- 
 ing. There is a minimum of essentials which all normal 
 people should get to a fair degree of mastery. Beyond 
 the minimum of essentials, all other skills should be 
 mastered with reference to the time when they will be 
 needed. As a general rule we should master first the 
 skills that we shall need first. A great deal of time 
 could be saved for the children of the country if com- 
 petent people would select from the various subjects 
 the minimum requirements of skill that are desirable 
 for all. Whenever the special work v/hich a person 
 takes up requires a skill additional to those already 
 acquired, that skill can then be acquired. The skills 
 acquired for distant and uncertain uses should be few. 
 The immediate demands are too great to substitute for 
 them uncertain future demands. But when future de- 
 
ECONOMICAL LEARNING — SPECIAL FACTORS 89 
 
 mands are not uncertain they should be provided for 
 at the proper time, and the proper time is the one that 
 is the most economical, provided there are no other 
 factors. 
 
 Function of the Teacher. — The relation of the teacher 
 to the learners is a close and important one. Book has 
 enumerated and discussed the various ways in which 
 the teacher can be of service to the pupils in their 
 learning. He gives ten points, which may be condensed 
 into eight, and stated concisely as follows: 
 
 The teacher can — 
 
 (1) Help pupils to overcome difficulties as they appear; 
 
 (2) Help pupils to discover the best methods of study 
 and work ; 
 
 (3) See that pupils use the most economic methods in 
 forming habits ; 
 
 (4) Minimise the f orma,tion of interfering tendencies ; 
 
 (5) Help pupils to organise and assimilate details in a 
 natural way; 
 
 (6) Provide special incentives to effort; 
 
 (7) Make hygienic and environmental conditions of 
 learning favorable ; 
 
 (8) Develop in pupils an attitude of permanent interest 
 in work. 
 
 To these may be added another: The teacher can 
 help the pupils to understand the purpose and nature 
 of the habits to be formed. 
 
 EXPERIMENTS AND EXERCISES. 
 
 The facts of this chapter do not readily submit them- 
 selves to class room demonstration. However, the stu- 
 dents might very well make application of the facts to 
 the various branches of the public school. These 
 branches can be examined and the habit-forming 
 
90 THE PSYCHOLOGY OF LEARNING 
 
 aspects of each determined. The students can then 
 work out a scheme of drill or other form of practice for 
 fixing the habits, and in doing this, provide for the use 
 of the psychological facts that are applicable. For ex- 
 ample, let us take arithmetic. The actual number of 
 habits that should be formed in the elementary study of 
 arithmetic can be listed, so many in adding, so many in 
 subtraction, so many in multiplication, and so on. The 
 order in v/hich these habits should be established should 
 be worked out. The use of early habits in acquiring 
 later ones should be noted. After the elementary 
 habits have been provided for, a study should be made 
 of the use and application of these in fractions and per- 
 centage. On this basis, a course of study in arithmetic 
 as habit-formation could be made. Let the student 
 make an estimate of the amount of time necessary to 
 master the course outlined. 
 
 Take other branches and make similar studies from 
 the point of view of habit-formation. The plans of 
 different students should be compared and discussed. 
 
 REFERENCES. 
 
 Book, W. F., The role of the teacher in the most expeditious and 
 economic learning, J. E. P., I, 1910, 183. 
 
 Brown, J. C, An investigation on the value of drUl ivork in funda- 
 mental operations of arithmetic, J. E. P., II, 1911, 81; also J. E. P., 
 Ill, 1912, 485 and 561. 
 
 Donovan, M. E. and Thorndike, E. L,, Improvement in a practice 
 experimeyit under school conditions, A. ,J. P., XXIV, 1913, 426. 
 
 Fulton, M. J., An experiment in teaching spelling, Ped. Sem., XXI, 
 1914, 287. 
 
 Kirkpatrick, E. A., An experiment in memorising vs. incidental learn- 
 ing, J. B. P., V, 1914, 405. 
 
 Peters, C. C, lite influence of speed drills upon the rate and the ef- 
 fectiveness of silent reading. J. E. P., VIII, 1917, 350 
 
 Phillips, F. M., raluc of daihj hill in arithmetic. J. E. P., IV, 1913, 
 159. 
 
 Thorndike, E. L., Practice in the case of addition, A. J. P., XXI, 
 1910, 483. 
 
 Walliu, J. E. W-, Has the drill hccome obsolescent? J. E. P., I, 1910, 
 200. 
 
Chapter VI. 
 IDEATIONAL LEARNING. 
 
 The Nature of Ideational Learning.— By ideational 
 learning, we mean acquiring knowledge. In the last 
 analysis, as pointed out in Chapter I, there is little if 
 any difference between habit and knowledge. Both de- 
 pend upon established neural connections. In habit, 
 stimulus is connected with its motor response ; in knowl- 
 edge, idea is connected with idea. In forming habits, 
 neural bonds are established which connect certain 
 stimuli v/ith certain muscular responses; in ideational 
 learning, neural bonds are formed which connect cer- 
 tain ideas with certain other ideas. In habits, the mus- 
 cular response is the important element ; in knowledge, 
 although muscular action may take place, it is not an 
 important element, not an essential part of knowledge. 
 In knowledge, the important thing is sequence of ideas. 
 If action takes place, it follows upon the sequence of 
 ideas, and has no causal relation with them. In the 
 last four chapters, we have had in mind chiefly habit- 
 formation. In this chapter, our exclusive concern is 
 with ideational learning, the getting of knowledge. 
 
 Knowledge and Ideas, — -Our first concern is to get a 
 clear notion of the meaning of the words idea and 
 hnoivledge. The meaning given to each of these terms 
 is a simple and common-sense one. We shall call our 
 sensory experience with the world, primary experience, 
 and the revivals of sensory experience in the absence 
 of sensory stimulation, we shall call secondary experi- 
 
 [91] 
 
92 THE PSYCHOLOGY OF LEARNING 
 
 ence. Sensory experience takes the form of perception ; 
 secondary experience takes the form of ideas. Percep- 
 tions are the conscious processes representative of the 
 objects of objective experience, sensory experience. 
 Ideas are the conscious processes representative of ob- 
 jects in rvived or secondary experience. To illustrate : 
 Suppose I take a walk through the v^oods. I see, hear, 
 touch, smell, and taste various things. I have primary, 
 sensory, perceptual experience of birds, trees, flowers, 
 fruits, breezes, sounds, temperatures. Afterward, when 
 night has come, I sit in my room and experience the 
 trip all over again in the form of ideas. We shall, then, 
 use idea as the general name for the bits or elements 
 of revived experience. We shall use knowledge as the 
 name for connected ideas, remembering that one ele- 
 ment of the connection may be a perception, but the 
 second element is always an idea. 
 
 As I sit in my room and live the wood trip over again, 
 many ideas come to me, not only ideas of the trip but 
 other ideas that come up from every part of my past 
 experience. Within an hour, ideas come, revived from 
 almost every part of my past life. Various experiences 
 of the trip, owing to similarity with past experiences, 
 revive the past experiences, so that the experiences of 
 the trip serve as a means of tapping or sounding almost 
 the whole of past experience. The intricacy with which 
 all of our past experience is bound together is well 
 shown by the free association experiment. In this ex- 
 periment, the subject is given a word and told to write 
 all the other words which com^e to mind. The author 
 has the following sequence : sky, sun moon, clouds, rain, 
 umbrella, wet, woman, hat, hair, dress, skin, comb, eyes, 
 glasses, water. Why did these words — representing 
 
IDEATIONAL LEARNING 93 
 
 ideas — come in this order? In the case of my wood- 
 trip, I find that my ideas correspond to the perceptual 
 experience as it came during the trip, not in strict 
 chronological order, but all the ideas of or from the 
 trip are closely bound together. Do the words, sky, 
 sun, etc., written above, represent any particular expe- 
 rience, or do they come from a great variety of experi- 
 ences? They, of course, come from a great variety of 
 experiences, and it shall now be our purpose to examine 
 into the laws that determine their coming. 
 
 The Law of Association. — Let us suppose that on my 
 walk through the woods, I passed a large boulder and 
 saw a snake coiled up beside it. When the experience 
 comes back to me, I have an idea of the boulder and 
 also of the snake. Or, when I go by the same place 
 again and see the boulder I also think of the snake. 
 This is typical of all experience. What is experienced 
 together in perception, comes back together as ideas. 
 Processes that take place together in the brain, or come 
 in close succession are connected, they are really a part 
 of one process, as they leave in the brain some form, a 
 trace of this connection, so that later a revival of one 
 of one process, and they leave in the brain some form, a 
 and boulder were together in the woods ; they are like- 
 wise together in my mind ; the creek, and the shore, and 
 the frogs, and the dragon-flies, and water lilies were 
 together in my experience; they are likewise together 
 in my mind. Clouds, and lightning, and thunder, and 
 rain, and mud are together in the world, they set up 
 simultaneous and immediately successive processes in 
 my brain and are consequently together in my mind. 
 Corresponding to all the outer world, with its spatial 
 and temporal and causal relationships, I build up an 
 
94 THE PSYCHOLOGY OF LEARNING 
 
 inner world of ideas. Whatever relationships and se- 
 quences exist in the objective world, hold also with my 
 ideas. The outer world is lawful, orderly, systematic, 
 so also is my world of ideas. 
 
 It is obvious that the various objects of perceptive 
 experience are experienced in myriad connections. A 
 horse, for example, is experienced in connection with 
 buggies, wagons, trucks; as being ridden, as being 
 driven; they are seen in pastures; they are seen run- 
 ning away; they are seen dying; they are seen lying 
 on the streets with broken legs ; they are seen with colts 
 following them; in a word, they are experienced in a 
 great variety of situations. At any particular time, if 
 horse is suggested to me, what one of its ideational 
 connections will follow? Both observation and labora- 
 tory experiment have revealed several determining fac- 
 tors, as follows: primacy, the original connection; 
 recency, the most recent connection. The other factors 
 are frequency ^ meaning the number of times the partic- 
 ular connection has been repeated in perceptual or idea- 
 tional experience; intensity, meaning the intensity or 
 vividness of the experienced connection ; and mental set 
 or attitude. By mental set we mean that the attitude 
 or feeling tone of the original experience is a determin- 
 ing factor in the revived experience. Other factors be- 
 ing equal, the mood or attitude that I am in will deter- 
 mine that sequence, in any given case, which corre- 
 sponds to previous sequences in the same attitude. At 
 any time, if I see a horse, or if the idea of horse comes 
 to mind, the idea which the idea hoi^se will first suggest 
 or arouse, will depend upon the most pervious neural 
 path at that moment, and the most pervious path or 
 connection will depend upon the various strengths of 
 
IDEATIONAL LEARNING 95 
 
 the factors above mentioned. The first idea may be of 
 a run-a-way horse, depending upon recency, the next 
 idea may be of a horse piilhng a plow depending upon 
 frequency, the next idea may be of a horse pulHng a 
 buggy depending upon primacy; and the next idea may 
 be of horses pulhng a hearse depending upon mood. 
 Always these various factors are at work ; always there 
 is a conflict for the determination of the course of asso- 
 ciation. The path that is most open leads to the first 
 connected idea; the path that has the least resistance 
 next in order determines the next idea; and so on, till 
 the trend of association is diverted by perception or till 
 it is carried internally to another center or matrix. 
 
 Thinking. — We shall use the v/ord thinking as a gen- 
 eral term to designate the free flow or passage of ideas. 
 This flow or passage depends entirely upon the law of 
 iissociation. While we are awake, there is in us a con- 
 stant succession of ideas. Usually a perception initiates 
 a series of ideas. For example : I walk along the street. 
 I receive impressions through all the senses. The sights 
 and sounds and odors arouse trains of ideas. I pass a 
 man. His name comes to me. Then I think of various 
 experiences which I have had with him — when I first 
 saw him, when I last saw him, of his business, his 
 family. These ideas are in process Vv^hen I pass a dough- 
 nut shop, v/hich sets up a different series of ideas. The 
 smell of doughnuts makes me think of my childhood, 
 the early home, my mother and her cooking. Nearly 
 everything which I pass as I go down the street is rich 
 in association and sets going a series of ideas. 
 
 All day long, wherever we are and whatever we may 
 be doing, we have perceptions. These perceptions set 
 going trains or series of ideas. A closely related series 
 
96 THE PSYCHOLOGY OF LEARNING 
 
 of ideas leads presently to another. And so the ideas 
 flow along till a new perception breaks in and sets up 
 a different series. 
 
 Reasoning.^ — Reasoning is the flow of ideas evoked by 
 a situation, new or partially new. Since the situation 
 is new, it arouses no response through habitual connec- 
 tions. For all situations that have been repeatedly met, 
 there follow habitual forms of response that have been 
 organised in connection with situations. When we meet 
 a situation partially new and no instinctive or habitual 
 response follows, we have to stop and wait for the asso- 
 ciative processes to suggest a solution. The process of 
 reasoning may be illustrated as follows ; The outlet of 
 my bath tub was stopped up so that the water would 
 not drain out. The following ideas came to me — there 
 is an obstruction just at the bend outside of the tub, 
 probably there is a way to get in there and get the ob- 
 struction out, I look and find that there is. I can get 
 in by unscrewing a nut. But the nut is in a place diffi- 
 cult to get at, and besides I have no wrench handy to 
 unscrew the nut. The pressure of the water in the tub 
 is not great enough to porce the obstruction out. If I 
 should fill the tub full of water perhaps the pressure 
 would be great enough to force the obstruction out. I 
 try it and it will not do it. How can I get greater pres- 
 sure ? The pressure of the water in the water pipes is 
 strong enough to force it out. But how can I utilise 
 this pressure ? If I could put a piece of rubber hose on 
 the pipe from which the water passes into the tub and 
 hold the other end of the hose at the outlet of the tub, 
 the force of the water might force the obstruction out. 
 I try it. It will not do it. If I could wrap a rag around 
 the end of the hose at the outlet of the tub so as to force 
 
IDEATIONAL LEARNING 97 
 
 all the water into the drain pipe, the pressure would be 
 great enough to force the obstruction out. I try it and 
 it works. The obstruction is forced out. The illustra- 
 tion is fairly typical of all reasoning. We meet a situa- 
 tion that has not been met before. The situation sug- 
 gests or arouses various ideas from our past experience. 
 Certain elements of the situation are like past experi- 
 ences. An idea comes to us. We try it out. It may 
 work. It may not. If it does not work, it is because 
 we do not have accurate enough information about the 
 forces with which we deal. In the above illustration, I 
 did not know how hard the obstruction would be to 
 force out. I did not know whether the pressure of the 
 tub full of water would be sufficient or not. I had to 
 try and see. Various ideas of getting pressure behind 
 the obstruction came to me. I tried out the ideas till 
 success came. 
 
 In the above illustration, an actual situation had to 
 be met. It often happens that we meet a hypothetical 
 situation as when we have to answer a question. The 
 thought processes are the same, but the illustration 
 shows that we can not be sure of an answer to a hypo- 
 thetical question unless the conditions are accurately 
 stated. Suppose I have a ring of iron and a sphere of 
 iron that will nearly but not quite go through the ring. 
 How can I get it through ? Now if my past experience 
 has taught me that heat expands iron, the idea may 
 come to me that if the ring is heated, it will expand and 
 possibly let the ball pass through. If I have accurate 
 knowledge about the expansion of iron due to heat, I 
 can answer the question by making accurate measure- 
 ments of the ring and ball. If I do not have this accu- 
 rate knowledge and can not make very accurate meas- 
 
98 THE PSYCHOLOGY OF LEARNING 
 
 urements, I shall have to try and see, I can not give an 
 accurate theoretical answer. The latter is often the 
 case in our theoretical reasoning. The facts that we 
 have are not full enough and not accurate enough to 
 enable us to make an exact answer. Much of the ordi- 
 nary reasoning of men has this defect. Much that is 
 taught in school is untrue because due to inferences 
 from incomplete or erroneous data. We have been 
 taught much in physiology and hygiene that is now dis- 
 covered to be untrue, much about digestion — what we 
 ought to eat, how we ought to eat it and when we ought 
 to eat it — mxuch about diseases, their causes and their 
 supposed cures. All the social sciences are full of false 
 doctrines because of incomplete and inaccurate knowl- 
 edge. 'It is so often true that v/e can not gauge cor- 
 rectly the forces or principles that enter into our rea- 
 soning, that we can seldom be sure of our conclusion 
 unless we can put out tentative solutions to the prac- 
 tical test. Reasoning is usually the means by which 
 our past experience suggests possible solutions to the 
 problem or situation which confronts us. The public 
 schools can be of great service to our people and to our 
 country by making these facts clear to children, but 
 this point we must take up in a later paragraph. 
 
 There is nothing new about the process of reasoning ; 
 nothing different from the ordinary flow of ideas; no 
 new law in operation. The flow of ideas, hov/ever, is 
 limited by the situation. We maintain the situation, or 
 external conditions maintain it for us — the water per- 
 sists in staying in the bath tub — and one series of ideas 
 after another is initiated. Each series of ideas is the 
 result of past experience. Our problem is solved, theo- 
 retically, when an idea comes that satisfies us, that 
 
IDEATIONAL LEARNING 99 
 
 seems in the light of our experience, to fit the situation. 
 The idea may be adequate; it may not. All depends 
 upon our past experience and the accuracy of our knowl- 
 edge of the situation. 
 
 Primary Experience.-— It is clear that in ideational 
 learning and in the processes of thinking and reasoning, 
 all depends upon primary experience. Ideas, the ele- 
 ments of knowledge, are derived from perceptual ex- 
 perience. We must therefore take pains to see that 
 children have wide experience, that the facts of experi- 
 ence are corrctly interpreted. The child must have a 
 first-hand experience with nature — with all the forces 
 of nature and with all the objects of nature — with ani- 
 mals and plants, and vvith machinery. It must learn 
 their names, and their uses or functions. It learns by 
 seeing, hearing, touching, tasting, smelling, etc. In 
 getting knowledge of things, nothing can fully take the 
 place of direct experience. No description of the taste 
 of sugar could ever give us a clear notion of the taste 
 of it if we had never tasted it. And if w^e have tasted 
 it, no description is necessary: No description of a 
 cow could give us a clear notion of one. All of our 
 knowledge must be built upon actual sensory experi- 
 ence. If a person lacks some sense, as sight, from 
 birth, no description can make clear to him what expe- 
 rience is like in the field of the sense that is lacking. 
 The basis of all knowledge must come through the 
 senses. Throughout all the years of a child's life, it is 
 getting this knowledge; it is learning the world in all 
 the world's manifold aspects. Day after day, and year 
 after year, the individual is having new experiences, is 
 learning new aspects of the world. The early life of the 
 child must be economised, useless knowledge eliminated 
 
100 THE PSYCHOLOGY OP LEARNING 
 
 and economical methods of learning used, so as to give 
 plenty of time for mastery of the material world in all 
 its various phases. 
 
 The Organization of Experience.-— Since we get expe- 
 rience through the senses, the getting of experience is 
 determined by the circum^stances of life, by the place 
 where we live and the time when we live. The boy on 
 the mountain has mountain experience ; the boy on the 
 plain has plain experience. Boys that lived at a certain 
 time and were at a certain place, witnessed the Battle 
 of Gettysburg, others saw Lincoln, others saw Wash- 
 ington, others witnessed the San Francisco earthquake. 
 The organisation of experience, however, is independ- 
 ent in some measure, of the original order of experi- 
 ence. We get experience as the chances of life deter- 
 mine. In accordance with the law of association, things 
 are bound together which have been experienced to- 
 gether. But things which we have experienced together 
 may have for us no useful connection. Chance may de- 
 termine that I see a rattlesnake and hear a hoot owl at 
 the same time, but this connection of the two may 
 have no use for me. We can organise our experience 
 by thinking it over in helpful relations. The important 
 idea for me to have when I see a rattler, is not of an 
 owl but that the snake may bite me and that its bite 
 is poisonous. Out of the raw ma^terial of the day's ex- 
 perience, we reconstruct a useful world, the world that 
 concerns us. The objective V\^orld is not lawless. The 
 order in it determines the order of our ideas. But cer- 
 tain relations in the world are more important to us 
 than others. We therefore reorganise the world by 
 thinking over our experience in the relations that are 
 important to us. The causal relation is usually most 
 
IDEATIONAL LEARNING 101 
 
 important to us. The causal sequences exist in nature, 
 and it is best for us to fix them in memory by repeated 
 thinking of the ideas in the causal sequence. But many 
 things that happen together have no causal relation, as 
 my seeing the snake and hearing the owl. Repeated 
 experience on our part is necessary to enable us to dis- 
 tinguish true causality from mere accidental concomi- 
 tance. Older people, as our parents and teachers who 
 have had more experience, can be of great service to us 
 in the organisation of our experience. They can point 
 out to us the important relationships and aid us in es- 
 tablishing them in the sequence of our ideas through 
 repetition. 
 
 Getting Knowledge. — To live in this world we must 
 be able to control it. To control it, we must have knowl- 
 edge of it. Since some aspects of the world are more 
 important to us than others, some knowledge is more 
 important than other knov/ledge. One function of the 
 school is to help the child to get and organise useful 
 knowledge, such as knowledge of the mechanical, phys- 
 ical and chemical aspects of the world ; knowledge about 
 animals and plants, knowledge about the human body — 
 its organs and their functions, how to keep healthy, 
 how to keep strong, — civic and social knov/ledge. Just 
 what this knowledge should be in detail we shall not 
 here enquire. We want simply to get a general view 
 of it. The child must learn certain aspects of the world 
 to enable him to control it and to direct his life in it. 
 The question we wish to raise here is how we can eco- 
 nomically get this knowledge. 
 
 But first let us recall fully the nature of knowledge. 
 Knowledge consists in the names of things coupled to 
 the ideas of the things and the functions or uses of the 
 
102 THE PSYCHOLOGY OF LEARNING 
 
 things coupled to the names. These we must know. 
 From infancy the process of getting this knowledge 
 goes on. The child learns how to move about in its 
 little world, learns v/hat will hurt it and what will not, 
 what it can manipulate and what, not. Year after year 
 more and more knowledge is added until at maturity we 
 have usable knowledge about most material things. Our 
 method of getting all this knowledge is through per- 
 ception. There is no other way. We see the objects 
 of the world, touch them, taste them, smell them, 
 hear them. We note their actions. We experiment 
 with them and thereby learn their characteristics and 
 their ways of reacting. We wish, for example, to learn 
 the various factors that determine the germination of 
 seeds. We therefore try to germinate seeds under all 
 possible conditions, v/ithout air, without moisture, with- 
 out heat, and with various combinations of these factors 
 and with various degrees of these three factors. As a 
 result of our experiment, we learn the factors that are 
 necessary for germination. All our basic knowledge of 
 the world must be got either by careful observation of 
 phenomena as we meet them or from the results of 
 carefully planned experiments. The child must be 
 taught not only to be a close observer, but a critical 
 observer. 
 
 Analysis of Experience. — The process of getting 
 knowledge and of organising knowledge is constantly 
 one of analysis. In the first place our notion of quali- 
 ties and all abstract characteristics comes through a 
 process of analysis, and analysis is possible through 
 experiencing things in a great variety of relationships. 
 We get the idea of redness by experiencing different 
 things of different sizes and shapes having this char- 
 
IDEATIONAL LEARNING 103 
 
 acteristic. We get the notion of triangularity, of 
 squareness, of roundness, etc., in a similar way. From 
 a variety of experiences our general notions crystallize 
 out. The world at first is largely unitary. Every day 
 of our life, it falls apart into its separate parts, and the 
 parts into their separate characteristics. The relation- 
 ships among the parts grow ever more subtle and intri- 
 cate. As v/e grow older, bonds are possible between 
 ideas that were not possible before because the ideas 
 did not and could not exist for us before. 
 
 Meaning.-— The most important thing about an idea is 
 its meaning. The meaning of an idea is another idea 
 closely associated with it. Since an idea may have 
 many such associations, it may have many meanings. 
 The mxost important meaning of an idea is the use to 
 which the thing represented by it may be put. Of all 
 the things that the idea of an object may bring to our 
 mind, v/hat most concerns us is what we can do with 
 the thing, how we can make it serve our purposes, how 
 it is likely to affect us, whether it will harm us or do 
 us good. 
 
 Ice is cold, hard, slick. It may be used to keep our 
 food cool in the refrigerator, to cool our drinking water, 
 for making ice cream, etc. A pencil is a thing made of 
 wood, with lead inside, but more important, it is a 
 thing with which I can write a letter. An ax is a thing 
 used for chopping, a saw for sawing, a fork for con- 
 veying food to the mouth. Shakespeare was a drama- 
 tist; Longfellow, a poet. Sympathy is feeling loith 
 another. The meaning of an idea is, then, another idea 
 that represents some characteristic, some relation, some 
 use. Most of our knevv^ledge is knowledge of meanings. 
 Most of our progress in knowledge consists in learning 
 
104 THE PSYCHOLOGY OF LEARNING 
 
 new meanings. The development or growth of the 
 meanings of the same thing are almost unlimited. A 
 flower to a child is not the same thing that it is to a 
 man or woman, and far from the same thing that it 
 is to a technical botanist. To the child it may be a 
 pretty thing that has an agreeable odor ; to the botanist, 
 it may still be this and in addition a means of develop- 
 ing seeds for propagation. A stone may be to a boy 
 merely a thing to be thrown at a dog, to the geologist 
 it tells a long story about a most interesting past. Long 
 and intimate experience with objects reveals meanings 
 before undreamed of, subtle relationships, resemblances 
 before unnoticed. How different from ours is the bota- 
 nist's notion of a plant! What different aspects does 
 the world take on to the mathematician, the physicist, 
 the chemist, the geologist, the zoologist, the psycholo- 
 gist! How different is the world of the farmer from 
 that of the banker or merchant ! 
 
 As the years go by we severally build up for our- 
 selves our own particular kind of world with meanings 
 suitable to our needs and uses. To the doctor it be- 
 comes a world of disease; to the preacher, a world of 
 sin; to the physicist, it is a world of forces. At bot- 
 tom, it is all the same world, but we have seized upon 
 different aspects of it and thereby narrowed our inter- 
 est in it, and circumscribed our knov/ledge of it. This 
 course is necessary, because since we of necessity must 
 deal with and manipulate different aspects of the world, 
 we must know these aspects with more fullness than 
 we knov/ other aspects that do not directly concern us. 
 Trouble comes v/hen we forget that our world is only 
 a partial world and not the whole world. I may be 
 interested only in the top of things; others are inter- 
 
IDEATIONAL LEARNING 105 
 
 ested in the bottom. I may be interested only in the 
 outside of things; others are interested in the inside. 
 It takes both top and bottom, both outside and inside to 
 make the whole. 
 
 The school should help the child to organise the world 
 with reference to useful meanings. The child should 
 always be asking the question, what does this thing 
 mean ? What is its true significance ? What caused it ? 
 What will it cause ? What is its relation to other things ? 
 In history, for example, what are the causes? What 
 results follow? What are the underlying movements? 
 What does Socrates mean? What, Napoleon? What, 
 Lincoln? What means the Renaissance? What, the 
 French Revolution ? In literature, what means Oedipus 
 Tyrrannuus? Macbeth? King Lear? Othello? Ro- 
 mola? The Scarlet Letter? 
 
 A defect in our acquiring knowledge and meanings 
 is that often we get the knovvledge in an abstract, iso- 
 lated sort of way apart from the real situations of life. 
 As a result, we do not really have knowledge of an 
 actual world, we do not have clear comprehension, and 
 when situations arise when the knowledge would be 
 available, the right idea does not readily come to us 
 because the knowledge was not learned in connection 
 with that particular kind of situation, consequently 
 the situation often fails to arouse the right association. 
 
 Reasoning Specific— Is reasoning capacity general or 
 is it specific? Can we speak of a person being a good 
 reasoner in general or do we have to say he is a good 
 reasoner in such and such fields and a poor reasoner 
 in other fields? These questions can be answered in 
 the light of the facts already given. Since reasoning 
 depends upon experience, one can not reason in fields 
 
106 THE PSYCHOLOGY OF LEARNING 
 
 where he has had no experience. One may have knowl- 
 edge ever so extensive in other fields, but if he have no 
 botanical knowledge he can not reason in the field of 
 botany. There are some aspects of reasoning, hov/- 
 ever, that are in a measure general. There are certain 
 criteria for the testing of truth that are of general 
 validity. The assumption that a phenomenon always 
 has a cause is one that can be made in all fields. The 
 practice of being cautious, of waiting until all the facts 
 are examined is of general validity, as is also that of 
 putting a conclusion to the practical test of experiment. 
 Training in reasoning must, therefore, be rather spe- 
 cific. If one wishes to be a great reasoner in the field of 
 physics, he must get a large experience in that field. 
 He must learn all the laws and principles and be familiar 
 with the methods of the science. One can be trained 
 in reasoning, (1) in getting a wide experience in the 
 field wherein he is to reason, (2) in the matter of be- 
 ing cautious and waiting to examine all the facts, and 
 (3) in putting the conclusion to the practical test of 
 experiment or experience. 
 
 Measuring Reasoning Capacity. — How well one can 
 reason, or how well one can learn to reason depends 
 upon hereditary factors as well as upon experience. 
 People of the same experience have different capacity 
 to reason. Some people, by reason of inheritance, seem 
 able to know what is significant, what relations are im- 
 portant, while other people have little ability to see 
 relative values, and it seems they can never learn to 
 do it. Our ability to reason, then, has certain hereditary 
 limitations. Since reasoning depends so largely upon 
 experience and since few people have had anything 
 like the same experience, its measurement is difficult, 
 
IDEATIONAL LEARNING 107 
 
 and can be accomplished only approximately. How well 
 one has established certain logical relationships, is to a 
 considerable extent indicative of one's ability to reason. 
 The measurement of these verbal relationships is then, 
 in a measure, a measurement of reasoning capacity. 
 Such is accomplished by the opposites test, the genus- 
 species test, the part-whole test, and other similar tests, 
 also by the analogies test. Certain tests, as those of 
 Bonser, have been specifically devised to test reasoning 
 capacity. Specific reasoning tests are likely to have the 
 defect that they demand specific experience and train- 
 ing ; for example, problems in arithmetic are good meas- 
 ures of reasoning capacity only if those measured have 
 had equal opportunity to learn the principles of arith- 
 metic. 
 
 Verbatim Learning.—- Learning the actual words of a 
 selection is a task not often required of people gen- 
 erally, but in all grades of school some verbatim learn- 
 ing is necessary. Experiments have shown that learn- 
 ing by small bits, a line at a time or a few lines at a 
 time, is not the most economical way. There is a sav- 
 ing of time in the learning of either prose or verse by 
 working on a large segment at a time, reading this 
 large segment clear through from beginning to end till 
 the part is learned. In the experimental work it ap- 
 peared that there is a saving in the whole method on 
 selections up to one 240 lines in length, as compared to 
 learning the same selections by small bits or the part 
 method. No experimenter has used selections longer 
 than 240 lines. 
 
 Each of the two methods, however, has its advan- 
 tages and disadvantages. If a selection is divided into 
 parts and each of the parts learned, and the time for 
 
108 THE PSYCHOLOGY OF LEARNING 
 
 learning the separate parts added, it is found usually 
 to be less than the time for learning the selection by 
 the whole method. But after the selection is learned 
 in parts, it takes more time to cement the parts to- 
 gether, so that the total is greater than the time re- 
 quired by the whole method. The weakness of the 
 whole method is that it is difficult to maintain a high 
 degree of attention throughout a long selection. Fur- 
 thermore, there is much over learning. Some parts are 
 learned long before the others. In spite of both of 
 these factors, the whole method proved more econom- 
 ical. A combination of the two methods in a way to 
 utilise the good points of each is probably to be recom- 
 mended in practice, such a combination, for example, as 
 learning a segment of several lines, then another; oc- 
 casionally going over it from the beginning as far as 
 learned, and also occasionally going through the entire 
 selection. 
 
 The experimental work on which the preceding state- 
 ments are based, was done with meaningful material in 
 prose and verse. Pechstein has recently reported that 
 with nonsense material, the part method is the best, 
 and gets favorable results by a combination of the two 
 methods. 
 
 How to Study. — ^When a student sits down to study a 
 lesson, how should he proceed ? Of course, the answer 
 depends somewhat on the kind of lesson, whether in 
 chemistry, history, or mathematics, for example. Gen- 
 erally, a lesson contains but a few facts or principles. 
 The student should first read through the assignment 
 as a whole to get the general drift of the argument or 
 description, then he should go back and pick out the 
 
IDEATIONAL LEARNING 109 
 
 essentials. As a rule, one should underline in the text 
 the crucial, fundamental statements, so as to make it 
 easy to turn through the pages and get the funda- 
 mental thoughts. The student should make sure of 
 what the author says, of the meaning and of its sig- 
 nificance. He should always put the question: What 
 does this mean ? Especially, What does it mean to me ? 
 The author should be translated into the understanding 
 and experience of the student. In reading a book, it 
 should always be our task to find out the author's mean- 
 ing. Certainly we should not attempt to criticise be- 
 fore we know what the author means. The student 
 should early learn to discriminate between fact and 
 theory or opinion, and should always note carefully 
 whether an author is giving facts, or the author's 
 opinions, or the opinions of some one else. Unfortu- 
 nately many people go through life without learning to 
 discriminate between a fact and an opinion. 
 
 EXPEPvIMENTS AND EXERCISES. 
 1. Association. The object of this experiment is to 
 make a study of the factors that determine the con- 
 nection of ideas. In a class experiment nothing beyond 
 illustration should be attempted. The free association 
 experiment will serve well enough for this purpose. The 
 instructor should prepare a list of v/ords, then pro- 
 nounce them one at a time to the class. When a word 
 is pronounced the students are to write down the first 
 word that comes to their minds. Let the students de- 
 termine the factors operative in the case of each word, 
 considering recency, primacy, frequency, intensity, 
 mental set and emotional factors. The experiment 
 should make it clear that there is nothing in the nature 
 
110 THE PSYCHOLOGY OF LEARNING 
 
 of ideas as such that bind them together, that the way 
 in which they become bound together is a matter of 
 experience. 
 
 2. Reasoning. The object of this experiment is to 
 show that reasoning depends primarily upon experience 
 and the abihty to recall the experience when it is 
 needed. The instructor should prepare a number of 
 problems and questions for solution and ansv/er. The 
 following will serve as examples: If a rope were 
 stretched entirely around the earth at the equator and 
 then lengthened six feet, how much space would there 
 be between the earth and rope, supposing the distance 
 to be made equal all around? Suppose three inch-cir- 
 cles are tangent, each to the other two, what is the 
 area of the space between the circles? Suppose a bar 
 of iron is riveted to a similar bar of copper and the 
 combined bar is then heated in the middle, which way 
 v/ill it bend? If a metal ball four inches in diameter 
 weigh 50 pounds, how large must a ball of the same 
 metal be to weigh 100 pounds ? If a cube of ice weigh- 
 ing ten pounds melts in an hour, how long will be re- 
 quired for a tv/enty-pound cube to melt, the tempera- 
 ture of the surrounding air being the same? 
 
 Give the above and similar problems to the students 
 and let each report on the reasoning processes that fol- 
 low. When a problem can not be solved, determine 
 whether it is because of lack of experience or inability 
 to recall the experience. 
 
 3. * 'Knowledge consists in the association of the 
 name of a thing with the idea of the thing, or the func- 
 tion or some characteristic of a thing with the name 
 or idea of the thing." Let the student verify this state- 
 ment by an examination of various kinds of knowledge. 
 
IDEATIONAL LEARNING 111 
 
 4. Give several illustrations showing the difference 
 between primary and secondary experience. 
 
 5. Make a study of meaning by getting the responses 
 of students to various characters in literature and his- 
 tory and to objects in nature. Put the questions as 
 follows: Who was Caesar? Pericles? Shakespeare? 
 Hamlet? Othello? Antigone? What is sulphur? 
 Iron? Platinum? Heliotrope? Water? Oxygen? 
 
 The experiment should shov/ that the same person or 
 thing may have different meanings to different people. 
 
 Interesting facts are learned by reading short bits of 
 literature to the class and having each student give his 
 interpretation. 
 
 6. "All knowledge of function and characteristics is 
 analytic." Let the student verify this statement by an 
 examination of many different kinds of knowledge. 
 
 7. The instructor can measure the reasoning capac- 
 ity of the members of the class by giving several con- 
 trolled association tests, several completion tests, and 
 by giving specific problems. The elements needed in 
 the solution of the problems must have been v/ithin 
 the experience of the students. In exercise 2 above, 
 the problems may very well go outside the experience 
 of the students, but in a study of reasoning ability, the 
 problems must be within the experience of the students. 
 The best reasoner is he who makes the best use of his 
 experience. 
 
 8. In Chapter VI, the students are asked to apply the 
 facts of the chapter to the habit-formation aspects of 
 the public school curriculum. Make a similar applica- 
 tion of the facts of this chapter to the ideational learn- 
 ing involved in the public school branches. Show, for 
 example, that in history, the child is learning the names 
 
112 THE PSYCHOLOGY OF LEARNING 
 
 of people and places and the causes and consequences 
 of events. The students should carefullj^ work out the 
 knowledge aspects of each school study. 
 
 REFERENCES. 
 
 Brown, W., Practice in associating color names with colors, P. R., 
 XXII, 1915, 45. 
 
 Practicing in associating numter names with nuniber symhols, 
 P. R., XXII, 1915, 77. 
 
 Dewey, J., Hoiv we think, 1910. 
 
 Hall, G. S., Some psychological asj^ects of language teaching, Ped. 
 Sem., XXI, 1914, 256. 
 
 Judd, C, H., The psychology of high school siihjects, 1915, Ch. XVII 
 on Generalised experience. 
 
 Lakenan, M. E., The tvhole and part methods of memorising poetry 
 and prose, J. E. P., IV, 1913, 189. 
 
 Otis, A. S., Do we think in words? P. R., XXVII, 1920, 399. 
 
 Pechstein, L. A., Whole vs. part methods in learning nonsensical 
 syllahles, J. E. P., IX, 1918, 381. 
 
 Alleged elements of waste iti learning a motor problem by the 
 ''part" method, J. E. P., VIII, 1917, 303. 
 
 Pyle, W. H. and Snyder, J- C, The most economical unit for com- 
 mitting to memory, J. E. P., H, 911, 133. 
 
 Thorndike, E. L., Repetition vs. recall in memorising vocabularies, 
 J. E. P., V, 1914. 
 
CHAPTER VII. 
 THE RETENTION OF EXPERIENCE. 
 
 Learning and Memory. — Learning would be impossi- 
 ble without memorj^ To learn means to become 
 changed, different from what we were before. If the 
 difference is not retained, at least for some length of 
 time, then we have not learned. Learning and memory- 
 each involves the other. To learn means that we must 
 have memory; to remember means we must have 
 learned. 
 
 Memory — the retention of experience — is one of the 
 most remarkable facts of human life. Any bit of ex- 
 perience leaves us different from what we were before. 
 Every new habit, every repetition of an old habit, every 
 bit of new knowledge, makes us a somewhat different 
 person. This is because we are the sum of our experi- 
 ences and tendencies; because we are the result of our 
 accumulated experience. 
 
 It is memory alone that puts meaning into life. A 
 merely perceptual life would be meaningless. Every 
 perception arouses ideas which have resulted from past 
 experience, and v/hich give the perception meaning. The 
 importance of m.emory, the part it plays in life, is at 
 once evident when one tries to imagine what life would 
 be without it. If every new experience were to us as 
 if it were our first, if it called forth no echoes of a 
 past, it would not really be experience. Experience 
 must have a longitudinal dimension as well as a trans- 
 verse one. It consists in a union of past and present.^^ 
 
 [113] 
 
114 THE PSYCHOLOGY OF LEARNING 
 
 Learning and memory are really different aspects of 
 the same thing. To learn means to become different. 
 To remember means to retain the difference. Much of 
 the experimental work in memory might well be called 
 studies of learning, and many learning experiments 
 might v/ell be called studies in memory. It is all a mat- 
 ter of emphasis. In this chapter we are to consider 
 those facts which bear more especially on the retention 
 of experience. We shall use the term memory in a 
 general sense as synonymons with the expression 
 retention of experience. 
 
 Memory and Age. — Ability to retain experience, as 
 far as it can be measured by experimental means, im- 
 proves with age from the earliest time in childhood 
 when such measurements are possible, up to maturity 
 or near maturity. Of course, retention is different with 
 different aspects of experience, but age brings improve- 
 m_ent in all aspects. If we show children a number of 
 objects, a number of pictures of objects, a number of 
 names of objects, or a number of abstract words, and 
 then determine later how well the experience is re- 
 tained, we find that it is better with each succeeding 
 year till near maturity. If v/e read a story to children, 
 and later determine hov; well the story is retained, we 
 find retention better up to near maturity. There seems 
 little increase in ability to retain experience after about 
 thirteen years of age. In fact experiments often show 
 a falling off of ability to reproduce stories during the 
 high school age. It is quite likely that this apparent 
 decrease in memory capacity is due to a certain atti- 
 tude of the subjects and not to any decrease in retentive 
 capacity. In the earlier years, children reproduce, par- 
 rot-like, stories read to them, without discrimination. 
 
THE RETENTION OF EXPERIENCE 115 
 
 Older children reproduce only the salient facts, through 
 habit omitting details. This attitude results in a lower 
 score. A careful study of all the experimental work 
 in this field leads one to the conclusion that retentivity, 
 as far as it can be measured by experiments, improves 
 considerably to about the age of adolescence, then more 
 slowly to the age of physical maturity. There is no 
 undoubted evidence of a decline of retentivity before 
 middle age. Improvement in retention due to age is 
 shown in tables 5, 6, and 7 : 
 
 TABLE 5. 
 
 THE RELATION OF MEMORY TO AGE AND SEX (PYLE). 
 
 CONCRETE ROTE MEMORY— CITY CHILDREN. 
 
 f Boys. ^ I Girls. ^ 
 
 Number Number 
 
 Age. Cases. Average. Cases. Average. 
 
 8 176 17.46 172 18.59 
 
 9 249 19.77 297 19.76 
 
 10 348 20.82 321 20.94 
 
 11 376 22.03 330 22.81 
 
 12 346 23.30 347 24.22 
 
 13 339 24.12 358 24.69 
 
 14 266 2.4.83 304 24.97 
 
 15 277 25.40 247 25.78 
 
 16 155 25.66 183 26.96 
 
 17 73 26.72 121 27.28 
 
 18 46 27.15 64 27.52 
 
 TABLE 6. 
 
 ABSTRACT ROTE MEMORY— CITY CHILDREN. 
 
 , Boys. ^ , Girls. ^ 
 
 Number Number 
 
 Age. Cases. Average. Cases. Average. 
 
 8 174 15.70 172 17.19 
 
 9 255 18.01 297 18.49 
 
 10 349 19.25 319 19.66 
 
 11 375 20.02 330 21.05 
 
 12 348 21.22 350 22.56 
 
 13 339 21.14 359 23.35 
 
 14 267 23.16 303 23.66 
 
 15 228 23.87 248 24.64 
 
 16 155 24.13 181 25.71 
 
 17 77 25.74 121 25.88 
 
 18 46 26.44 64 27.13 
 
116 
 
 THE PSYCHOLOGY OF LEARNING 
 
 TABLE 7. 
 
 THE RELATION OF MEMORY TO AGE AND SEX (PYLE). 
 
 MATERIAL USED, THE MARBLE STATUE. 
 
 CITY CHILDREN. 
 
 , Boys. V ( Girls. ^ 
 
 Number Number 
 
 Age. Cases. Average. Cases. Average. 
 
 8 102 24.3 89 28.5 
 
 9 148 28.7 158 21.0 
 
 10 142 30.0 138 33.5 
 
 11 149 32.9 156 36.4 
 
 12 156 35.1 191 38.1 
 
 13 163 33.8 164 38.5 
 
 14 129 36.1 146 39.0 
 
 15 89 36.5 99 39.1 
 
 16 60 34.4 94 37.3 
 
 17 45 34.6 81 36.6 
 
 18 65 38.3 86 40.1 
 
 Adults 65 38.3 86 40.0 
 
 Figure 13. Rote Memory Graphs, somewhat smoothed, ages 8 to 
 18, broken line girls, solid line boys. 
 
THE RETENTION OF EXPERIENCE 
 
 117 
 
 . 
 
 'g 9 /o // /z yJ ^'^ /•*" ''^ '/ ''* 
 
 Figure 14, Logical Memory : results from the Marble Statue test. 
 Broken line, girls ; solid line, boys. 
 
 Memory and Sex. — The fact that there is an improve- 
 ment in the various aspects of memory with age, shows 
 that retention is a function of development. Since girls 
 mature faster than boys we should expect their reten- 
 tivity to be better. Experiment reveals this to be the 
 case. In tables 5, 6 and 7 the relative memory efh- 
 
118 THE PSYCHOLOGY OF LEARNING 
 
 ciency of boys and girls is shown for the different ages 
 from eight to eighteen. The comparison is shown 
 graphically in figures 13 and 14. It can be seen from 
 the tables that girls are better in rote memory at every 
 age except nine, in concrete rote memory. Here the 
 boys excel by a hundredth of a word. While the girls, 
 with this one exception, are uniformly better, the dif- 
 ference is very small. The average yearly improvement 
 of boys from nine to eighteen is 3.94%, while that of 
 girls is 4.03%. The large gain shown for the boys in 
 concrete rote memory from eight to nine is doubt- 
 less spurious, due to the inaccuracy of measuring eight- 
 year-old boys. 
 
 In logical memory, as determined by using The Mar- 
 ble Statue test, the girls are better at every age from 
 eight to eighteen. And university women are better 
 than university men. It must be said, however, that in 
 tests of logical memory, the results depend upon the 
 kind of material used. This may be true of rote mem- 
 ory as well. In The Farmei^'s Son test used by the 
 author, boys are better from eleven to fifteen. The 
 author used three tests of logical memory, one called 
 Willie's Dog in the lower grades, one called The Far- 
 mer's Son in the upper grades, and one called Costly 
 Temper in the high school. There are records for age 
 13 in all three tests. In the Willie's Dog and Costly 
 Temper tests girls are better at age 13, while in the 
 Farmer's Son test, at the same age, boys are better. At 
 age 13, the average score for boys in the three tests 
 is 38.46, and for girls, 39.57. There are so many factors 
 effective in learning, which have their influence on re- 
 tention, that it is difficult for us to compare the reten- 
 tion of girls and boys. We can say that in any given 
 
THE RETENTION OF EXPERIENCE 119 
 
 test one sex is better than the other, but whether it is 
 better because of better retentive capacity or because 
 of difference in attitude, attention, interest, or famil- 
 iarity of the material, it is difficult to say. However, 
 in the extensive experimental work in memory, the 
 superiority of girls over boys is so general that v/e are 
 warranted in concluding that they have a better re- 
 tentive capacity. 
 
 Some experimenters have reported boys as having 
 better retention in certain fields of memory. Mulhall, 
 for example, reports boys as having better memory for 
 fonn, and cites Chamberlain as having found no sex 
 difference. But both these experimenters make their 
 comparisons by grade and not by age. Now, in the 
 same grade the boys average older than the girls. An 
 adequate sex comparison should be made by age, not 
 by grade. Several writers have stated conclusions 
 about sex differences, with this same error entering into 
 their calculations. However, even if we are careful to 
 make sex comparisons by age and -not by grade, we may 
 expect to find boys superior if they are favored by 
 either material or interest, i. e., if the experience of 
 boys makes a certain material used in the test more 
 suitable to them than to girls. It seems clear, how- 
 ever, as stated above, that if we rule out the influences 
 due to differences in experience and interest, the reten- 
 tion of girls is better than that of boys. 
 
 Girls have better retention than boys, but do women 
 excel men ? Gates, basing his conclusion on an exami- 
 nation in psychology, finds women better than men in 
 both immediate and delayed recall. It is possible that 
 the women had spent more time in study, and remem- 
 bered better because they had learned better. In the 
 
120 THE PSYCHOLOGY OF LEARNING 
 
 author's experiments, men and women were on an 
 equality as far as opportunity of learning was con- 
 cerned, for in both rote and logical memory, the mate- 
 rial was presented to both men and women in the same 
 manner, at the same time. In concrete rote memory 
 38 university men make a score of 28.5, and 61 univer- 
 sity women make a score of 28.6. With abstract 
 words, 40 men make ascore of 28.4, and 61 women 
 make a score of 27.9. There is but little difference, the 
 v/omen excelling by one-tenth of a word in the test 
 with concrete words and the men excelling by five- 
 tenths of a word in abstract rote memory. 
 
 In logical memory, with The Marble Statue as mate- 
 rial, 65 university men make a score of 38.3, and 86 
 university women make a score of 40.1. The women 
 have a superiority of 1.8 words, or 4.7%. There is 
 the possibility that this story makes a stronger appeal 
 to women than it does to men. The difference between 
 boys and girls with this same test is nearly twice as 
 great. The average score of boys from 8 to 18 is 33.3, 
 while the average score for girls is 35.9, a difference of 
 2.9 words, or 8.7%. The comparison of boys with girls 
 is based on tests of 1215 boys and 1364 girls. 
 
 Briefly, in summary : The retention of girls is better 
 than that of boys because of their more rapid develop- 
 ment. At maturity, there is no clearly demonstrated 
 sex difference not due to experience, training or atti- 
 tude. 
 
 Effect of Practice.— Can we improve memory by 
 practice? We are constantly having experience, and 
 consequently have constant practice in retention. We 
 have as much practice in retention as we do in sensa- 
 tion. Sensation does not need, in ordinary cases, sp§- 
 
THE RETENTION OF EXPERIENCE 121 
 
 cial exercise to develop it. Will special exercise in 
 memorising permanently improve the retentive capac- 
 ity of the brain? James, some thirty years ago, an- 
 swered this question as follows : "All improvement of 
 the memory lies in the line of elaborating the asso- 
 ciates of each of the several things to be remembered. 
 No amount of culture would seem capable of modifijing 
 a man's general retentivenes. This is a physiological 
 quality, given once for all with his organization, and 
 which he can never hope to change".* The great 
 amount of experimental work in memory gives us no 
 reason for modifying James' statement. The experi- 
 ments do prove beyond doubt, however, that we can 
 greatly improve our ability to memorise. In all kinds 
 of material at least some improvement comes through 
 practice, from the learning of nonsense syllables on the 
 one hand to the learning of the logical material of a 
 book on the other. But that this improvement is in 
 any sense due to an improvement of the retentiv- 
 ity of the brain it would be difficult to prove. 
 
 Winch using consonants as material, presenting them 
 visually and later in another experiment, auditorially, 
 found an improvement in ability to memorise, due to 
 practice. Bolton using digits as material, also found 
 improvement from practice. Miiller and Schumann 
 found an imiprovement in ability to memorise nonsense 
 syllables. The author found a large miprovement m 
 ability to learn the substance of a text book in sociology. 
 The experiment was conducted as follows: About a 
 page of material was read to the subject, who then re- 
 peated all the ideas he could recall. The selection was 
 then read again, and the ideas reported. This pro- 
 
 *James, Priuciplos of psychology, Vol. I, p. 003. 
 
122 THE PSYCHOLOGY OF LEARNING 
 
 eedure was continued til] the subject could report all 
 the ideas of the selection. After three months of prac- 
 tice, the subject could memorise in fifteen minutes an 
 amount of material that required an hour at the begin- 
 ning of the experiment. The improvement doubtless 
 depended upon an improvement in methods of getting 
 and organising the facts, and in an increased famil- 
 iarity with the matter discussed in the book used as 
 material. Improvement in memorising comes with so 
 httle practice and in so short a time that it is absurd 
 to believe that the retentivity of the brain could be 
 affected. 
 
 However, we must not forget that v/e can tremen- 
 dously improve our ability to fix our experience so that 
 it will be retained. We can become immensely better 
 at remembering not because we have changed the brain 
 so that it retains simple impressions better than before ; 
 through better attention, better and more repetition, 
 better and more associations, we so organise our ex- 
 perience that it is much more lasting than if got with 
 poorer methods. Poor retention is usually due to poor 
 attention, lack of repetition, poor organisation. We can 
 improve retention by improving these factors. The pos- 
 sibility of improvement in m.emorising is much greater 
 in logical material, connected ideas, than it is with dis- 
 crete material, as letters, words, nonsense material, be- 
 cause of the greater possibility of organisation on the 
 asis of meaning. 
 
 Pelation of Learning to Retention. — Our question 
 b^re simply stated is, whether facility in learning and 
 facility in retention are positively or negatively related. 
 Do quick learners retain well and poor learners retain 
 poorly, or is the reverse the truth? In general, the 
 
THE RETENTION OF EXPERIENCE 123 
 
 results of the experiments are in accord. The work of 
 Miiller and Schumann, Ogden, Henderson, Norsworthy, 
 Lyon and the author leads to one conclusion : in general, 
 the fast learner is also good in retention. This is espe- 
 cially true with logical material, such as prose and 
 poetry. The experimenters used all kinds of material, 
 nonsense syllables, words, digits, and selections in prose 
 and verse. We give some typical results. 
 
 In the author's experiments, each subject learned the 
 ideas in 21 separate segments of prose material. The 
 segments were of equal length, each containing 40 
 ideas. The procedure v/as to read the matter to the 
 subject till all the ideas could be reproduced. The num- 
 ber of readings was taken as the score. Twenty-four 
 hours later a written reproduction was required, and 
 the number of ideas correctly reproduced was taken 
 as the score. The results were as follows: 
 
 TABLE 8. 
 
 Number of Average Ideas Average 
 
 Subjects. Repetitions. Deviation. Retained. Deviation 
 
 C 4.7 2.24 37.5 2.0 
 
 F 2.9 0.78 38.5 1.7 
 
 K 5.2 1.40 34.2 4.6 
 
 J 3.6 1.90 36.7 3.2 
 
 The quickest learner is F, the slowest is K. Sub- 
 ject F requires only 55.7% as many repetitions as K, 
 but retains 11% more ideas. It will be noticed that 
 there is less difference in retention as measured by the 
 absolute amount that could be reproduced than there 
 is in learning. There is general agreement among all 
 experimenters on this point. 
 
 Norsworthy found that the students who learned the 
 greatest number of words in a German-English vocabu- 
 lary, in a given time, retained the largest percentage of 
 what had been learned. The most extensive work on 
 
124 THE PSYCHOLOGY OF LEARNING 
 
 this problem was done by Lyon. He used three dif- 
 ferent methods of testing retention. (1) The first was 
 by the absokite amount retained. (2) The second was 
 by the amount reproduced after hearing the material 
 read once again. (3) The third was by the amount of 
 time taken to relearn the material. Using method 1 
 as the criterion of retention, we find the fast learners 
 are the best in retention in all forms of mateiial used, 
 poetry, prose, nonsense syllables, words, and digits. 
 Using method 2 as our criterion, v/e find the fast learn- 
 ers the best in retention except in the case of digits. 
 But by using method 3 as our test of retention, we find 
 the fast learners the best in retention in the case of 
 logical material, prose and poetry, only. By all the 
 tests of retention, the fast learners show up the poor- 
 est in retention in the case of the digits. The greatest 
 variability in learning time was shown in the case of 
 digits. Commenting on his results, Lyon says: 'Tak- 
 ing all three methods into consideration, we are en- 
 titled to say that with material that is logical in char- 
 acter, those ivho learn quickly remember the longest. 
 With digits, however, we find the conditions, so far as 
 method 3 is concerned, reversed, for here it is the 
 quick learners who seem to forget the most. One 
 might make the inference that those who learn slowly 
 remember long, if the material used is such as involves 
 motor associations, but that they forget quickly if the 
 material is logical in character." Lyon found that, as 
 far as method 3 was concerned, nonsense syllables gave 
 the same sort of results as prose material, and that 
 words gave the same sort of results as digits. For this 
 apparently strange result, Lyon has "no satisfactory 
 explanation to offer." 
 
THE RETENTION OF EXPERIENCE 125 
 
 An experiment performed in the author's laboratory 
 confirms the results given above. The material used 
 was nonsense syllables, 25 in a series. The procedure 
 was to learn the series at one sitting, by repeating the 
 syllables to the stroke of a metronome, then relearning 
 them on the second and succeeding days until they could 
 be said from memory twenty-four hours after the last 
 learning. The number of repetitions required to leam 
 the syllables the first time is called the learning time. 
 The total number of repetitions at the different sit- 
 tings is called the total learning time. The ratio is 
 found by dividing the total learning time by the first 
 learning time. As a rule, those who are quick in first 
 learning are quick also in relearning. 
 
 TABLE 9. 
 NONSENSE SYLLABLES. 
 Learning" Time, Total Learning 
 Subject. No. Repetitions. Time. Ratio. 
 
 Wi 123 150 1.2 
 
 Th 121 129 1.08 
 
 We 118 - 147 1.2 
 
 Sw 99 136 1.4 . 
 
 Sn 72 85 1.2 
 
 B 59 70 1.2 
 
 Average 96 117 1.2 
 
 Another experiment somewhat similar gives results 
 in harmony with those given above. The author gave 
 a logical memory test to about 2000 school children in 
 grades three to eight. Whipple's Marble Statue was 
 read to the pupils and an immediate reproduction re- 
 quired. One month later a second reproduction was 
 called for. Standing in the immediate memory test was 
 correlated with standing in the retention test taken one 
 month later, the correlation being computed by grades. 
 The correlations by the Pearson formula ranged from 
 .50 to .80. 
 
126 THE PSYCHOLOGY OF LEARNING 
 
 The evidence seems conclusive that quick learning 
 and good retention are positively related. However, 
 the relation is not simple and there are many things 
 that must be taken into consideration. If slow learning 
 is due not to poor learning capacity but to caution and 
 care in learning, then slow learning will be coupled 
 with good retention. On the other hand, if the quick 
 learning is at the same time poor learning, it will be 
 coupled with poor retention. The factors that make a 
 person a good learner are the factors that will make 
 him good at retention, a high degree of attention and 
 concentration, quick and accurate apprehension, quick 
 grasp of meaning and significance, organisation of ma- 
 terial through schemes of meaning and association. 
 The reason that a quick learner often seems to remem- 
 ber poorly is because he is content with j)oor learning, 
 and especially because he does not take advantage of 
 the value that comes from repetition. In comparing 
 the retention of different individuals we should always 
 take into consideration the quality and method of 
 learning. It is worth noting that in Lyon's extensive 
 experiments, he found not a single case of a consistently 
 good learner who was consistently poor at retention. 
 There is little doubt that Lyon would have found still 
 stronger evidence of the positive relationship of learn- 
 ing and remembering if he had used a different criterion 
 of learning. Some of his subjects doubtless had the 
 material better learned than others. 
 
 Memory Material. — Memory of objects or pictures of 
 objects serially presented is better than memory for 
 the names of objects either seen or heard. Typical re- 
 sults are those of Calkins shown in the following table. 
 The delayed recall is for results three (?) days after 
 the first presentation. 
 
THE RETENTION OF EXPERIENCE 127 
 
 TABLE 10. 
 
 Words Words Pictures of 
 
 Heard. Seen. Objects. 
 
 Immediate 84.2 89.8 93.5 
 
 Delayed 34.9 48.2 74.5 
 
 Ratios Imme 1 1.066 1.110 
 
 Ratios Del 1 1.208 2.135 
 
 It will be seen that the pictures of objects are re- 
 membered much better than the names of objects either 
 seen or heard, and that in delayed memory more than 
 twice as many pictures of objects are recalled than 
 names of objects heard. Patterson compared memory 
 for objects seen with memory for words seen and words 
 heard. The results for immediate memory were : words 
 heard, 6.85; words seen, 6.92; objects seen, 8.28. The 
 ratio of objects seen to words heard is 1.209 to 1. By 
 comparison with the table above it is seen that objects 
 are remembered somewhat better than pictures of ob- 
 jects. 
 
 The educational inferences to be made from these 
 facts are obvious. Visual illustrations by means of 
 drawings or pictures will be helpful, and demonstra- 
 tions by means of the actual objects will be of the high- 
 est value. 
 
 As for other types of material: digits are remem- 
 bered better than consonants ; meaningful words better 
 than meaningless words; related words better than 
 nonrelated words; material that submits itself to 
 grouping better than that which does not ; lists of con- 
 crete words better than lists of abstract words. In the 
 author's study of the rote memories of 2654 boys and 
 2744 girls, it was found tliat the boys remembered 7.3% 
 more concrete words than abstract words, while the 
 girls remembered 5.7% more. 
 
 Manner of Presentation. — The effect on memory of 
 the form and manner of presenting the material for 
 
128 THE PSYCHOLOGY OF LEARNING 
 
 leaning has not been determined beyond doubt. We 
 shall not, therefore, go into a detailed discussion of the 
 numerous experiments. A very general statement will 
 suffice. On the one hand we find Meumann, Miinster- 
 berg and Bigham, Pohlmann, Smedley, and others 
 claiming on the basis of their experiments that visual 
 presentation is better in the case of children, especially 
 with meaningless material. On the other hand, Kem- 
 sies, Hawkins, Von Sybyl, and Henmon find auditory 
 presentation better. 
 
 There are evidently many factors entering into the 
 problem, particularly the kind of material and the age 
 of the subjects, as well as the habits and training of the 
 subjects, possibly also their ideational type. For ex- 
 ample, Pohlmann and others find auditory presentation 
 best for significant material, and visual best for mean- 
 ingless material. Henmon found the auditory presen- 
 tation best for nouns, nonsense syllables, and numbers. 
 But his subjects were six adults. He did not experi- 
 ment with children. 
 
 Several experimenters find that memory is better if 
 the material is presented to both vision and audition. 
 Others find little or no advantage by such presentation. 
 Some have reported that if the subject is allowed to 
 articulate the learning material in addition to hearing 
 and seeing it, memory is better. Others report no nH- 
 vantage. 
 
 Unfortunately, although the question of the most 
 favorable method of presenting material to different 
 ages of pupils and for different kinds of material is an 
 important one for education, it awaits the solution of 
 a future experimenter. The solution can come only 
 from a careful consideration and isolation of all thp 
 factors involved. 
 
THE RETENTION OF EXPERIENCE 129 
 
 Length of Series. — What is the relation of memory 
 to the amomit learned at one time? Strong's studies of 
 advertising showed the importance of the length of 
 series, i. e., of the number of impressions at one time. 
 He found that if only five advertisements were shown 
 at one time, the subject could recognise 86% of them 
 immediately afterward, while if 150 were shown, only 
 47% could be recognized immediately afterward. The 
 per cent, of correct recognitions decreased as the length 
 of the series increased. In Strong's experiments v/e 
 have evidence of the relation of retentiveness to the 
 number of impressions; in the earlier experiments of 
 Ebbinghaus, we have evidence of the relation of the 
 time of memorising to the amount to be memorised. 
 Ebbinghaus could memorise a series of seven syllables 
 in one repetition, but a series of tv/elve syllables re- 
 quired on the average 16.6 repetitions. Series of 16, 
 24, and 36 syllables required respectively 30, 44 and 55 
 repetitions. While 12 syllables Vv^ere learned in a little 
 over 16 repetitions, a series of 24, or twice as many, 
 required not 33 repetitions, but 44 repetitions. As the 
 length of series is lengthened beyond one's memory 
 span, the learning time is enormously increased at 
 first, then more slowly. The results from the experi- 
 ments of Ebbinghaus and Strong are shown in tables 
 11 and 12. 
 
 TABLE 11. 
 
 SHOWING THE RELATION OF THE NUMBER OF ADVERTISEMENTS 
 SEEN TO THE NUMBER THAT COULD BE RECOGNISED IMMEDI- 
 ATELY AFTERWARD. THE FIRST HORIZONTAL COLUMN SHOWS 
 THE NUMBER OF ADVERTISEMENTS SEEN, AND THE SECOND 
 ROW SHOWS THE PER CENT. THAT COULD BE RECOGNISED IM- 
 MEDIATELY AFTERWARD. (STRONG.) 
 
 No. seen 5 10 25 50 100 150 
 
 Per cent, recognised 86 85 78 67 63 47 
 
180 THE PSYCHOLOGY OF LEARNING 
 
 TABLE 12. 
 
 SHOWING THE RELATION OF THE LENGTH OF A SERIES OF NON- 
 SENSE SYLLABLES TO THE NUMBER OF REPETITIONS REQUIRED 
 TO LEARN THEM. THE UPPER HORIZONTAL COLUMN SHOWS 
 THE NUMBER OF SYLLABLES IN THE SERIES AND THE LOWER 
 COLUMN SHOV/S THE NUMBER OF REPETITIONS REQUIRED TO 
 COMMIT THEM TO MEMORY. (EBBINGHAUS.) 
 
 No. syllables 7 12 16 24 36 
 
 No. repetitions 1 16.6 30 44 55 
 
 Several other psychologists have repeated the experi- 
 ments of Ebbinghaus, and have also used other mate- 
 rial, — digits, prose, and poetry. The experimenters 
 were Binet and Henri, Meumann, Henmon, and more 
 recently, Lyon. In Table 13 are shown the results of 
 some of their work in parallel columns for comparison. 
 Lyon's results are shown in Tables 14, 15 and 16. Lyon 
 used two methods of learning. One he calls the ''con- 
 tinuous" method, and the other, the ''once-a-day" 
 method. In the former, all the learning was done at 
 one sitting. In the latter, the learning was done at the 
 rate of one reading a day. The results show the latter 
 much the better way to learn, especially in the case of 
 non-meaningful material. 
 
 A careful study of all the tables shows that the diffi- 
 culty of learning increases when the length of the ma- 
 terial is increased. How much does it increase? If 
 the increasing difficulty is in proportion to the increase 
 in length, then the number of repetitions will remain 
 constant, and the time will increase as the length. This 
 is not the case. There is a general increase in the num- 
 ber of repetitions, showing that the difficulty increases 
 more than the length increases. On the other hand, 
 the number of repetitions does not increase in propor- 
 tion to the increase in length. The increase in number 
 of repetitions is greater than the increase in length at 
 
THE RETENTION OF EXPERIENCE 131 
 
 first and then much less when the learning is done at 
 the rate of one reading a day. When the learning was 
 all done at one sitting the added length increased the 
 difficulty more than when the learning was spread out. 
 In Table 15 it is seen that 8 syllables require two repe- 
 titions. Sixteen syllables require not four repetitions, 
 but 23. On the other hand, 32 syllables require not 46 
 repetitions, but only 24. 
 
 The results in the tables show some irregularities and 
 discrepancies due to individual differences, differences 
 in material and in some cases to difference in method. 
 All the results agree in showing that increasing length 
 of material increases the difficulty of learning. They 
 disagree as to what the law of this increase is. 
 
 The educational importance of the facts shov/n by 
 these experiments is very great. From the point of 
 view of the economy of learning and memory, the num- 
 ber of impressions received at one time should be few. 
 If in a lecture, for example, too many ideas are pre- 
 
 TABLE 13. 
 
 COMPARATIVE TABLE SHOWING THE RESULTS OF SEVERAL INVES- 
 TIGATORS ON THE TIME REQUIRED TO LEARN SERIES OT^ 
 SYLLABLES OF DIFFERENT LENGTHS. 
 MATERIAL USED— NONSENSE SYLLABLES. 
 METHOD— SYLLABLES ALL LEARNED AT ONE SITTING. 
 
 Number of 
 
 Meumann. 
 
 Syllables. 
 
 Ebbinghaus. 
 
 8 
 
 
 1* 
 
 10 
 
 
 
 12 
 
 
 17 
 
 14 
 
 
 
 16 
 
 
 30 
 
 18 
 
 
 
 20 „ 
 
 
 
 24 
 
 
 44 
 
 30 
 
 
 
 32 
 
 
 
 36 
 
 
 55 
 
 48 
 
 
 
 72 
 
 
 
 ♦Ebbinghaus 
 
 use<l 
 
 7 syllables 
 
 10 
 
 17 
 21 
 
 80 
 33 
 
 nmon. 
 
 Lyon. 
 
 
 5 
 
 7 
 
 
 8 
 
 69 
 
 8 
 
 
 9 
 
 83 
 
 11 
 
 
 14 
 
 
 13 
 
 94 
 
 20 
 
 
 103 
 
 120 
 306 
 
132 
 
 THE PSYCHOLOGY OF LEARNING 
 
 sen ted to the students, the results are confusion in 
 memory and poor retention. At any one time, a teacher 
 should present to his pupils only a feio facts. These 
 should be illustrated, elaborated, related, organised, 
 
 TABLE 14 (FROM LYON). 
 
 SHOWING THE NUMBER OF READINGS REQUIRED TO MEMORISE 
 
 SELECTIONS OF VARIOUS LENGTHS. 
 
 PROSE. POETRY. 
 
 No. of No. of Repe- Time, No. of No. of Repe- Time, 
 
 Words. tltions. Min. Stanzas. titions. Min. 
 
 15 1 3/40 1 1 Vs 
 
 20 1 1/10 2 3 % 
 
 25 1 % 4 7 3 
 
 50 9 214 5 8 5 
 
 100 23 9 6 12 8 
 
 200 24 24 8 11 10 
 
 400 28 56 10 18 21 
 
 800 34 136 12 16 23 
 
 1000 33 105 18 22 46 
 
 1500 40 300 21 24 60 
 
 2000 35 350 25 20 58 
 
 3000 52 780 35 24 98 
 
 4000 38 760 40 22 103 
 
 5000 65 1625 50 25 146 
 
 7000 59 2065 75 30 262 
 
 10000. 84 4200 100 33 385 
 
 15000 73 5475 
 
 TABLE 15 (FROM LYON). 
 
 SHOWING THE RELATION OF THE LENGTH OF SERIES OF NON- 
 SENSE SYLLABLES AND OF DIGITS TO THE TIME REQUIRED TO 
 LEARN THEM, WHEN THE LEARNING WAS DONE AT THE RATE OF 
 ONE READING A DAY. THE LAST COLUMN GIVES THE NUMBER OF 
 REPETITIONS AS THY WOULD BE IF THE NUMBER INCREASED IN 
 PROPORTION TO THE NUMBER OF SYLLABLES. 
 
 No. No. Repe- No. No. Repe- 
 
 Digits. titions. Syllables. titions. 
 
 ^ 8 1 8 2 2 
 
 12 5 12 15 3 
 
 16 14 16 23 4 
 
 24 18 24 .25 6 
 
 32 25 32 24 8 
 
 48 24 48 35 12 
 
 72„ 35 72 43 18 
 
 104 41 104 43 26 
 
 200 51 200 56 50 
 
 400 70 300 78 75 
 
THE RETENTION OF EXPERIENCE 
 
 133 
 
 TABLE 16. (FROM LYON.) 
 
 SHOWING THE RELATION OF THE LENGTH OF SERIES OF NON- 
 SENSE SYLLABLES AND DIGITS TO THE TIME REQUIRED TO LEARN 
 THEltl. WHEN THE LEARNING WAS DONE ALL AT ONE SITTING. 
 
 No. Time, No. Time, 
 
 Dig-its. Min. Syllables. Min. 
 
 8 1/12 8 14 
 
 12 :... 14 12 6 
 
 16 2 16 9 
 
 24 5 24 16 
 
 32 10 32 28 
 
 48 18 48 43 
 
 72 24 72 138 
 
 104 - 56 
 
 200 154 
 
 Forgetting. — As soon as we learn we begin to forget. 
 It is well to know the rate of this forgetting, its prog- 
 ress for different materials, and the laws that seem to 
 control it. The experimental work of chief importance 
 here is that of Ebbinghaus, Radossawljewitsch, and 
 
 r^is 
 
 Figure 15. The solid liue is a logarithmic curve. The broken line 
 is a forgetting cui-ve constructed from the actual data. 
 
 Finkenbinder. The results of these three investigators 
 are shown in table 17. The differences in the results 
 from the three sources are due to differences in 
 method, — differences in method of presentation of the 
 
134 
 
 THE PSYCHOLOGY OF LEARNING 
 
 material and differences in determining when learning 
 was completed. The work of Finkenbinder was thp 
 more accurate. A study of the table and the graphs 
 shows that forgetting, in the case of nonsense material, 
 is very fast at first and then slower. The fourth col- 
 umn in the table gives the figures for a theoretical 
 logarithmic cui*ve, which the cuiTe for the nonsense 
 syllables as sho\vn in column one approximates. Such 
 a curve is shown in Fig. 15, also an actual curve con- 
 structed from the data in column 2. 
 
 TABLE 17. 
 
 AMOUNT FORGOTTEN AFTER LAPSE OF DIFFERENT INTERVALS OF 
 TIME. (AFTER FINKENBINDER.) 
 
 Finken- Radossawljevv- Ebbing- 
 
 binder. itsch. haus. 
 
 Interval 
 
 Nonsense Syllables. 
 
 Av. 
 
 of Time. 
 
 120 days 
 
 minutes 
 
 minutes 
 
 minutes 25.0 
 
 hour 27.2 
 
 hours 30.6 
 
 4 hours 33.6 
 
 8 hours 34.5 
 
 12 hours 36.2 
 
 16 hours 37.0 
 
 24 houi-s 42.2 
 
 5 
 
 20 
 
 30 
 
 1 
 
 2 
 
 hours 41.2 
 
 days.. 
 days, 
 days 
 days, 
 days- 
 days.. 
 
 14 days.. 
 
 21 days.. 
 
 30 days. 
 
 44.5 
 47.9 
 
 RE. 
 
 1.3 
 .6 
 1.8 
 1.7 
 1.4 
 1.4 
 .9 
 1.8 
 1.9 
 1.8 
 1.7 
 
 10 Log. 
 62.3 
 16.9 
 23.0 
 24.7 
 27.7 
 30.7 
 33.8 
 36.8 
 38.5 
 39.8 
 41.5 
 43.3 
 44.5 
 46.3 
 47.6 
 48.6 
 49.3 
 50.0 
 53.0 
 54.8 
 56.3 
 
 N. S. 
 Av. 
 97.2 
 2.5 
 11.4 
 
 29.3 
 
 52.6 
 
 31.1 
 39^1 
 
 50.7 
 
 59.0 
 
 62.2 
 79.8 
 
 Mean 
 
 Av. 
 
 3.9 
 
 2:L7 
 
 41.9 
 
 20. 
 
 33.2 
 43.5 
 45.5 
 43.5 
 57.6 
 50.0 
 70.0 
 52.4 
 76.1 
 
 Av 
 
 41.8 
 55^S 
 
 6472 
 
 12^ 
 
 74.6 
 
 78.9 
 
 Bean finds five factors which contribute to an ex- 
 planation of the cui"ve of forgetting. 
 
THE RETENTION OF EXPERIENCE 135 
 
 (1) The rate of forgetting depends on the degree to 
 which the material had been learned before the com- 
 mencement of the period of forgetting. Overleaming 
 makes forgetting slower. 
 
 (2) The rate of forgetting depends on the fnethod 
 of learning, i. e., on such matters as the concentration 
 and distribution of practice. Distributing learning 
 makes forgetting slower; concentrating learning makes 
 forgetting faster. 
 
 (3) The rate of forgetting varies with the kind of 
 performance in habit-foiTnation, or the kind of material 
 in ideational learning. Significant material, for exam- 
 ple, is forgotten more slowly than meaningless material. 
 
 (4) The rate of forgetting varies with the method 
 by which forgetting is measured. Bean thinks forget- 
 ting is faster if measured by the reproduction method, 
 slower if measured by the releaming method. 
 
 (5) The rate of forgetting varies with different indi- 
 viduals. 
 
 To these factors should be added two others. First, 
 the rate of forgetting varies with the amount learned 
 at one time. If a small amount is learned, forgetting 
 is relatively rapid at first. If a large amount is learned 
 at once, forgetting is slower at first. This fact is a 
 corollary of (2) above, for learning a large amount 
 makes necessary spreading out learning. If twenty- 
 four syllables are learned at one time, the learning is 
 spread out over much more time than if only twelve 
 syllables are learned. Second, fast learners, with other 
 factors equal, are slow forgetters. 
 
 The main facts of the psychology of forgetting, 
 briefly summarised, are as follows: If the material 
 learned is new material, the ratio of what is retained 
 
136 THE PSYCHOLOGY OF LEARNING 
 
 to what is forgotten varies inversely as the logarithm 
 of the time. This is the conclusion to which Ebbing- 
 haus came, and nothing that has since been discovered 
 gives any cause to modify the statement. The rate of 
 forgetting is different with different m.aterials, differ- 
 ent methods, and for different individuals, but it is 
 always very fast at first, then slov/er and slower. 
 In Ebbinghaus' experiments, more than one-third of 
 the syllables were forgotten in the first twenty min- 
 utes; more than one-half in one hour; nearly two- 
 thirds in nine hours ; more than two-thirds in 24 hours. 
 In the experiments of Radossawljewitsch with mean- 
 ingful material, one-third was forgotten within two 
 days, while one-third of meaningless material was for- 
 gotten in one day ; one-half of meaningful material was 
 forgotten in 7 days, v/hile one-half of meaningless ma- 
 terial was forgotten in 6 days. In 30 days three-fourths 
 of the meaningful material was forgotten and four- 
 fifths of the meaningless material. 
 
 Meaningful material is not forgotten as quickly as 
 meaningless material, because it is not new material, 
 and the learning of it has been going on perhaps for 
 years, in some cases, all our lives. The learning of 
 meaningful material means merely the re-arrangement 
 of what we already know\ If we undertake to commit 
 to memory Evangeline, we read: 'This is the forest 
 primeval, the murmuring pines and the hemlocks," etc. 
 Neither the v/ords nor the thought is new. This fact 
 makes the learning of meaningful material a wholly 
 different matter from the learning of meaningless ma- 
 terial. 
 
 Retention Over Long Periods. — The longest interval 
 shov^Ti in Table 17 is 120 days. We have now to enouire 
 
THE RETENTION OF EXPERIENCE 
 
 137 
 
 concerning the effect on retention of a much longer 
 interval. We shall consider only the work of Book 
 and Swift. The former tested the retention of type- 
 writing skill after a lapse of about six months and of 
 a year and six months. The original practice had cov- 
 
 li.OC 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i~06 
 
 
 
 
 '-^v'^' 
 
 
 
 
 
 Figure 1G. Typewriting Graph from Book. The graph from 
 the left up to A shows the results of the 60 days practice. At A 
 is shown "the result of the final ten-minute practices for the last 
 ten days of the experiment. B shows the efficiency reached by ten 
 days of practice six months later. C shows the efficiency reached 
 by ten days of practice a year and six months later. 
 
 end a period of 60 days, an hour a day. Practice 
 cejRpd on January 16, 1906. On June 1, a test began 
 and r/>ntinued ten minutes a day for ten days. On June 
 1, 1907, a second test of ten minutes a day for ten days 
 wa:5 begun. The results are shown in Table 18 and also 
 in Fig. 16. 
 
138 THE PSYCHOLOGY OP LEARNING ' 
 
 TABLE 18. 
 RETENTION OF TYPEWRITING SKILL (BOOK). 
 
 Last regular 
 
 parctice, 
 
 Jan. 7-16, 
 
 1906 1503 1509 1404 1572 1494 1436 1501 1455 1508 1698 1508 
 
 1st memory 
 
 test, June 
 
 1-10, 1906 1365 1421 1421 1433 1529 1443 1523 1504 1313 1472 1443 
 2nd memory 
 
 test, June 
 
 1-10, 1907 1390 1344 1345 1537 1681 1694 1634 1845 1761 1850 1611 
 
 In the regular practice there had been 50 minutes of 
 practice and a ten-minute test each day. In Fig. 16 
 are shown the learning curves for these ten-minute 
 tests for the last ten days of practice, also for the ten 
 ten-minute tests made on ten successive days about six 
 months, and a year and six months later. The striking 
 thing about this experiment is the quickness with which 
 the skill was regained. Although nearly a year and a 
 half had passed since the regular practice ceased, in ten 
 minutes daily practice for ten days the skill was re- 
 gained, and an average of 1611 words was made as 
 against an average of 1508 made on the last ten days 
 of practice. 
 
 Swift's experiments were with ball-tossing. The 
 practice was finished on Dec. 11, 1902. Memory tests 
 were made in 1903 and 1904, and again Dec. 28, 1908, 
 six years and 17 days after practice had ceased. The 
 original practice had covered 42 days. In the relearn- 
 ing test, covering 11 days, the original speed was re- 
 gained and exceeded. A score of 1600 was made as 
 against a score of 1051 in the earlier work. Swift re- 
 ports that the work was done in the last test "with 
 greater skill, greater ease." 
 
 Swift thinks that such experiments as these pix)ve 
 that learning is to some extent a grov/th, that learning 
 goes on after practice has ceased. Book, on the other 
 
THE RETENTION OF EXPERIENCE 139 
 
 hand, thinks the facts can be explained by the disap- 
 pearance of interfering associations. There is no ques- 
 tion in either experiment that forgetting had taken 
 place. When Book sat down for his first memory test, 
 he had fbrgotten the key board, but he quickly re- 
 learned ivhat he had forgotten. The fact seems to be 
 that v/hen learning is carried to a high degree of per- 
 fection, when it has required a large number of repe- 
 titions, it persists with great tenacity. It may be that 
 the associative connections begin to lose strength at 
 once, but tfiey can be easily re-established and restored 
 to their former strength. 
 
 Hill found that three years after learning mirror 
 writing, skill was nearly as good as at the start and 
 very soon regained. In three or four days of practice 
 the skill originally acquired in 48 days of practice was 
 regained. The author has also noted the peculiar per- 
 sistence of the ability to do mirror writing once it has 
 been mastered. Most people can learn to do mirror writ- 
 ing in an hour, try it no more for a year, and then find 
 themselves able to do it. 
 
 Perhaps a related fact was that found by Downey 
 and Anderson that there is "considerable retention of 
 capacity to maintain two processes (reading and writ- 
 ing) after lapse of practice for more than tv/o years, 
 with a rapid releaming and approximation of one's last 
 record." However, in both these cases, it is not so 
 much a matter of having established a number of 
 bonds, as having acquired a principle of procedure or 
 attitude, or perhaps a method, a new way of taking the 
 world. We learn this and get considerable practice in 
 an hour or two, the results of which persist for a long 
 time, but perhaps no longer than would a new idea that 
 had a similar amount of repetition. 
 
140 THE PSYCHOLOGY OF LEARNING 
 
 Thomdike holds the opinion that associative bonds 
 of the neuromuscular kind probably are better retained 
 than bonds that are foimed in ideational learning-, and 
 that the explanation is to be found in the fact that the 
 bonds connecting ideas have their basis in the later de- 
 veloped, more instable parts of the brain. Such may 
 be the case, but it is not at all sure. There are on an 
 Under^^ood typewriter 46 keys. Book practiced strik- 
 ing these 46 keys for 60 daj^s, an hour a day. For the 
 last ten days of practice this gave an average of about 
 270 strokes for each key. In the 60 days of practice, 
 each key was struck many thousands of times. In 
 Sv/ift's experiments the muscular movements involved 
 in keeping two balls in the air were few. These muscu- 
 lar movements were repeated many thousands of times 
 in the 42 days of practice. Owing to the immens<3 
 amount of repetition in both of these experiments, it 
 is no wonder that the bonds persisted. If we should 
 practice as much in the case of ideational learning, who 
 knows but that the associative bonds would be fully as 
 tenacious? Suppose we should take a poem involving; 
 46 words or 46 ideas and practice saying it for 60 days, 
 and then after a year and a half releam it, the results 
 might be much similar to those of the experiments 
 mentioned above. The reason why we seem to forget, 
 in the case of ideas, faster than is the case in neuro- 
 muscular learning, is more probably because in the case 
 of ideational learning practice in fixing bonds is insig- 
 nificant in amount when compared to practice in the 
 case of motor learning. We study a lesson once, or at 
 most, a few times. We hear a lecture once. Perhaps 
 the facts of the lecture are reviewed ? time or two. We 
 
THE RETENTION OF EXPERIENCE 141 
 
 could not expect the fixation in such learning to com- 
 pare favorably v/ith motor learning involving hundreds, 
 perhaps thousands of repetitions. 
 
 The facts available do not enable us to settle the 
 matter of the relative persistence of ideational learn- 
 ing as compared to motor learning. But the facts do 
 enable us to say that the strength of ideational bonds 
 could be much increased by more repetition. 
 
 The results of motor learning suggest that great im- 
 provement is possible in ideational learning. This im- 
 provement is in the direction of association, organiza- 
 tion, repetition. Some people seem to have the idea 
 that if a thing is well learned once, it is learned for all 
 time. Such is not the case. The only way known to 
 fix and perpetuate a neural bond of any kind is by repe- 
 tition. It makes no difference whether the bond is one 
 connecting ideas, or stimulus and response, the facts 
 are the same. Poems and songs that are learned in 
 childhood through countless repetitions persist through 
 life, or at least can be quickly and readily relearned. 
 The tables which we learn in childhood by countless 
 repetitions also persist in our minds. In all cases of 
 ideational learning involving a great num!)er of repeti- 
 tions, retention is good over long periods of time. 
 
 Relation of Memory to Intelligence. — The relation of 
 memory to intelligence has been determined in various 
 ways: (1) Many investigators have compared pupils* 
 standing in logical merr>ory tests with their standing in 
 school studies. The relation is found to be positive, 
 those having good retention standing well in their 
 studies, and those having poor retention standing low 
 in their studies. The relation is not absolute, of course, 
 
3 42 THE PSYCHOLOGY OF LEARNING 
 
 but high and positive. The author has found the cor- 
 relation to be as high as .76 in the upper grades. 
 
 (2) Standing in logical memory tests has been cor- 
 related v/ith other mental tests. The author has ob- 
 tained the following correlations : 
 
 TABLE 19. 
 
 CORRELATION OF LOGICAL MEMORY TEST WITH SIX OTHER 
 
 MENTAL TESTS. 
 
 With r. P. E. 
 
 Rote memory .44 .049 
 
 Substitution 26 .026 
 
 Opposites 77 .033 
 
 Free Association 41 .069 
 
 Word building 53 .059 
 
 Completion 77 .034 
 
 Average of six 64 .049 
 
 It is evident from the table of correlations that log- 
 ical memory has a high positive relation to other aspects 
 of intelligence, giving the highest correlation with 
 completion, the best single measure of intelligence so 
 far devised. 
 
 (3) If all the children of the same age in any school 
 system are compared with reference to their logical 
 memory, those having the better memory are found in 
 the higher grades as shown in Fig. 17. 
 
 All psychologists v/ho have seriously investigated the 
 relation of memory to intelligence have found the facts 
 as stated above. For example, Lyon, after years of 
 experimental study of memory, says:i "The students 
 who stand highest in their various studies, and who 
 prove upon examination to be the most intelligent, have, 
 as a rule, the best memories. They not only learn more 
 quickly, but they retain better." Winch in his exten- 
 sive studies of the memories of London school children, 
 came to the same conclusion, as have Bolton, Bourdon, 
 
THE RETENTION OF EXPERIENCE 
 
 143 
 
 Ji 
 
 
 
 
 
 ' 
 
 34 
 
 
 
 /y/7 
 
 
 / 
 
 // 
 
 
 
 
 
 ""^ 
 
 SC 
 
 
 
 
 
 
 28 
 
 
 . 
 
 
 
 
 Zi 
 
 ^ 
 
 ^ 
 
 
 
 
 t4 
 
 
 
 GjeADi- 
 
 
 
 i 
 
 T 
 
 jT 
 
 
 j&- 
 
 :^ir 
 
 Figure 17. The graph shows the logical memory efficiency of twelve 
 year old children in the fourth, fifth, sixth, and seventh grades. 
 
144 THE PSYCHOLOGY OF LEARNING 
 
 Pohlmann, and many others. It is interesting to note 
 here that Winch also found a positive correlation be- 
 tween ability to memorise stories and the ability to in- 
 vent stories, the correlation being higher with the more 
 intelligent. 
 
 When we say memory has a high relation to intelli- 
 gence, we mean particularly logical memory. People 
 of low intelligence may have good rote memories, may 
 remember well discrete material, but only the intelli- 
 gent have good memory for logical material. The 
 unintelligent have no comprehension of significance. 
 Such comprehension is necessary in logical memory. 
 Some teachers have the mistaken notion that a good 
 memory is to be deplored, that children should under- 
 stand rather than menfiorise, that children should 
 reason things out. But one cannot reason unless one 
 remembers the facts of experience. Remembered ideas 
 are the raw material of thought. Other things equal, 
 the one who remembers best can reason best. Of 
 course, remembering things without knowing their sig- 
 nificance is of little value, but we must remember if 
 we are to know their significance. 
 
 Individual Differences in Retentiveness.— Individual 
 differences in the various aspects of memory are very 
 great. The author once determined the logical memory 
 of 100 high school students and found the best memory 
 to be four times as good as the poorest. The distribu- 
 tion of 1032 university students is shov/n in Fig. 18. 
 The test on which the graphs of the figure are based is 
 the Marble Statue. The records, as represented on the 
 horizontal axis of the figure, are for the number of 
 ideas correctly reproduced after hearing the story read. 
 
THE RETENTION OF EXPERIENCE 
 
 145 
 
 The number of students represented is 516 men and the 
 same number of women. 
 
 The range of ability in rote memory is not quite so 
 great as shown in Fig. 19. The graphs are based on 
 measures of 419 university men and an equal number 
 of women. 
 
 The differences in retentive capacity are thus seen 
 to be very great. It is of very great consequence to a 
 teacher to know that in an ordinary class retentive 
 
 
 
 T"\ 
 
 
 
 
 rii^ /s 
 
 '■■'' / 
 
 V 
 
 \ 
 
 \ 
 
 
 / 
 
 / 
 
 
 \ 
 
 \ 
 
 
 // 
 
 / 
 
 
 
 \\ 
 
 
 ^^■^ 
 
 
 
 
 \/K 
 
 N,. 
 
 Figure 18. Frequency Surfaces showiug the distribution of 516 
 university men and the same number of women in logical memory 
 efficiency ; solid line women, broken line men. 
 
 capacity may be two or three or even four times as 
 good in the best pupil as it is in the poorest pupil. 
 
 Different Aspects of Memory.— We have treated re- 
 tention as if it were a definite characteristic of the 
 nervous system. It probably is, but when we under- 
 
146 THE PSYCHOLOGY OF LEARNING 
 
 A 
 
 r.f. 
 
 Figure 19. Frequency Surfaces showing the distribution of 419 
 university men and the same number of women in rote memory 
 efficiency. 
 
THE RETENTION OF EXPERIENCE 147 
 
 take to measure it, we have to use some paiticular kind 
 of impression, some particular stimulus, some definite 
 kind of material. It turns out that we get different 
 results from different kinds of material. We can not 
 therefore treat memory as if it were a definite, unitary 
 function. We must always speak of memory for this 
 or that kind of material, presented in such and such 
 manner, and tested in such and such way. It is prob- 
 ably true that our nervous systems have very definite 
 capacities for retention, different for different people, 
 but our various interests, and attentions, and preju- 
 dices, our various experiences with their different con- 
 sequences, bring it about that different things make 
 very different appeals to us. As a result, our retention 
 of one type of impression may be very different from 
 our retention of another. If a group of people are 
 measured to ascertain their retention of various types 
 of material, such as numbers, letters, objects, words, 
 sentences, or the ideas of some selection of prose or 
 poetry, it is found that their relative standing is not 
 the same. Indeed, their retention of several different 
 selections of meaningful material will vary. The author 
 has tv/ice computed the intercorrelation betv/een the 
 various pairs of measures for four different logical 
 memory tests. In the first experiment, the intercor- 
 relations were .379, .524, .555, .556, .568, .590. The 
 average of these correlations is .5287, and the average 
 P. E. is .089. In the second experiment, the correla- 
 tions were .545, .559, .550, .518, .617, .476. The average 
 is .5441, and the average P. E. is .074. For raw cor- 
 relations in any sort of mental tests, these are high, and 
 we may conclude that the relative ranks of individuals 
 in logical memory tests will be about the same unless 
 
148 THE PSYCHOLOGY OF LEARNING 
 
 some are specifically favored by the type of material. 
 Thomdike found the correlation between memory for 
 words and m.emory for numbers to be .45, and the cor- 
 relation between immediate memory for words and de- 
 layed meniory — 24 hours later — ^to be .55. 
 
 So called rote memory is of considerable importance 
 in early education, and of some importance throughout 
 life. Rote memory has a positive correlation with 
 logical memory but it is not very high. The average 
 raw correlation is .355. The true correlation being 
 about .53. 
 
 There are two definitely different aspects of mem- 
 ory. One v/e may call retention; the other, organisa- 
 tion. By retention, we mean the after-effects of ner- 
 vous excitation, the persistence in the brain of the 
 effects of neural activity. By organisation, we mean 
 the associative connections between the separate im- 
 pressions. These associative connections are the basis 
 of recall. For the practical pui-poses of life, not only 
 must impressions be retained, but v/e must be able to 
 recall them. The significant thing for life is therefore 
 logical memory, and it depends on more than mere 
 retention. It depends on organisation, on meaning. Our 
 ability to organise our experience in helpful ways de- 
 pends on our capa^city to see the significance of experi- 
 ence, and this is something entirely apart from reten- 
 tion. Indeed, our simple retention may be good v/hile 
 we entirely lack the power of organisation. Organisa- 
 tion can not exist without retention, but we can have a 
 high degree of retentive capacity and lack the power 
 of organization. A low type of mind may have an ex- 
 perience as rich as that of a Newton, and retain that 
 experience as well, but lacking the power of organisa- 
 
THE RETENTION OF EXPERIENCE 149 
 
 tion, the person uses this experience to no advantage 
 because there are no helpful, meaningful associations 
 in it. 
 
 Immediate and Delayed Memory. — The problem here 
 is virtually the same as that discussed above under the 
 head of relation of learning to retention, for the imme- 
 diate memory span is taken as the measure of quickness 
 of learning. We shall cite here no further evidence of 
 the relationship, but merely recall that it was said that 
 the quick learner is in general good in retention also. 
 We may here put the matter in a somewhat different 
 way. Suppose we present a given amount of matter 
 to a group of people and find what amount of the matter 
 they can immediately reproduce, then a day later, or a 
 month later, we ascertain hov/ much of the matter is 
 still retained, v/e find in general that those who had 
 the most facts immediately after the presentation still 
 have the most facts. In any particular experiment, the 
 relation will not be found perfect and absolute, because 
 it is impossible to keep the conditions equal. The mat- 
 ter will come back to the minds of some and be repeated 
 more than will be the case v/ith others. Their retention 
 Vv^ill consequently be better relatively. 
 
 Measurements of Retention. — ^There are three main 
 miothods of measuring retention. (1) The reproduction 
 Tfiethod. By this method, we determine the amount of 
 material previously learned that can be recalled. This 
 is the most widely used method of measuring retention, 
 and it is the test which the practical affairs of life put 
 upon us, for no matter how easily we could relearn the 
 material, no matter how much of it is just below the 
 conscious level, if we can not recall it, it is not imme- 
 diately available, and does us no more good for the 
 
150 THE PSYCHOLOGY OF LEARNING 
 
 moment of need than if it were in no sense retained. 
 (2) The relearning method. By this method we really 
 measure the amount forgotten by the time required to 
 relearn. The procedure is to learn a certain material 
 to the point of just being able to reproduce it, and then 
 later determine how^ much time is required to relearn 
 the amount forgotten. The relearning time is taken 
 as the measure of the amount forgotten and is inverse- 
 ly related to the amount retained. A relatively short 
 relearning time indicates a good memory. This is prob- 
 ably a better method of measuring absolute retention 
 than is the reproduction method, for it gives weight to 
 any retention, no matter hov/ vague or how faint. The 
 reproduction method emphasises organisation, for re- 
 call depends on organisation and, as said above, for the 
 practical purposes of life is perhaps more useful than 
 the relearning method. (3) The recognition method 
 By this method we do not have the subject either recall 
 or relearn but recognise a previously experienced stimu- 
 lus. The method may be illustrated by Strong's adver- 
 tisement experiments. Strong presented to his subjects 
 a certain number of advertisements and later pre- 
 sented these same advertisements along with many 
 others. The subjects were to say whether each adver- 
 tisement had been seen before or not. It is evident that 
 the one of these methods w^e should use in any given 
 case would depend upon our purposes. 
 
 Usually we are most interested in the logical mem- 
 ory of our students. This can be accurately measured 
 by the form of tests now in use for this purpose. The 
 procedure is to divide a story into ideas or units. The 
 story is read to the subjects and, for immediate mem- 
 ory, we then require a written reproduction of the 
 
THE RETENTION OF EXPERIENCE 151 
 
 story. In scoring it, we give credit for each idea ade- 
 quately reproduced- Another method is to have the 
 subjects answer a number of questions concerning the 
 story. This procedure makes the grading of the work 
 easier and more accurate. The questions can be so 
 constructed that they can be answered in each case 
 by a single word. 
 
 The best standardised method of measuring rote 
 memory is as follows: Prepare six lists of words of 
 one syllable each. Put three words in the first list, four 
 in the second, five in the third, six in the fourth, seven 
 in the fifth, and eight in the sixth. The first list is 
 read to the subjects and an immediate reproduction 
 required. The second list is then read and an immediate 
 reproduction required, and so on to the eighth list. We 
 can score the reproduction by giving one credit for 
 each word reproduced without regard to position in 
 the list, or we can give credit for the v/ord and one 
 credit for its position. The results are about the same, 
 and of course, the former method is much simpler and 
 easier. 
 
 Recognition. — A few facts have been discovered in 
 the study of recognition memory that should be men- 
 tioned. Myers found recognition memory to be about 
 two-and-a-half times as efficient as recall, and the cor- 
 relation between the two to be low. For boys, recogni- 
 tion was three times as efficient as recall and for girls 
 recognition was only twice as efficient. But girls were 
 more efficient in recognition than v/ere boys. 
 
 Strong studied the effect of the time interval upon 
 recognition. The procedure was to present 20 words 
 and then later present the same 20 along with 40 others. 
 He tried different intervals up to 42 days. In immediate 
 
152 THE PSYCHOLOGY OF LEARNING 
 
 recognition, 84% were recognised. After 7 days, only 
 10% were recognised. The decrease was rapid at first, 
 then slow, as it has been found to be in other tests of 
 memory. 
 
 Voluntary Recall. — Until matter is well learned, at- 
 tempt at voluntary recall is a hindrance and a waste of 
 time, but after matter is learned, voluntary recall has 
 great value just as re-presentation has great value in 
 fixing the material. Myers has studied the former of 
 these factors. To 332 subjects he gave four series of 
 tests, presenting orally ten words in each series. He 
 found a decided gain in final recall as a result of inter- 
 vening recall without re-presentation of the stimuli 
 The effects of recall were greater if recall took place 
 five minutes after presentation than when recall was 
 immediate. This is in harmony with Jost's law — ^that 
 repetitions have greater value for the older associa- 
 tions. 
 
 Guillet showed the value of presenting matter again 
 two to five days after first presentation. Both recall 
 and re-presentation are valuable. Which we should use 
 at any particular stage of learning depends upon the 
 kind of material and how far learning has progressed. 
 
 As long as learning is uncertain and doubtful, re-pre- 
 sentation is preferable. When recall is easily possible, 
 recall is preferable. The importance for permanent 
 fixation of re-presentation and recall can not be over- 
 estimated. In the practical work of studying our les- 
 sons, we should learn the material of the lesson; then 
 after an interval of some hours go over the matter 
 again; then after another interval go over the matter 
 in the process of recall; then with longer and longer 
 
THE RETENTION OF EXPERIENCE 153 
 
 intervals go over the matter again and again, using re- 
 presentation or recall as the conditions demand. 
 
 In poor learning, as Myers has pointed out, the 
 wrong elements become coupled together in recall. We 
 should not try to recall when this is likely to happen. 
 The remedy for poor learning is re-presentation, drill, 
 better organisation. 
 
 Position in Series. — Numerous experiments have 
 shown that in committing to memory a series of words, 
 letters, or other items, the first ones and the last ones 
 are learned first. Dell considered the matter with ref- 
 erence to meaningful material. He found that, other 
 things being equal, the beginning and end of material 
 v/ere learned first. Generally speaking, other things are 
 not equal; other factors are usually more important 
 than position in series. Logical and systematic ar- 
 rangement are probably more important. If we can 
 combine logical and systematic arrangement with put- 
 ting-important items at the beginning and end, the re- 
 sults will be favorable. 
 
 Cramming. — Cramming may be defined as the learn- 
 ing or trying to learn a great amount of material in a 
 short time. It is a legitimate process in two definite 
 situations. If one has to organize a large am^ount of 
 material for a particular occasion, it is legitimate and 
 economical to concentrate the learning and do it all just 
 before the knowledge is needed. Secondly, in the learn- 
 ing of material that is to be organised and remembered 
 for life, or for a long period of time, the economical 
 procedure is to spread the learning out, and occasionally 
 to go over in concentrated fashion the whole of the 
 material, for example, go over a Vv^hole book in one eve- 
 ning. This concentrated consideration of a large amount 
 
154 THE PSYCHOLOGY OF LEARNING 
 
 of material helps one to see each part in the light of 
 other parts. By the very process of being considered 
 together, facts become related in memory. The oftener 
 we can think over bits of experience that belong to- 
 gether, and that are needed together for use, the better 
 these bits become cemented in memory, and the more 
 likely one idea will be to recall the other related ones. 
 The cramming process is not effective for lasting mem- 
 ory, however, unless there has previously been learning 
 of the parts in detail. A bird's eye view of a lot of ma- 
 terial is profitable provided we have previously had a 
 microscopic view. 
 
 EXPERIMENTS AND EXERCISES. 
 
 1. The logical mem.ory of all the students in the class 
 should be determined by giving at least four tests, on 
 the order of Whipple's Marble Statue test. The in- 
 structor should select four short passages of prose that 
 are likely to make a strong appeal to all members of 
 the class. Divide the passage up into its ideational units 
 for ease in scoring. Read each passage and give plenty 
 of time for each student to reproduce it. The score 
 should be the number of ideas adequately reproduced. 
 Combine the scores of the four tests to obtain each stu- 
 dent's logical memory efficiency. Compute the inter- 
 correlations, six in all, to determine the co-efficient of 
 reliability of the tests. 
 
 2. Determine each student's rote memory for words 
 by using lists of one syllable words, the lists ranging 
 in length from five words to eight. Make at least four 
 such tests. Determine the co-efficient of reliability as 
 in experiment 1 above. Determine each student's 
 standing by combining the results of the four tests, and 
 
THE RETENTION OF EXPERIENCE 155 
 
 correlate the results with those from the logical mem- 
 ory tests. 
 
 3. Make a study of memory for objects in the follow- 
 ing way: Select four groups of objects, ten in each 
 group. Conceal them from the class. Expose the ob- 
 jects of each group serially, one at a time, and require 
 the students to write down the names of the objects, 
 without regard to the order of presentation. Combine 
 the four records for the total score. Determine the co- 
 efficient of reliability as above, and correlate the com- 
 bined scores with those in the two tests above. 
 
 REFERENCES. 
 
 Allin, A. M., Recognition, A. J. P., VII, 1896, 249. 
 
 Angell, J. R., Meaning in memory and in attention, P. R., X, 
 1903, 267. 
 
 Baldwin, J. M. and Shaw, W. J., Memory for square size, P. R., 
 II, 1895, 236. 
 
 Bean, C. H., The curve of forgetting. Arch, of Psych., XX, No. 3, 
 1912. 
 
 Bennett, F., Tlie correlations between different memories, 
 J. Exp. P., I, 1916, 404. 
 
 Bergstrom, J. A., Experiments upon physiological memory J)y 
 means of the interference of associations, A. J. P., V, 1892, 356. 
 Effect of changes in the time variable in memorising, to- 
 gether with some discussion of the technique of memory experi- 
 mentation, A. J. P., XVIII, 1907, 206. 
 
 Burnham, W. H., Retroactive amnesia, illustrative cases and a 
 tentative explanation, A. J. P., XIV, 1903, 382. 
 
 Memory, historically and experimentally considered, A. J. P., 
 II, 1888, 39. See also, II, 568. 
 
 Bigham. J., Memory, P. R., I, 1894, 453. 
 
 Bolton, T. L., The growth of memory in school children, A. J. P., 
 IV, 1891, 362. 
 
 Bolton, F. E., The accuracy of recollection and odservaMnn. 
 P. R., Ill, 1896, 286. 
 
 Calkins, M. W., Short studies in memory and in association^ 
 P. R. v., 1898, 451. 
 
 Chamberlain, A. H., A memory test with school children PR. 
 XXII, 1915, 71. 
 
 Collgrove, P. W., Individual memories, A. J. P., X, 1899, 228. 
 
 Cowan, E. A., An experiment on the influence of trainina on 
 memory, J. E. P., VII, 1916, 31. 
 
156 THE PSYCHOLOGY OF LEARNING 
 
 Dearborn, G. V. N., Recognition under objective reversal, P. R., 
 VI, 1899, 395. 
 
 Dell, J. A., Som-e observations on the learning of sensible 
 material, J. E. P., Ill, 1912, 401. 
 
 Downey, J. E. and Anderson, J. E., Retention of skill after 
 lapse of practice, A. J. P., XXVIII, 1917, 396. 
 
 Ebbinghaus, H., Memory, 1885, Tr. by H. A. Ruger, 1913. 
 
 Finkenbinder, E. 0., The curve of forgetting, A. J. P., XXIV, 
 1913, 8. 
 
 Gates, A. I., Experiments on the relative efficiency of men and 
 women in memory and reasoning, P. R., XXIV, 1917, 139. 
 
 Recitation as a factor in memorising. Arch, of Psych., No. 
 40, 1917. 
 
 Correlations of immediate and delayed recall, J. E. P., IX, 
 1918, 489. 
 
 The mnemonic span for visual and auditory digits, J. Exp. P., 
 I, 1916, 393. 
 
 Guillet, C, A study of the memory of young loomen, J. E .P., 
 VIII, 1917, 65. 
 
 Retentiveness in child and adult, A. J. P.. XX, 1909, 318. 
 
 Hall, G. S., Note on early memories, Ped. Sem., VI, 1899, 485. 
 
 Hawkins, C. J., Experiments on memory types, P. R., IV, 1897, 
 289. 
 
 Hayden, E. A., memory for lifted weights, A. J. P., XVII, 
 1906, 497. 
 
 Henderson, E. N., A study of memory, P. R. Mon. Sup., No. 23, 
 1903. 
 
 Henmon, V. A. C., The relation between mode of presentation 
 and retention, P. R., XIX, 1912, 79. 
 
 The relation betiveen learning and retention and the amount 
 to be learned, J. Exp. P.. II, 1917, 476. 
 
 Hill, D. S., Minor studies in learning and relearning, J. E. P., 
 V, 1914, 375. 
 
 HoUingworth, H. L., Characteristic differences between recall 
 and recognition, A. J. P., XXIV, 1913, 532. 
 
 Kennedy, F., On the experimental investigation of memory, 
 P.R.V., 1898, 477. 
 
 Kirkpatrick, E. A., An experimental study of memory, P. R., 
 I, 1894, 602. 
 
 Kuhlmann, F., The place of mental imagery and memory among 
 mental functions, A. J. P., XVI, 1905, 337. 
 
 Lyon. D. O., The relation of quickness of learning to retentive- 
 ness. Arch, of Psych., No. 34, 1916. 
 
 Mulhall, E. F., Tests of the memories of school children, 
 J. E. P., VIII, 1917, 294. 
 Miinsterberg, H. and Blgham, J., Memory, P. R., I, 1894. 34. 
 
 Myers, G. C., A comparative study of recognition and recall, 
 P. R., XXI, 1914, 442. 
 
 Recall in relation to retention, J. E. P., V, 1914, 119. 
 Confusion in recall, J. E. P., VIII, 1917, 166. 
 
 Norsworthy, N., Acquisition as related to retention, J. E. P., 
 Ill, 1912, 214. 
 
THE RETENTION OF EXPERIENCE 157 
 
 Patterson, T. L., Pedagogical suggestions from memory tests, 
 J. E. P., IX, 1918, 497. 
 
 Perkins, N. L., The Value of distributed repetitions in rote 
 learning, B. J. P., VII, 1914-15, 253. 
 
 Peterson, H. A., On the influence of complexity ana dissimi- 
 larity on memory, P. R. Mon. Sup., No. 49, 1909. 
 
 McC. Gamble, E. A., A study in memorising various materials 
 hy the reconstruction method. P. R. Mon. Sup., No. 43, 1909. 
 
 Pohlmann, A., Experimentelle Beitrdge ziir Lehre vom Oeddcht- 
 nis, 1906. 
 
 Potwin, E. B., Study of early memories, P. R., VIII, 1901, 596. 
 
 Pyle, W. H., One function of the teacher in memory work, 
 J. E. P., I, 1910, 474. 
 
 The most economical unit for committing to memory, J. E. P., 
 II, 1911, 133. 
 
 Retention as related to repetition, J. E. P., II, 1911, 311. 
 
 Shaw, J. C, A test of memory in school children. Fed. Sem., IV, 
 1896, 61. 
 
 Smith, W. G., The Place of repetition in memory, P. R., Ill, 
 1896, 21. 
 
 Smith, M. and McDougal, W., Some experiments in learning 
 and retention, B. J. P., X, 1920, 199. 
 Smith, T. L., On muscular memory, A. J. P., VII, 1896, 453. 
 
 Stetson, G. R., Some memory tests of whites and blacks, P. R., 
 IV, 1897, 285. 
 
 Strong, E. K., Jr., The effect of time interval upon recognition 
 memory, P. R., XX, 1913, 339. 
 
 The effect of length of series upon recognition memory, P. R., 
 XIX, 1912, 447. 
 
 Swift, E. J., Memory of a complex skillful act, A. J. P., XVI, 
 1905, 131. 
 Thorndike, E. L., Memory for paired associates, P. R., XV, 1908, 
 122. 
 
 The relation between memory for ivords and memory for 
 numbers and the relation between memory over short and mem- 
 ory over long intervals, A. J. P., XXI, 1910, 487. 
 
 Titchener, E. B. and Talbot, E. B., An attempt to train the 
 visual memory, A. J. P., VIII, 1897, 414. 
 
 Tolman, E. C, Retroactive inhibition as affected by conditions 
 of learning, P. Mon., No. 107, 1917. 
 
 Watkins, S. H., Immediate memory and its evaluation, B. J. P., 
 VII, 1914-15, 319. 
 
 Whitehead, L. G.. A Study of visual and aural memory pro- 
 cesses, P. R., Ill, 1896, 258. 
 
 Winch, W. H., Some relation between substance memory and 
 productive imagination in school children, B, J, P., IV., 1911, 95. 
 Immediate memory in school children, B. J. P., II, 1906, 52. 
 Immediate memory in school children, B. J. P., I, 1904, 127. 
 
Chapter VIII. 
 THE NATURE OF LEARNING CAPACITY. 
 
 Our problem is now to enquire into the nature of 
 learning capacity. The question which we shall try 
 to answer is this: Are the characteristics of the ner- 
 vous system such that a person may be spoken of as a 
 "good learner" or a "poor learner" in general? Or 
 must we say that a person is good at learning one type 
 of material and perhaps poor at learning other types? 
 Must we say that a certain person can learn mathe- 
 matics easily, but learns some other subject, let us say 
 history, with difficulty? Or can we say that a person 
 is a good learner or a poor learner at whatever he un- 
 dertakes to learn? In the schools, for example, is a 
 person known as a good student or a poor student in 
 general, or is he known as a good student in history and 
 language, for example, and a poor student in science 
 and mathematics ? Generally students are spoken of as 
 good or poor without reference to any particular sub- 
 ject of study. There are, however, numerous excep- 
 tions. Sometimes a student is known as an excellent 
 student in science and as a poor student in language. In 
 such cases the question arises whether the explanation 
 may not be quite other than difference in ability to 
 learn the different subjects. The explanation may be 
 found in difference in preliminary training, difference 
 in interest, and other factors of experience. 
 
 [158] 
 
THE NATURE OF LEARNING CAPACITY 159 
 
 The question which we have raised in this chapter is 
 of the highest importance to education, for if learn- 
 ing, whatever its nature, depends upon fundamental 
 characteristics of the brain and nervous system, then 
 it will be found a comparatively simple matter to de- 
 termine by experiment the learning capacity of the 
 pupils in the public schools and of students in the col- 
 leges. On the other hand, if learning is a specific mat- 
 ter, we can not speak of learning capacity in a general 
 sense, but must speak of ability to learn this or that. 
 But let us turn to the experiments. 
 
 Interrelation of Different Learning Capacities. — We 
 shall first consider extensive experiments performed in 
 the author's laboratory with several different kinds of 
 material. The learning falls under four different types. 
 (1) First, what we may call motor learning, consisting 
 of card sorting and marble sorting. The card sorting 
 consisted in distributing 150 numbered cards into trays 
 numbered correspondingly. There were 30 trays ar- 
 ranged with five in a row, and in six rows. There 
 were therefore five cards to each tray. An experiment 
 consisted in distributing the 150 cards into their appro- 
 priate trays. In marble sorting 96 marbles of different 
 colors and sizes were distributed into six different 
 receiving boxes by means of a complicated machine 
 demanding the use of both feet and both hands. (2) 
 The second type of learning is what we may call 
 semi-motor in character, the substitution experiments. 
 There were three types of these experiments, turning 
 digits into symbols, turning symbols into digits, and 
 transcribing the alphabet into an arbitrary symbol 
 alphabet. The former two we shall call the digit sub- 
 stitution experiment and the latter the alphabet substi- 
 
160 THE PSYCHOLOGY OF LEARNING 
 
 tution experiment. The essential difference in these 
 experiments was that in the digit substitution experi- 
 ments, there were only nine bonds to form, while in the 
 alphabet substitution experiment there were twenty-six 
 bonds to form. In the substitution experiments 
 while there was a motor element, it was not so im- 
 portant as in the sorting experiments, and the idea- 
 tional element was of more importance. (3) The 
 third type of experiment we may call ideational, 
 because the motor element was relatively unimportant. 
 In this experiment series of ten nonsense syllables were 
 learned. They were exposed visually in series of ten, 
 one second exposure to each syllable. The exposure of 
 the syllables was continued until the subjects could 
 write them in their proper order. (4) The fourth type 
 of learning we may call motor-inhibited. It consisted in 
 learning mirror writing. The method consisted in learn- 
 ing to write, with a mirror-reflected stimulus, letters 
 that appeared normal in the mirror. It will be seen 
 that these experiments were widely different in type. 
 Even the two we have called motor were widely differ- 
 ent. In the case of. card-sorting, simple associations 
 had to be built up between numbers and places. There 
 were 30 such associations. In marble-sorting, the 
 matter was much more complicated. Appropriate re- 
 sponses had to be learned for marbles of four different 
 colors and two different sizes, only six in all, but there 
 were four different holes for the reception of the mar- 
 bles, and the feet had to manipulate pedals when cer- 
 tain marbles were deposited. Most of the experiments 
 were repeated with several different groups of stu- 
 dents. There were usually about 75 in a group. The 
 
THE NATURE OF LEARNING CAPACITY 161 
 
 learning capacity of each student for each type of 
 learning was determined, and all the possible correla- 
 tions were computed, 41 in all. The results were very 
 constant- Only the averages will be given here. In 
 table 20 are shown the average correlations of each 
 type of learning with all other types. 
 
 TABLE 20. 
 
 SHOWING THE AVERAGE CORRELATION OF EACH OF FIVE TYPES 
 OF LEARNING WITH OTHER TYPES. 
 
 Type of Average of 
 
 Learning. Correlations. 
 
 Digit substitution 594 
 
 Alphabet substitution 547 
 
 Nonsense syllables 441 
 
 Card sorting 496 
 
 Marble sorting 503 
 
 Average of all 516 
 
 Only one correlation could be computed with mirror 
 writing, and that was with learning nonsense syllables. 
 Between these two tests, the correlation was .505, and 
 is included in the averages as given for nonsense syl- 
 lables. 
 
 What is the significance of these correlations? A 
 raw correlation of a little more than .5 indicates a very 
 high positive relationship between the abilities required 
 to learn the different types of material. But if learn- 
 ing capacity is general, should not the correlation be 
 unity? In the author's opinion, if the disturbing fac- 
 tors could be eliminated, the correlation would be unity. 
 There were several disturbing factors. In the first 
 place, the score obtained was not always a correct indi- 
 cation of quickness of learning. In card-sorting there 
 were wide differences in ability to manipulate the cards 
 which affected the score but had no relation to ability 
 to learn. These differences in facility with the cards 
 
162 THE PSYCHOLOGY OF LEARNING 
 
 were due to past experience. Another factor which 
 affected the score but had no relation to capacity to 
 learn was attitude toward the different experiments. 
 Owing to difference in interest, students would try much 
 harder in some experiments than in others. Previous 
 experience, then, affected the scores through giving 
 the students some advantage or disadvantage and 
 through their effects on the students' attitude and in- 
 terest. In motor learning, reaction time, which prob- 
 ably has no relation to ability to learn, affected the 
 scores. We conclude that as far as these learning 
 experiments furnish evidence, learning capacity is a 
 general characteristic. 
 
 Evidence From Memory Experiments. — Experiments 
 In immediate memory are virtually experiments in 
 ability to learn, and their results throw light on our 
 present problem. In Table 21 are shown the intercor- 
 relations among four logical memory tests. Each mem- 
 ory test was given to two separate groups of students. 
 This made possible two complete sets of correlations, 
 twelve in all. The different correlations together with 
 the averages are shown in the table. The four tests 
 were The Dutch Homestead, Cicero, and Marble Statue 
 from Whipple, and The Costly Temper from the au- 
 thor's "Manual." 
 
 The intercorrelations of the memory tests, .537, is 
 but a trifle more than the intercorrelations of the learn- 
 ing tests. Ability to reproduce stories heard is cer- 
 tainly as nearly a unitary function as one could ima- 
 gine, still the correlations between story reproductions 
 are practically the same as the correlations among 
 widely different types of learning. 
 
THE NATURE OF LEARNING CAPACITY 163 
 
 TABLE 21. 
 
 SHOWING THE INTERCORRELATIONS OF FOUR DIFFERENT TESTS 
 OF LOGICAL MEMORY. 
 
 Costly Dutch Marble 
 
 Temper. Homestead. Cicero. Statue. 
 
 524 .568 .556 
 
 Ccstlv Temper .617 .559 .518 
 
 571 .564 .537 
 
 .524 555 .590 
 
 Dutch Homestead 617 .545 .476 
 
 .571 550 .533 
 
 .568 .555 379 
 
 Cicero 559 .545 .550 
 
 .564 .550 465 
 
 .556 .590 .379 
 
 Marble Statue 518 .476 .550 
 
 .537 .533 .465 
 
 Average 557 .551 .526 .512 
 
 Average of the four averages is .537. 
 In Lyon's memory studies we find correlations com- 
 puted for five different types of memorising. In one 
 series of studies with 24 subjects he found an inter- 
 correlation of .51 ; and in another study with 17 sub- 
 jects he found a correlation of .42. His correlations 
 were computed by the rank-difference method. His re- 
 sults are therefore about the same as those reported 
 above. His correlations for different types of memor- 
 ising are practically the same as those in logical mem- 
 ory. 
 
 Regularity of Learning. — Few characteristics of 
 learning are more remarkable than its regularity in the 
 same individual. Two instances will be sufficient for 
 illustration. In an experiment in learning lists of 26 
 nonsense syllables, the number of repetitions necessary 
 to learn them to the point of first reproduction was de- 
 termined. The number of repetitions necessary for re- 
 learning on succeeding days was also determined. It 
 
164 THE PSYCHOLOGY OF LEARNING 
 
 was found that relearning time had a very definite re- 
 lation to learning time and this ratio was fairly con- 
 stant and about the same for all subjects. While the 
 experiment threw no light on the matter of learning 
 different types of material, it showed great definiteness 
 and constancy in learning the same kind of material. 
 The other illustration is from card-sorting. In this 
 kind of learning, nothing was more noticeable than the 
 definiteness and regularity of the different subjects. 
 They retained their relative positions or ranks from 
 day to day with great constancy. At the beginning of 
 an experiment the experimenter could predict, on the 
 basis of the previous day's work, the relative ranks of 
 the subjects. The learning capacity of the various 
 learners became definitely known to the experimenter. 
 The characteristics underlying learning capacity seem 
 to persist and to be as definite as anything in human 
 nature. Such variations as usually occur in a given 
 subject's learning from day to day, have definite causes 
 and can be predicted when the causes are known. Of 
 course, constancy in learning depends upon constancy 
 of the conditions and factors involved. These factors 
 are numerous. The condition of nearly every organ in 
 the body can have its effect on the temporary efl^iciency 
 of the learner. But there is probably a constant factor 
 which may be called general learning capacity, depend- 
 ent upon the characteristics of the central nervous sys- 
 tem. 
 
 Interrelation of Mental Functions. — -The question of 
 the relations of different types of learning capacity 
 raises the more general question of the relation of all 
 mental functions. The question may be put in this way : 
 Are mental functions positively or negatively related? 
 Or is there any dependent relationships among them at 
 
THE NATURE OF LEARNING CAPACITY 165 
 
 all ? If the efficiency of one mental function is known, 
 can we predict the efficiency of others? These ques- 
 tions can be ansv/ered with some degree of certainty. 
 The answer comes from mental tests given in various 
 countries during the last twenty years. We shall show 
 certain typical results and then discuss their signifi- 
 cence. In table 22 are shown the intercorrelations for 
 a group of mental tests given to three different groups 
 of students in the author's laboratory. There were 21 
 correlations for each group, 63 correlations in all. Only 
 the averages are shown in the table. 
 
 What is the significance of the figures shown in the 
 three tables of correlations? In table 22 the correla- 
 tions are all positive but one. Free association gives 
 with substitution a negative correlation. It is small, 
 however, and may be considered zero. The averages of 
 all the correlations, although small, are positive. A 
 careful study of all the correlation tables published 
 shows that all important complex mental functions are 
 positively related. The raw correlations are never very 
 high. The disturbing factors are so many that we can 
 
 TABLE 22 (PYLE). 
 
 SHOWING THE INTERCORRELATIONS AMONG DIFFERENT MENTAL 
 
 FUNCTIONS. 
 
 d d S3 
 
 O 73 . ® 2 
 
 'Z ^' '^ -' bn 'C '^ 
 
 <u -r rj- i; C' a « s 
 
 F— 1 w r< 05 "? ,„ .« ,^ 
 
 S St r ^ ir ? c "z: a^ ^ ^ 
 
 Completion 382 .546 .522 .357 .286 .177 .274 
 
 Logical memory .546 .523 .291 .436 ,232 .261 .122 
 
 ODposites 522 .291 .558 .154 .438 .165 .240 
 
 Rote memory 357 .436 .154 .655 .022 .265 .016 
 
 Word building 286 .232 .438 .022 .654 .200 .050 
 
 Free association .177 .261 .165 .265 .200 .679 —.027 
 
 Substitution 274 .122 .240 .016 .050 —.027 .805 
 
 Averages 330 .315 .302 .208 .205 .174 .113 
 
166 THE PSYCHOLOGY OF LEARNING 
 
 TABLE 23 (FROM BURT). 
 
 SHOWING THE INTERCORRELATIONS (CORRECTED CO-EFFICIENTS) 
 
 IN THE CASE OF THE MORE COMPLEX MENTAL 
 
 FUNCTIONS TESTED. 
 
 ^s S si ^ i I 
 
 a=^ o -^f^ ^ ^ ^ 
 
 w?-i C f^H -<1 J^ Q 
 
 Spot pattern 80 .75 .96 .41 .66 
 
 Dotting 80 84 .85 .22 .83 
 
 Mirror tracing 75 .84 71 .13 .72 
 
 Alphabet 96 .85 .71 47 .83 
 
 Memory 41 .22 .13 .47 18 
 
 Dealing 66 .83 .72 .83 .18 
 
 Averages 72 .71 .63 .76 .28 .64 
 
 TABLE 24 (FROM SIMPSON). 
 
 SHOWING THE ESTIMATED TRUE CORRELATION FOR PEOPLE IN 
 GENERAL, I. E., THE PROBABLE CORRELATIONS AS THEY WOULD 
 BE IF THE SUBJECTS WERE A VERY LARGE NUMBER OF PERSONS 
 REPRESENTING A RANDOM SAMPLING OF ALL THE PEOPLE, IN- 
 STEAD OF TWO SMALL. SELECTED GROUPS. 
 
 s 
 
 fcX) 
 
 
 r3 O 
 
 Or- o a 
 
 at, 
 
 j^B ^^ «S ?^ S.S So 
 
 -2© 5a ^,° a^oj (DOS ot> 
 
 SU Wo g!^ ^Ph mCh U^ 
 
 Ebbinghaus completion 85 .82 .71 .60 .60 
 
 Hard opposites 85 84 .70 .56 .72 
 
 Memory for words 82 .84 80 .65 .61 
 
 Memory for passages 71 .70 .80 30 .31 
 
 Learning pairs 60 .56 .65 .30 44 
 
 Completing v/ords 60 .72 .61 .31 .44 
 
 Averages 72 .73 .74 .56 .51 .54 
 
 never expect to get a very high correlation between any 
 two different mental functions. In table 22 it is seen 
 that the raw correlation between two successive tests 
 of the same kind is not nearly unity. In the seven tests 
 they range from .382 to .805, the average being only 
 .608. This last coefficient indicates the reliability of 
 our measure, for, of course, the true correlation is not 
 .608 but unity. The average raw correlation in this 
 table is .239. The average true correlation would be 
 about .40. 
 
THE NATURE OF LEARNING CAPACITY 167 
 
 We said above that all important mental functions 
 are positively related. This is not quite the truth. 
 What we should say is that all mental functions which 
 may be considered a part of general intelligence, gen- 
 eral mental ability, are positively related. For, of 
 course, it is not true that any aspect of human behavior 
 that we may measure will be found positively related 
 to all other aspects of human behavior. Quickness of 
 reaction time, for example, probably has no relation to 
 intelligence. Rapidity in running has no relation to 
 judgment of color. Strength of grip has no relation 
 to accuracy of ethical judgment. Now, a few perfor- 
 mances that have been used as mental tests have given 
 negative correlations with other tests. The author has 
 found Whipple's ink-blot test, the free association test, 
 and cancellation tests given for speed only, to show 
 small negative correlations with certain other tests. 
 These correlations are usually small, and probably 
 should be considered zero, for they are sometimes posi- 
 tive. Some writers, Simpson for example, have at- 
 tached some importance to these negative correlations. 
 But their explanation is plain. Older children make 
 poorer records in the ink blot test than younger chil- 
 dren, on account of a change in attitude. If we make 
 the test consist in ability to see objects in the blots, 
 as children become older, they do not report many ob- 
 jects owing to their great sophistication. They will 
 not admit that the blots look like anything. The free 
 association test is little if any more than a test of quick- 
 ness of nervous response, and this, as far as we know, 
 has no relation to the value or quality of response. 
 And in the cancellation test when we do not consider 
 accuracy, we are measuring little more than reaction 
 time. None of these three tests has shown any positive 
 
168 THE PSYCHOLOGY OF LEARNING 
 
 relation to general intelligence otherwise determined, 
 and none of these tests is of a'ny value as a measure of 
 intelligence. We repeat, in agreement with Thorndike, 
 that no test that is a measure of general intelligence has 
 a negative relation with any other such test. On the 
 other hand, every test that has proven to be a high 
 criterion of general intelligence, has shown high cor- 
 relation with other such tests. Of all single tests, the 
 completion test has proven to be the best measure of 
 intelligence. This test gives the highest average cor- 
 relation with other tests. The other tests of proven 
 value as measures of general intelligence — opposites, 
 logical memory and analogies — all give high positive 
 correlations v/ith other important tests. From this 
 evidence, only one conclusion seems to us possible. 
 There is a human characteristic which we may call 
 general learning capacity, just as there is another 
 which we may call general intelligence. 
 
 Specific Abilities. — We have a general learning capac- 
 ity. We are quick learners, slow learners or mediocre 
 learners. But when we learn, we work with some par- 
 ticular kind of material. There are therefore two addi- 
 tional factors that must always be taken into account 
 besides the general factor. One of these factors in- 
 volves the specific characteristics necessary to some 
 type of learning. For example, in music, one will learn 
 fast if he has a good general learning capacity and also 
 those specific characteristics necessary to music, par- 
 ticularly ability to differentiate pitch, judgment of har- 
 mony and dissonance, perception of rhythm. In art, 
 one might learn slowly although having good general 
 learning capacity, because of lack of muscular control, 
 or because of being color blind. In all learning requir- 
 ing specific abilities, quickness of learning will depend 
 
THE NATURE OF LEARNING CAPACITY 169 
 
 upon the degree of general learning capacity possessed 
 and also upon the degree to which the specific abilities 
 are possessed. 
 
 The other secondary factor that must always be 
 taken into account is the effect of experience. In the 
 realm of habit, when we start in to form a new one, we 
 usually find some old habit that either helps or hinders, 
 so that our speed of learning is affected by the old habit. 
 And likewise in the realm of knowledge, the knowledge 
 we have already acquired is of great consequence. Our 
 learning is fast because of the previously acquired 
 knowledge or slow because we do not have it. Great is 
 the effect, also, of the attitude which previous experi- 
 ence gives us toward new tasks. Because of past ex- 
 perience we like certain things and dislike others, and 
 these affective attitudes are of the greatest consequence 
 m learning ; helping if favorable and greatly hindering 
 if unfavorable. If by any chance a pupil gets the idea 
 that he can not do a thing, he does not like that thing 
 and he does not try to do it. As time passes, his inabil- 
 ity increases because of the increase of the force of the 
 inhibition and also because of lack of practice. While 
 we therefore have a general learning capacity, it never 
 operates entirely freely but is always complicated by 
 specific factors and by the effects of previous experi- 
 ence. 
 
 Nature of General Learning Capacity. — It is clear 
 that specific learning capacity depends upon the pos- 
 session of certain specific characteristics, such as reac- 
 tion time, muscular co-ordination, quality of sense per- 
 ception, and sensory discrimination, color sense, vari- 
 ous auditory characteristics, ideational type, etc. But 
 on what does general learning capacity depend ? What 
 the characteristics of the central nervous system on 
 
170 THE PSYCHOLOGY OF LEARNING 
 
 which quick learning depends, are, we do not know. In 
 psychological terms, however, something can be said 
 of what characterises a quick learner. The quick learner 
 has what is called in ordinary terminology, the power 
 of concentration. All the available cerebral energy 
 seems to participate in the learning; there is no waste 
 of energy. This characteristic is what is usually spoken 
 of as power of attention. The fact that good learners 
 also remember well, and the further fact that good 
 retention is known to depend upon the vividness and 
 intensity of impression which are secured only in a high 
 state of attention, lend evidence to the assumption that 
 power to learn depends upon power to attend. Since 
 attention is merely sensory clearness, it is probable 
 that a low level of attention means a low level of men- 
 tality. The mental processes of the slow learner would, 
 on this assumption, be more vague, less potent, more 
 poorly knit together. Another characteristic that un- 
 doubtedly affects learning capacity is ability to perceive 
 significance. This factor is particularly effective in the 
 higher realms of intellectual learning, where meaning 
 and organisation are the important things. All of us 
 have about the same sensory experience; the good 
 learner is he who knows what, in all this experience, is 
 significant and what is not. The essential thing in intel- 
 lectual learning is organisation. Organisation depends 
 upon knoioing what and hoiv to organise. The good 
 learner sees what characteristics of experience are sig- 
 nificant for the purposes of his life, he attends to these 
 characteristics to the exclusion of unimportant charac- 
 teristics. The result is a helpful and useful organisa- 
 tion of the items of experience. As a consequence ex- 
 perience is more vital to the good learner, more useful 
 in the after application to life's purposes and more help- 
 
THE NATURE OF LEARNING CAPACITY 171 
 
 fill in learning other related things. The good learner, 
 then, has available a large amount of cerebral energy 
 which he can bring to bear in the learning process, and 
 is able to recognise the significant and distinguish it 
 from the unimportant. The result is a good organisa- 
 tion of what is useful. The poor learner lives on a lower 
 intellectual level. His mental processes are more 
 vague, more loosely knit together. The poor learner 
 can not discriminate. One item of experience is to him 
 about as important as another; there is therefore no 
 selective attention, and as a result no helpful organisa- 
 tion of the items of experience. Experience is therefore 
 of little use in the practical affairs of life or in learning. 
 
 In the very simplest of learning, these two factors 
 can be seen to operate. Let us apply them to card- 
 sorting. In the case of the quick learner, the mental 
 power is all concentrated on the learning. There is no 
 division of attention. The quick learner further sees 
 significant and helpful connections and associations. 
 He says, for example, "such a number is in the corner, 
 another is in the middle, still another is by such and 
 such a number," and so on. I have found that in non- 
 sense learning, the fastest learners are always those 
 that put meanings into the syllables, and hit upon vari- 
 ous schemes of organisation. The poorer learners are 
 dependent solely upon repeated presentation of the 
 stimuli. 
 
 A General Mental Factor.— -Our consideration of the 
 nature of learning leads us to the question of the nature 
 of general intelligence. We have found reason to be- 
 lieve that there is a general factor that operates in all 
 learning, a factor which we call general learning capac- 
 ity. Certain psychologists claim that there is a gen- 
 
172 THE PSYCHOLOGY OF LEARNING 
 
 eral or central intellectual factor that participates in all 
 intellectual activity. This central factor is claimed to 
 be the chief basis of the positive correlation found be- 
 tween any two important intellectual activities. Spear- 
 man is the leading psychologist who advocates the cen- 
 tral factor. The main item of proof of this factor is 
 that in a table of correlations, if the correlations them- 
 selves are high and dependable, any two pairs of col- 
 umns give a high positive correlation. This could be 
 true, Spearman claims, only if there is a common factor 
 in all the tests. Thorndike, Brown, Simpson, and 
 others oppose this theory. Thorndike holds that posi- 
 tive correlation betv/een two activities depends only 
 upon some identity of the functions involved, a certain 
 identity of processes. The evidence available does not 
 make possible a present solution of this question. In 
 the light of such evidence as we now have, I am inclined 
 to believe that there is a general learning factor, and 
 also a general intellectual factor, a factor operative in 
 all intellectual processes. The general learning factor 
 and the general intellectual factor are probably the 
 same. We stated above that the general learning factor 
 may be considered to be the power of attention. Burt 
 made a similar claim for the general intellectual factor. 
 But a high type of intellect, as pointed out above, pos- 
 sesses two characteristics, good learning ability and the 
 ability to recognize significance. It may be that these 
 two characteristics have a common basis. The physio- 
 logical processes and the anatomical structures which 
 underlie the psychological factors we have named and 
 discussed are not known. About all we can say is that 
 some brains are better for the important purposes of 
 life than are other brains. Our brains probably have 
 
THE NATURE OF LEARNING CAPACITY 173 
 
 general characteristics which are potent in all intel- 
 lectual operations, hence, the general learning factor, 
 and the central intellectual factor. But just as surely, 
 brains have certain specific characteristics, effective 
 in specific acts of behavior, characteristics that make 
 one a better seer, or hearer, or taster, for example; 
 characteristics that help or hinder in specific processes, 
 and that must always be considered in connection with 
 the general factor. 
 
 EXEECISES AND EXPERIMENTS 
 1. The object of this experiment is to determine to 
 what extent learning capacity is general. Method: 
 Compute the inter-correlations among all the learning 
 experiment's perfoi*med in this course. The experiments 
 should include card-sorting, different forms of substitu- 
 tion, all the memory tests included in the exercises of 
 Chapter VII, verbatim learning, learning nonsense syl- 
 lables, and any other learning experiments performed 
 with sufficient care to give valid results. How do the 
 correlations compare with those reported in the chap- 
 ter? 
 
 REFERENCES. 
 
 Aiken, H. A. and Thorndike, E. L., Correlations among the per- 
 ceptive and associative processes, P. R., IX, 1902, 374. 
 
 Brown, W., Some experimental results in the correlation of 
 mental abilities, B, J. P., Ill, 1910, 296. 
 
 Brown, W., The essentials of mental measurement, 1911. 
 
 Burt. C, Experimental tests of general intelligence, B. J. P., 
 Ill, 1909, 94. 
 
 Chapman, J. C, Individical differences in ability and improve- 
 ment and their correlations, T. C. Cont. to Ed., No. 63, 1914. 
 
 Cogan, L. G. and Conkliu, A. M. and Hollingworth, H. L., An 
 experimental study of self analysis, estimates of associates and 
 psychological tests, S. and S., II, 1915, 171. 
 
 Dunlap, K., The biological basis of the association of ideas and 
 the development of perception, Psychobiology, II, No. 1, 29. 
 
 Elderton, E. M., On the association of drawing with other 
 capacities in school children, Biometrica, III, 1909-10, 222. 
 
 Haggerty, M. E., The laws of learning, P. R., XX, 1913, 411. 
 
174 THE PSYCHOLOGY OF LEARNING 
 
 Hart, B. and Spearman, C, General aUlity, its existence and 
 nature, B. J. P., V, 1912, 51. 
 
 Hollingworth, H. L., Correlation of abilities as affected hij 
 practice, J. E. P., IV, 1913, 405. 
 
 Individual differences lefore, during, and after practice, 
 P. R., XXI, 1914, 1. 
 
 King, I, The relationship of abilities in certain mental tests to 
 ability as estimated by teachers, S. and S., V, 1917, 204. 
 
 Kline, L. W. and Owens, W. A., Preliminary report of a study 
 in the learning process involving feeling tone, transference and 
 interference, P. R., XX, 1913, 206. 
 
 Krueger, F. and Spearman, C, Die correlation zwischen ver- 
 schicdenen (jeistlgen leistungsfUhigkeiten, Zeit., f. Psych. Bd., 
 XLIV, 1906, 50. 
 
 Lashly, K. S., Studies of cerebral function in learning, Psycho- 
 biology, II, No. 1, 55. 
 
 Lyon, D. O., Memory and the learning process, 1917. 
 
 The relation of quickness of learning to retentiveyiess, 1916. 
 
 McCall, W. A., Correlation of some psychological measurements 
 with special attention to the measurement of mental ability, 
 S. and S., V, 1917, 24. 
 
 Peterson, A., Correlation of certain mental tests in normal 
 school students. P. R., XV, 1918, 323. 
 
 Pyle, W. H., Standards of oriental efficiency, J. E. P., IV, 1913, 
 61. 
 
 Rietz. H. L., 0?^ the correlation of the marks of students in 
 mathematics and in lau\ J. E. P., VII, 1916, 87. 
 
 Simpson, B, R., Correlations of mental abilities, T. C. Cont. to 
 Ed. No. 53, 1912. 
 
 Spearman, C, General intelligence objectively determined a7id 
 measured, A. J. P., XV. 1904, 201. 
 
 Manifold sub-theories of the "Tivo factors:' P. R., XXVII 
 1920, 159. 
 
 Starch, D., Correlations among abilities in school studies 
 J. E. P., IV, 1913, 411. 
 
 Thompson, G. H., A hierarchy without a qeneral factor, B. J. P. 
 VIII, 1916, 271. 
 
 Thorndike, E. L., The effect of practice in the case of a purely 
 intellectual function, A. J. P., XIX, 1908, 374. 
 
 and Lay W. and Dean, P. R., The relation of accuracy in 
 sensory discrimination to general intelligence, A. J. P., XX 
 1909. 364. 
 
 Educational Psychology, Vol. Ill, 1914, 315. 
 
 Weglin, D. E., 77? c correlation of abilities of high school pupils 
 1917. 
 
 Whitley, M. T., An empirical study of certain tests for indi- 
 vidual differences. Arch, of Psych., No. 19, 1911. 
 
 Wissler, C, The correlation of mental and physical tests, P. R. 
 Mon. Sup., Ill, No. 16, 1901, 62. 
 
Chapter IX. 
 MEASURING LEARNING CAPACITY. 
 
 In the last chapter we came face to face with the 
 problem of measuring learning capacity, for in order 
 to compare ability in one field of learning with ability 
 in another, we must have some means of measuring 
 capacity to learn. Since the question of measuring 
 learning capacity is an important one and has many 
 details that demand specific investigation, we have re- 
 served its treatment for a separate chapter. Our prob- 
 lem here is but a part of the more general question of 
 measuring all mental functions. Into the general ques- 
 tion of mental measurements, we shall not enter. The 
 possibility of such measurements and their practical 
 feasibility have been fully demonstrated. 
 
 Learning is forming bonds between stimuli and re- 
 sponses or between ideas. Common observation as well 
 as experiment shows that the rate and effectiveness 
 with which we form these bonds vary from one person 
 to another. In order to measure the efficiency of a 
 learner we have but to set him a task of bond-forming 
 and determine the amount accomplished in a given 
 time ; by comparing this amount with the results from 
 other learners we can determine whether the learner 
 being measured is fast or slow. 
 
 Such a procedure seems quite simple, as indeed it is. 
 But before we can feel confidence in the results of such 
 measurements, we must have information on several 
 
 [175] 
 
176 THE PSYCHOLOGY OF LEARNING 
 
 details of the procedure. In the first place, if we are to 
 compare the efficiency of one learner with that of an- 
 other, the learners must not only be set the same task 
 but each must work with all his might. If one learner 
 puts forth all possible energy, and the other does nat 
 try, of course, the results are not comparable. A mental 
 test of any kind is valid only on the assumption that the 
 learner does his best. The validity of a mental test de- 
 pends further on those tested being in good physical 
 and mental condition. By good physical condition, we 
 mean the subject must be free from fatigue, sickness 
 or any other physical ailment that would interfere with 
 learning. By proper mental condition, we have refer- 
 ence to emotion and attitude- The person being meas- 
 ured should not be excited unduly or nervous. A favor- 
 able attitude is always desirable, because it is necessary 
 to get the subject to do his best. So much for the con- 
 ditions required on the part of the subject. There are 
 also certain problems of method that must be consid- 
 ered. In comparing the learning efficiency of different 
 learners, we compare their learning curves. But what 
 point in the curves shall we take for comparision ? Shall 
 we measure a group of learners which we wish to com- 
 pare after they have worked a short time or after they 
 have worked a long time? The results may be differ- 
 ent. We do not know in the absence of experiment. We 
 do not know whether some learners may start out 
 slowly and become much faster later, while some may 
 start out fast and become slower later. If we allow our 
 subjects to work till they have reached their limit in 
 the given task, and then measure them, we are evi- 
 dently measuring final efficiency, which may be quite 
 a different thing from learning capacity. For all we 
 
MEASURING LEARNING CAPACITY 177 
 
 know to the contrary, a quick learner may soon reach 
 his limit while a slow learner may continue to improv.^ 
 and reach a final efficiency much greater than that 
 reached by the fast learner. It is evident therefore 
 that the length of time we allow our subjects to work 
 before taking our measure is an important matter. 
 These various questions of method we shall consider 
 below. We must first see what sort of tests have been 
 used as measures of learning capacity. 
 
 Learning Tests. — The type of learning test most ex- 
 tensively used is the substitution test, which is one of 
 the simplest of tests and has a wide range of possibili- 
 ties in the materials that can be used and the varia- 
 tions that can be made in them. The principle of the 
 test is as follows: The learner is given a number of 
 symbols with which he is familiar and is required to 
 couple with these symbols others which have not been 
 so coupled with them in the experience of the subject. 
 In a form of the experiment standardised by the au- 
 thor, the known symbols are the nine digits, and the 
 symbols that are to be coupled with them are letters of 
 the alphabet. The procedure in giving the test is as 
 follows: The subject is given the test sheet which 
 contains columns of numbers, five digits in a number. 
 After the numbers are five squares in which the proper 
 letters are to be written. At the top of the test sheet 
 is a key showing what letter is to be coupled with each 
 digit. The child, after hearing an explanation of the 
 nature of what he is required to do, proceeds at once to 
 fill out the squares, looking up at the key to see how to 
 do it. The theory of the test is that the quick learner 
 will soon know the key and will not have to look at it ; 
 he therefore makes a higher score in a given time. As 
 
178 THE PSYCHOLOGY OF LEARNING 
 
 the author has used the test, eight minutes were allowed 
 in grades three, four, and five, and five minutes in the 
 grades above. The score — indicating the learning capa- 
 city — is the number of correct substitutions made in a 
 minute. 
 
 In practice, this test does not turn out to be just what 
 it is in theory. The discrepancy consists in the fact 
 that some subjects do not try to learn the key, but con- 
 tinue to refer to it throughout .the experiment. Since 
 these are the less intelligent pupils, the error is not so 
 great as might be expected. Nevertheless a better way 
 to administer a substitution test is to provide the sub- 
 jects with the key and allow them to study it for a cer- 
 tain length of time, then require them to write the 
 substitutions from memory. In Table 25 are the norms 
 for public school children by age and sex; method, 
 pupils using the key while practicing. 
 
 TABLE 25. 
 
 SUBSTITUTION TEST, CITY CHILDREN. METHOD— USING KEY WHILE 
 
 PRACTICING. SCORE— NUMBER CORRECT SUBSTI- 
 
 TIONS PER MINUTE. 
 
 Boys Girls 
 
 Age. No. Ca;=es. Averag-e. No. Cases. Averaj^e. 
 
 8 223 7.95 213 9.11 
 
 9 296 10.08 355 10.86 
 
 10 410 11.81 380 13.82 
 
 11 443 13.43 387 15.88 
 
 12 399 15.48 433 18.29 
 
 13 401 16.80 424 20.31 
 
 14 „ 308 19.26 344 22.21 
 
 15 255 22.13 274 24.17 
 
 16 173 23.71 217 26.92 
 
 17 93 26.42 141 28.12 
 
 18 52 24.41 86 28.39 
 
 The substitution test has a high co-efficient of relia- 
 bility, and it gives the highest average correlation with 
 the other learning tests, as shown in table 20. With the 
 
MEASURING LEARNING CAPACITY 179 
 
 other mental tests, it gives low but constant correla- 
 tions. While it gives low correlations with the other 
 types of mental tests, if its correlations with the other 
 tests are themselves correlated with the averages of the 
 correlations of each test with all the others as shown 
 at the bottom of the correlation table (p. 165), it gives 
 a higher correlation than any other test similarly 
 treated. This indicates the most constant presence and 
 regularity of the factor common to the tests. 
 
 Many different forms of the substitution test have 
 been used. Their validity depends upon how they are 
 given and how they are scored. If they are properly 
 administered and properly scored, and repeated so as 
 to give a stable measure, they give a measure of con- 
 siderable reliability. 
 
 Tests of Motor Learning. — A motor learning test ex- 
 tensively used by the author and, in various forms, 
 used in other laboratories, is the card-sorting test de- 
 scribed in Chapter VIII. This test has many advan- 
 tages. It has no relation to book-learning or any of the 
 ordinary learning that goes on at school. By using 
 playing cards or cards with colors instead of numbers 
 it can be given to illiterates or very young children. The 
 demands of the test are easily understood. By sorting 
 into a small number of boxes, the learner soon gets suf- 
 ficient data for the construction of a learning curve. 
 The test practically always arouses interest and the 
 subjects do their best. Its only disadvantage is that in 
 the case of older subjects, those who are adept in the 
 handling of cards, make a somewhat higher score than 
 they otherwise would. 
 
 The card-sorting experiment has many possibilities. 
 The author uses a card-sorting box having six rows of 
 
180 THE PSYCHOLOGY OF LEARNING 
 
 trays with five trays in a row, thirty in all. These are 
 numbered in miscellaneous order from 11 to 41. On 
 the reverse side of the box there is the same number 
 of trays. For the study of inhibition, these have the 
 same numbers with a different arrangement. There 
 are five cards of each number, 150 in all. In six days 
 of practice not only can reliable measures of learning 
 capacity be obtained but data be secured for the con- 
 struction of six comparative learning curves, and facts 
 be discovered which throw light on the problem of 
 transfer of training. The procedure is as follows : On 
 the first day of practice, the learners spend an hour 
 sorting into the first row of boxes. Since there are only 
 five boxes in the row, their location is soon learned and 
 the learning curve rises fast. On the second day, the 
 learners spend an hour on the second row of boxes. On 
 the third day, they take the third row, and so continue 
 to take a new row each day till all the rows are used. 
 This requires six days if only one side of the box is 
 used. By going on with different numbers, both sides 
 of the box can be used, giving twelve days of learning 
 with a new set of boxes for each day. Such an experi- 
 ment teaches students more about habit-formation and 
 the laws of learning than they could learn from a book 
 in a long time. Indeed such facts can not be adequate- 
 ly learned from a book in any length of time. Facts 
 discovered in this experiment which throw light on the 
 problem of inhibition and also that of the transfer of 
 training are discussed in later chapters. 
 
 The marble-sorting experiment, discussed in Chapter 
 VIII, is a valid measure of capacity in motor learning. 
 It has the advantage also of being different from the 
 ordinary school work, and arouses great interest. It 
 
MEASURING LEARNING CAPACITY 181 
 
 requires a high degree of attention and concentration. 
 It can be given to illiterates and young children. Sub- 
 jects who have had much experience v/ith the piano or 
 typewriter have a slight advantage in the score. The 
 author used this experiment with success in comparing 
 the learning capacity of negro children with v/hite chil- 
 dren. The comparisons are shown in Table 26. 
 
 In the field of ideational learning the number of 
 kinds of learning tests that can be devised is as great 
 as the number of kinds of learning material. Nonsense 
 syllables, properly constructed and arranged, furnish 
 us with one of the best means of measuring learning 
 capacity. Their learning demands a high degree of 
 concentration. In a few hours several series can be 
 learned, affording an adequate measure of this form of 
 learning. The material has the advantage that it is 
 absolutely new to experience. Previous experience gives 
 no bias. The associations must be established ab novo. 
 
 It may be claimed that the best types of learning 
 tests should use meaningful material. The author has 
 some sympathy with this claim. Such material has 
 many advantages, but has the disadvantage that when 
 used for comparative purposes, it will favor certain 
 subjects or groups of subjects, because of differences in 
 previous experience and interest. If we take great care 
 in the selection of test material, such tests as those 
 usually given under the head of logical memory are of 
 great value in measuring differences in learning capac- 
 ity. There are various ways in which they can be ad- 
 ministered. A description of two methods will be suf- 
 ficient. A short story can be divided into ideational 
 units for ease in grading the results. This story is 
 read to the subjects once. They then make a written 
 
182 THE PSYCHOLOGY OF LEARNING 
 
 TABLE 26. 
 THE LEARNING CAPACITY OF NEGPvO CHILDREN. 
 
 The numbers in the table represent the percentages which the 
 Negro scores were of the scores of white children for the various 
 ages. Test used, marble-sorting experiment. 
 
 Ages 9 10 11 12 13 14 15 Av. 
 
 Girls 81 76 72 79 73 75 73 75.6 
 
 P>oys 84 79 80 74 77 83 81 79.7 
 
 reproduction of the ideas. The number of ideas ade- 
 quately reproduced is the score. The other method of 
 giving the test is to read the story once, have a repro- 
 duction, then continue to read it and have it reproduced 
 until all the ideas are reproduced. This procedure has 
 been used by the author for individual tests, the repro- 
 ductions being made orally. The former method re- 
 quires much less time, and if repeated with four differ- 
 ent kinds of material, gives a valid measure of learning 
 capacity. 
 
 To summarise: Learning is connecting. Any test 
 that measures the capacity of a subject to establish 
 connections between two processes not before asso- 
 ciated, is a valid measure of general learning capacity. 
 In practice, we should use several different types of 
 test. Each test should be repeated at least twice, and 
 preferably four times. The results of the several tests 
 should then be combined and taken as a measure of 
 learning capacity. Any test of learning capacity that 
 uses only one kind of material and is given but once 
 can be only a rough measure of learning capacity. In 
 ordinary practice, only group tests can be used. The 
 author recommends the use of the substitution test, the 
 second procedure mentioned above, repeated four 
 times ; card-sorting, an hour test with each of six boxes ; 
 the learning of four ten-syllable series of nonsense syl- 
 lables ; and four tests of logical memory. The combined 
 
MEASURING LEARNING CAPACITY 183 
 
 results are to be taken as the measure of general leani- 
 ing capacity. 
 
 Where to Make the Measure.-— If our purpose is 
 merely to measure the quickness of learning, we have 
 but to set our subjects to work forming new bonds, let 
 them work for a time, then measure the degree of per- 
 fection with which the bonds are formed. But how 
 long should we let them work ? If there are only a few 
 bonds to form, and the subjects work for a considerable 
 time, and we then determine the subjects' relative effi- 
 ciencies in using the bonds, we are not measuring learn- 
 ing capacity at all, but the various individuals' capaci- 
 ties to do a certain thing. We should take our measure 
 early in the practice period in order to measure learning 
 capacity and not final capacity. We should take our 
 measure while the learning curve is still rising steeply. 
 How many minutes or hours should precede the appli- 
 cation of our measure depends upon the number and 
 complexity of the bonds. In the case of card-sorting, 
 the efficiency attained with five boxes in an hour gives 
 a good measure. With thirty boxes, a better measure 
 would be had after a longer period. The only advan- 
 tage of a longer period, however, is to offset minor 
 errors that would unduly affect a shorter period. In 
 general, the longer the period covered by any mental 
 test, the more reliable the test. For example, in a test 
 covering a minute, the various accidents that happen 
 in getting started — accidents with paper, pencil, etc. — 
 have an appreciable effect on the score. Such accidents 
 do not have a measurable effect on a test covering an 
 hour. In learning involving the formation of very 
 many bonds, we should not apply the measure until all 
 the bonds are formed at least weakly. 
 
184 THE PSYCHOLOGY OF LEARNING 
 
 The importance of knov/ing when to apply our meas- 
 ure is at once apparent when we undertake to measure 
 the learning capacity of people having widely different 
 learning capacities, as negroes and whites. If we take 
 the efficiency at the end of five minutes practice with a 
 nine-digit substitution test, we find the whites very 
 much superior. If we should take as our measure the 
 efficiency attained by twenty minutes of practice, the 
 negroes are found to be as good as the whites. In the 
 first case we are really measuring learning capacity ; in 
 the second case, we are measuring not learning capacity 
 at all but quickness at copying. At the latter task, the 
 negroes are as good as the v/hites. 
 
 It is quite evident that at different points on the 
 learning curve, we measure different things. We said 
 above that in the case of card-sorting involving 30 
 boxes, we should not make our measure too early. The 
 reason for this is that in such a case we are measuring 
 neither learning capacity nor final efficiency, but how 
 quickly a person can look over a list of numbers and 
 find a certain one. 
 
 Correlations in Learning Experiments.— The various 
 questions raised in the above discussions can be scien- 
 tifically studied by applying the correlation formula. 
 We can put a group of people to vv^orking at a learning 
 experiment and allow them to continue till considerable 
 fixation is reached, taking measures of efficiency at fre- 
 quent intervals. We can then compare the relative 
 standing of our subjects at any point with their stand- 
 ing at any other point. In nine-digit substitution, for 
 example, in an experiment involving 29 subjects v/ho 
 worked for five-minute periods, the correlation of 
 standing at the end of the first five minutes with that 
 
MEASURING LEARNING CAPACITY 185 
 
 at the end of five practice periods was .64, while the 
 correlation of the standing at the end of the second 
 period with that at the end of the fifth was .96. In a 
 learning experiment so simple as this, the subjects very 
 soon assume relative positions that are indicative of 
 their final efficiencies. Even in a complicated learning 
 experiment such as sorting cards into 30 boxes, a group 
 of subjects soon reach a relative position that changes 
 little. In the case mentioned, four subjects on the fourth 
 day attain relative positions that do not change during 
 the rest of the experiment. 
 
 Relation of Learning to Other Functions. — The rela- 
 tion of learning capacity to other mental functions and 
 to general intelligence can be determined by correlating 
 the results of learning tests with the results of other 
 types of mental tests and with other measures of gen- 
 eral intelligence, such as class standing or the estimates 
 of teachers. In Table 27 are shown the correlations of 
 three learning tests with the results obtained from the 
 Army Alpha test. 
 
 TABLE 27. 
 
 LEARNING CACAPITY CORRELATED WITH GENERAL INTELLIGENCE 
 AS DETERMINED BY THE ARMY ALPHA TEST. 
 
 Digit-symbol substitution with Army Alpha .32 
 
 Marble-sorting with Army Alpha 30 
 
 Card-sorting with Army Alpha 25 
 
 If learning capacity is determined by combining the 
 results of the substitution test, marble-sorting test, 
 and card-sorting test, and the combined rating corre- 
 lated with general intelligence as determined by seven 
 group tests, the correlation is found to be .42. The 
 group tests were tests of logical and rote memory, 
 word-building, substitution, opposites, completion test. 
 
186 THE PSYCHOLOGY OF LEARNING 
 
 analogies. The corresponding correlation obtained from 
 a different group of students was .467. With this sec- 
 ond group of students the substitution test alone gave 
 a positive correlation of .42 with general intelligence; 
 marble-sorting test, a correlation of .33 ; and nonsense- 
 learning, a correlation of .46. 
 
 In a class of 40, class standing was determined by- 
 three examinations, general intelligence from the re- 
 sults of seven group tests mentioned above and learn- 
 ing capacity by the three learning tests — marble-sort- 
 ing, card-sorting, and digit-symbol substitution. The 
 following correlations were obtained: 
 
 Class standing: with 
 
 Average of mental tests gives 41 
 
 Average of learning tests. 436 
 
 Logical memory alone 22 
 
 Completion alone 29 
 
 The high diagnostic value of learning tests is at once 
 evident. If learning capacity could be determined with 
 absolute accuracy and class standing determined with 
 equal accuracy, the correlation between the two would 
 probably not be over .50 or .60, because of the many 
 other factors that enter into the determination of class 
 standing. Learning capacity is probably the most im- 
 portant single factor, but previous preparation for the 
 course, interest in the course, amount of time spent in 
 study upon the course are impoi1:ant factors in deter- 
 mining class standing. 
 
 In conclusion it may be said that general learning 
 capacity can be determined by the use of several forms 
 of learning tests, and that the results of such tests are 
 valid in practical diagnosis. 
 
MEASURING LEARNING CAPACITY 
 
 187 
 
 Fjgure 20. Learning Curves, Digit-letter Sustitution. The 
 group average is shown by the heavy line 5, the other lines show 
 individual records. After the first few practices, the individuals 
 maintain their relative ranks with considerable constancy. How- 
 ever, number 6 starts below the average and finishes above the 
 average, aDd number 1 starts highest and soon falls to near the 
 average, finishing but little above. 
 
188 THE PSYCHOLOGY OF LEARNING 
 
 The amount of regularity to be found in a practice 
 experiment, and the relative ranks of the subjects at 
 different periods of practice, are shown in Table 28. 
 The experiment which furnished the data for this table 
 was as follows : Thirteen subjects practiced at a substi- 
 tution test, five minutes at a time, four times a day for 
 three days. The results for each practice are converted 
 to a group average of 50. This enables the reader to 
 determine at a glance how each subject stood at any 
 time with reference to the group average. It will be 
 noticed that some subjects are always above the group 
 average and others are always below, while some who 
 are close to the average are sometimes a little above and 
 at other times a little below. If the first two tests are 
 combined and taken as a measure of initial speed, and 
 the last two are similarly combined as a measure of 
 final speed, and the two arrays correlated, the correla- 
 tion is found to be .554. If all those subjects above 
 the average are combined into one group and those 
 below the average are combined into another group and 
 graphs constructed to show their relative position above 
 and below the average, we get the results shown in 
 Fig. 21. It will be seen that the fast group as a whole 
 is fast throughout, and the slow group as a whole is 
 slow throughout. 
 
 The actual raw correlation of each practice period 
 with the last is as follows: 1, .383; 2, .700; 3, .755; 
 4, .834; 5, .793; 6, .767; 7, .861; 8, .939; 9, .965; 10, .952; 
 11, .976. 
 
 In comparing learning capacity in one kind of mate- 
 rial with that in another, it is of special importance 
 that we know just where to apply our measure. If we 
 apply it at a point where the curves are still steep, we 
 
MEASURING LEARNING CAPACITY 
 
 189 
 
 6o 
 
 30 
 
 ^o 
 
 A*-?/ 
 
 / 2 Ji 4 -^ € 7 S 9 ^o // /z 
 
 Figure 21. Graphs fob Substitution Experiment. The horizont- 
 al line shows the group average. The upper graph is the learn- 
 ing curve for those who averaged above the group average, and the 
 lower curve is for those who were below the group average. 
 
 are really comparing learning efficiency. If we apply 
 it after the curves have become relatively flat, we are 
 comparing the efficiency of two different functions. In 
 the latter case, the correlation might be high and posi- 
 tive or it might be high and negative. 
 
 An important question arises here. What is the re- 
 lation of learning capacity to final efficiency? The 
 question can be answered by correlating standing early 
 
190 THE PSYCHOLOGY OF LEARNING 
 
 TABLE 28. 
 
 SUBSTITUTION EXPERIMENT. TWELVE PRACTICE PERIODS. THIR- 
 TEEN SUBJECTS. SCORES ALL REDUCED TO A GROUP 
 AVERAGE OF 50. 
 
 The successive practices are shown from left to right. 
 Sub- 
 ject. 1 2 3 4 5 6 7 8 9 10 11 12 Av. 
 
 A 43 42 44 50 51 52 52 56 55 49 58 56 51 1 
 
 B 46 54 45 52 37 40 47 47 50 52 61 57 49 — 1 
 
 C 45 49 43 40 37 33 31 30 32 30 29 28 36 —14 
 
 D 59 57 53 50 49 56 53 55 58 55 63 59 56 6 
 
 E 59 52 59 52 57 61 57 57 58 57 61 56 57 7 
 
 F 53 46 52 51 50 55 54 52 49 51 50 51 51 1 
 
 G 46 31 33 41 43 44 40 42 43 42 42 44 41 — 9 
 
 H 48 42 46 45 47 44 43 39 39 40 38 35 42 — 8 
 
 1 69 69 64 62 61 50 55 60 65 70 68 66 63 13 
 
 J 48 67 73 78 74 73 76 68 69 65 70 67 69 19 
 
 K 44 39 38 33 40 41 40 40 36 39 35 32 38 —12 
 
 L 38 53 58 53 64 62 61 62 53 59 61 59 57 
 
 M 51 47 42 43 40 38 40 41 42 40 43 40 43 — 7 
 
 in the learning curve v^ith standing after the curve has 
 become relatively flat. In all such computations made 
 by the author, the correlations are found to be positive. 
 As a rule, the quick learners have the highest final effi- 
 ciency. It therefore turns out that although learning 
 capacity and final efficiency are different things, they 
 are positively related. In a group of learners there is 
 not much change of position after practice has pro- 
 ceeded a little way. Of course there is some change of 
 position due to the fact that the particular material 
 may demand specific abilities that are not possessed in 
 the same proportion as learning capacity, and also due 
 to different methods used by the learners. In some 
 cases an early score is low because of a method that 
 does not give good results immediately, but later. In 
 card-sorting, for example, a low early score may be due 
 to the fact that the subject is trying to fix the location 
 of boxes in mind. A valid measure of learning should 
 be taken late enough for such slow learning to show 
 
MEASURING LEARNING CAPACITY 191 
 
 its effects, perhaps at a point where the curve is most 
 convex. In general ,then, the learners who stand high 
 early in the practice periods, stand high later. This is 
 shown in Fig. 20, page 187. It will be noticed that 
 those who are high early are high at the end of a prac- 
 tice. 
 
 Holling*worth has made a study of the effects of 
 practice on correlations, but his study did not involve 
 learning capacity in the sense that we are using it. He 
 studied the following functions: adding, opposites, col- 
 or-naming, discrimination, cancellation, co-ordination, 
 and tapping. The correlation in the preliminary test 
 was .41. In the 5th, 25th, 50th, 80th, 130th, and 175th 
 practices, the average correlations of each with all the 
 others v/ere respectively .61, .73, .77, .85, .92, and 1.00. 
 This increase in the correlation is due to the stabilizing 
 of the performances. After 175 practices each of these 
 tests measured the same thing, namely reaction time. 
 In any kind of motor performance, final efficiency de- 
 pends upon reaction time. In the case of learning ex- 
 periments the author has often found a higher correla- 
 tion between early practices than between later ones. 
 In studying correlations, therefore, we should always 
 bear in mind what the things are which we are corre- 
 lating. Burt finds a higher correlation with intelligence 
 before than after practice in the case of mental tests. 
 This is because intelligence is more closely related to 
 quickness of learning than to final efficiency, although 
 all three are generally in some degree positively related. 
 
 EXPERIMENTS AND EXERCISES. 
 1. The learning capacity of the members of the class 
 may be determined by combining the results of all the 
 learning experiments so far used in the course. Others 
 
192 THE PSYCHOLOGY OF LEARNING 
 
 can be given as described in this chapter. In combining 
 the results, the class averages in each test should be 
 reduced to 50 and all the scores expressed in relation to 
 this average. The scores should then be added and 
 divided by the number of tests. All students whose 
 combined score average is above 50 are of course above 
 the class average, those below 50 are below the class 
 average. 
 
 For references, see page 173. 
 
Chapter X. 
 DIFFERENCES IN LEARNING CAPACITY. 
 
 Individual Differences. — When we measure the learn- 
 ing capacity of a group of students, we find great indi- 
 vidual differences. In Figure 22 is shown the distribu- 
 tion of 443 university students with reference to their 
 learning capacity. The test used was the substitution 
 test. The subjects were all students in the same course 
 in educational psychology in successive classes. Men 
 and women are included together in the records shown 
 in the figure. There were about one-fifth as many men 
 as women. Efficiency is indicated by the scores shown 
 on the horizontal axis. The efficiency scores range 
 from 28 to 72. The average of all the scores is 50. The 
 peaks just above and just below 50 are probably due to 
 our combining both sexes in the same curve, for the 
 women average ten per cent, better in learning capacity, 
 as shown by this test, than the men. 
 
 A child's status in school is determined chiefly by his 
 general intellectual ability, but learning capacity is 
 doubtless the main factor in intellectual ability. The 
 variability in learning capacity is well shown in Figures 
 23 and 24. In Fig. 23 is shown the age distribution of 
 2943 boys and 3152 girls in the eighth grade of the 
 Detroit schools. It will be seen that eighth grade chil- 
 dren range in age from 10 to 17. In Fig. 24 difference 
 in ability is shown in a different way. The graphs show 
 the grade distribution of all the eleven-year old children 
 
 [193] 
 
194 
 
 THE PSYCHOLOGY OF LEARNING 
 
 /<' 
 
 ^'f'' 
 
 4o\ 
 
 M 
 
 M 
 
 FiGUHE 22. Frequency Surface showing the distribution of 443 
 university students with reference to their learning capacity as 
 determined by the substitution test. 
 
DIFFERENCES IN LEARNING CAPACITY 195 
 
 in the city of Detroit. It will be seen that they range 
 from the first grade to the ninth. The graphs repre- 
 sent 2382 boys and 2457 girls. 
 
 Collings found the distribution of 112 twelve-year old 
 boys in the rural schools of a certain county to be as 
 shown in Table 29. The author studied experimentally 
 the mental differences of forty-four twelve-year old 
 pupils in the schools of a certain small city of Missouri. 
 Their distribution through the grades and their men- 
 tality as shown by four tests are given in Table 30. 
 
 TABLE 29. 
 
 SHOWING THE GRADE DISTRIBUTION OF 112 RURAL BOYS. 
 
 Grade. 1234537 8 
 Number pupils 4 4 14 17 39 13 18 3 
 
 TABLE 30. 
 
 SHOWING THE DIFFERENCES IN ABILITY OF 44 12-YEAR-OLD PUPILS. 
 
 School Number Completion Logical Word Oppo- 
 
 Grade. Pupils. Test. Memory. Building". sites, 
 
 4.... 4 17.8 22.2 3.5 2.5 
 
 5 10 41.6 33.0 8.5 6.4 
 
 6 23 50.4 34.2 6.6 7.9 
 
 7 7 59.9 42.5 7.5 , 9.4 
 
 In four different high schools, the author selected the 
 brightest and the dullest pupils on the basis of class 
 standing and teachers' judgments, and determined their 
 general ability by means of various tests. The tests 
 v/ere somewhat different in different schools. The re- 
 sults of all these studies are shown in Table 31. The 
 figures, expressed in per cents, in every case indicate 
 the amounts by which the good pupils excelled the poor 
 ones. 
 
 The scores of the bright pupils are, on the average, 
 about 21 % better than those of the dull pupils. 
 
196 THE PSYCHOLOGY OF LEARNING 
 
 SOfi 
 
 A 
 
 / \ 
 
 / \ 
 
 / \ 
 
 / \ 
 
 / \ 
 
 / \ 
 
 d'oa 
 
 Boo 
 
 /?^ 
 
 23 
 
 /O 
 
 /2 /^ 
 
 /S /^ 
 
 Figure 23. Frequency Surfaces, solid line, boys ; broken line, girls ; 
 showing the distribution of eighth grade children in the city ot 
 Detroit with reference to age; 2943 boys; 3152 girls. 
 
DIFFERENCES IN LEARNING CAPACITY 
 
 197 
 
 Figure 24. Frequency Surfaces showing the distribution of eleven 
 year old children, city of Detroit, with reference to grade, solid line 
 boys, broken line girls; 2382 boys, 2457 girls. 
 
198 
 
 THE PSYCHOLOGY OF LEARNING 
 
 TABLE 31. 
 
 SHOWING THE AMOUNTS— EXPRESSED IN PER CENTS.—BY WHICH 
 BRIGHT PUPILS EXCELLED DULL ONES . 
 
 .i -2 i li 
 
 ■sc t: a ^ ^ o| 
 
 -=) o'l' ca- »r3 c :r om 
 
 j^ ^Jl^ p^;^ ?^M U CO U*< 
 
 A .. 23 36 39 13 24 
 
 B 23 15.0 28 25 11 8 
 
 C 23 28 13 
 
 D 33 2.4 18 17 28 
 
 Average... 26 8.7 27 29 17 18 
 
 The variation in ability in learning capacity in the 
 case of university men and women is shown in Figure 
 25. The type of learning was marble-sorting. There 
 were 213 women and 127 men. It will be seen that the 
 range of ability is about 1 to 5. Still another compari- 
 
 FiGUEE 25. Frequency Surfaces showing the distribution of 
 university men and women with reference to learning capacity as 
 determined by a marble sorting experiment. Solid line represents 
 women and the broken line, men. 
 
 son of university students is shown in Figure 18. The 
 test was the immediate reproduction of a story read to 
 the subjects, who were 516 university women and 277 
 university men. In order that the surfaces might show 
 
DIFFERENCES IN LEARNING CAPACITY 199 
 
 3o 
 
 r,f.z^ 
 
 /c 
 
 Q X /0 7s zo Z^ Jo 
 
 9 ,4 ^9 Zf ^^ ^ 
 
 Figure 2G. Frequency Surfaces showing the distribution of all 
 the children in a school with reference to learning capacity as 
 determined by a substitution test. 1 represents girls, and 2 
 represents boys. 
 
200 THE PSYCHOLOGY OF LEARNING 
 
 the same area, the number of men was multiplied by 
 the ratio 1.86. In this experiment, the ratio of the 
 poorest to the best is about 1 to 9. 
 
 The distribution of the pupils of an entire school 
 system on the basis of learning capacity (Webster 
 Groves, Mo.) is shown in Figure 26. These frequency 
 surfaces represent all the pupils of both sexes and in 
 all grades. In Figure 27 are shown the very gi'eat 
 diiferences in ability found in the same grade. The 
 curves in this figure are based on mental measurements. 
 The subjects are the pupils in a small school system. 
 It will be noticed that the range within a grade is very 
 great and that the grades overlap. A large number 
 of pupils in the fifth grade, for example, have no better 
 ability than many in the fourth. 
 
 Figure 27. These graphs show the range of ability in the same 
 grade and the overlapping of grades. Ability was determined by 
 seven group tests. The pupils were all the children in the grades 
 indicated, in the schools of Webster Groves, Mo. 
 
 Definite difi:erences in capacity are forcibly shown 
 in Figure 28. The graphs were constructed as follows: 
 Four subjects sorted cards one hour daily for 15 days. 
 After several days they sorted seven times in one hour. 
 The average of each sorting for all the days was taken 
 for each person and from these averages, the graphs 
 were constructed. The records for the graphs are in 
 the form of number of seconds required to sort 150 
 cards. It is evident that these four subjects are definite 
 
DIFFERENCES IN LEARNING CAPACITY 201 
 
 Figure 28. Learning Curves, four subjects, card-sorting. The 
 scores are the number of seconds required to sort 150 cards. 
 
 and distinct in their several abilities in mastering card- 
 sorting and in sorting them after the various habits 
 were formed. It will be seen that on the third day, 
 the subjects assumed relative positions that did not 
 afterward change. 
 
 Learning Capacity of Country Children. — A compari- 
 son of the learning capacity of country children with 
 that of city children is shown in the following Table. 
 The children studied were all the pupils in the country 
 schools of a certain Missouri county. The city norms 
 with which they are compared were obtained by giving 
 
202 
 
 THE PSYCHOLOGY OF LEARNING 
 
 the same test — substitution — to the children of various 
 Missouri cities. The test was the digit-symbol substi- 
 tution test, and the scores shown in the tables represent 
 the number of substitutions made per minute. 
 
 Age, 8 
 
 City 7.9 
 
 Country 6.0 
 
 Age. 
 
 City 
 
 Country 
 
 8 
 
 9.1 
 7.1 
 
 9 
 10.0 
 
 7.7 
 
 10.8 
 8.4 
 
 10 
 
 11.8 
 
 9.2 
 
 10 
 13.8 
 11.9 
 
 BOYS. 
 
 11 12 
 13.4 15.4 
 12.4 14.6 
 
 GIRLS. 
 11 12 
 15.8 18.2 
 15.2 19.8 
 
 13 
 
 16.8 
 17.3 
 
 13 
 20.3 
 21.9 
 
 14 
 19.2 
 19.8 
 
 14 
 22.2 
 23.5 
 
 15 
 22.1 
 20.5 
 
 15 
 24.1 
 24.8 
 
 16 
 
 23.7 
 21.2 
 
 16 
 26.9 
 27.6 
 
 17 
 26.4 
 22.4 
 
 17 
 28.1 
 28.6 
 
 18 
 24.4 
 25.5 
 
 18 
 28.3 
 28.3 
 
 In Figure 29 city and country children are compared 
 with reference to their standing in several mental tests, 
 substitution, as shown in the above, and also, logical 
 memory, rote memory, association, and completion. 
 The figure is constructed as follows: The horizontal 
 line at the top represents the standing of city children. 
 The graphs of the country boys and girls are drawn so 
 as to show the percentage of the city scores which the 
 country boys and girls make at the different ages. As 
 the children grow older there appears less and less 
 difference between city and country. It will be seen 
 that the country pupils make only about 60 <^ 
 a score at age 8 as do the city children. 
 
 -/c as high 
 
 ciTr CMiLCittK 
 
 /'/.f.^S 
 
 Figure 21). The graphs show the mental ability of country boys and 
 girls as compared with city children. 
 
DIFFERENCES IN LEARNING CAPACITY 203 
 
 Racial Differences in Learning Capacity. — Careful, 
 though not extensive, studies have been made of the 
 learning capacity of the American Negroes and of the 
 native Chinese. In Table 32 Chinese boys and girls 
 are compared with American boys and girls. The 
 numbers in the table represent the per cent, which the 
 Chinese scores are of the American scores for the ages 
 shown. In the substitution test alone, the average 
 efficiency of the Chinese children is, in the case of boys, 
 86.6 per cent, of American learning efficiency, and, in 
 the case of girls, 77.9 per cent. If the results of the 
 various tests are combined, it is seen that the average 
 efficiency of the Chinese is for boys, 84 per cent, and 
 for girls, 77 per cent, of that of Americans. In Table 
 33 a similar comparison is made with the Negroes. The 
 average of the norms for the Negro boys is 57 per cent, 
 of the average for whites. The average for the Negro 
 girls is 60 per cent, of the average for white girls. In 
 the substitution test alone, the average score for the 
 Negro boys is 44.5 per cent, of the white average. The 
 average for Negro girls is 43.7 per cent, of the average 
 for white girls. 
 
 It will be seen that the Chinese, although they were 
 at many disadvantages in experience, training, and 
 language, compare very favorably with American white 
 children, while the Negroes, although at no such disad- 
 vantage, show up very poorly. 
 
 The author made an extensive study of Negro ability 
 to learn, using for the purpose the marble-sorting 
 apparatus mentioned above. The comparisons arp 
 shown in Figures 30 and 31. 
 
204 
 
 THE PSYCHOLOGY OF LEARNING 
 
 
 
 
 
 1 
 
 
 
 
 
 
 "^ 
 
 
 
 
 
 
 
 
 - — 
 
 i 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 ' . 
 
 
 
 
 /^,fJO 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 r 
 
 '/i 
 
 
 
 
 
 
 
 
 ^s^s 
 
 I'iGURE 30. Frequency Surfaces showing the distribution of white 
 children and negro cliildren in ability to learn as determined by 
 the marble-sorting experiment. Solid line represents negro child- 
 ren ; broken line, white children. 
 
 Figure 31. Graphs showing the learning capacity of negroes and 
 whites, ages nine to fifteen ; upper graphs represent whites ; lower 
 graphs, negroes; broken lines, girls; solid lines, boys. 
 
DIFFERENCES IN LEARNING CAPACITY 205 
 
 TABLE 32. 
 
 A PER CENT. COMPARISON OF CHINESE WITH AMERICAN CHILDREN. 
 
 Age. 12 13 14 15 16 17 Av. 
 
 Rote memory Boys 125 118 122 116 122 109 117.0 
 
 Girls 122 114 103 102 97 112 108.3 
 
 Logical memory Boys 82.2 77.3 85.2 89.1 90.1 99.6 87.3 
 
 Girls 97.9 96.7 89.6 93.5 94.9 95.3 94.7 
 
 Substitution Boys 91.3 85.6 93.0 81.8 83.4 96.6 88.6 
 
 Girls 85.2 83.8 75.4 77.2 66.5 79.5 77.9 
 
 Analogies Boys 23.1 33.1 42.5 39.1 40.4 37.8 36.0 
 
 Girls 85.2 83.8 75.4 77.2 66.5 79.5 77.9 
 
 Spot pattern Boys 116.0 100.0 80.6 65.2 98.1 82.4 90.4 
 
 Average Boys 87.5 82.8 84.7 78.2 86.0 85.1 84.0 
 
 Girls 81.7 84.1 73.5 74.3 69.5 78.6 77.0 
 
 TABLE 33. 
 
 The numbers in the table show the percentage which Negro 
 ability is of the ability of white children: 
 
 GIRLS. 
 
 Ages 8 9 10 11 12 13 14 15 16 Av. 
 
 Logical memory, immediate 68 79 74 85 68 76 92 88 72 78.8 
 
 Logical memory, delayed 90 88 52 86 51 75 68 115 81 79.6 
 
 Rote memory, concrete 55 76 59 87 76 85 97 95 110 83.0 
 
 Rote memory, abstract 44 61 45 56 66 69 84 79 96 69.8 
 
 Substitution digit-symbol .12 36 23 38 29 50 60 43 67 43.7 
 
 Opposites 25 47 30 55 46 55 60 63 68 53.7 
 
 Genus-species 18 31 13 28 28 42 49 43 47 36.5 
 
 Part-whole 21 47 28 43 49 53 61 50 45 46.6 
 
 Word building 42 62 45 75 31 52 66 24 46 49.1 
 
 BOYS. 
 
 Logical memory, immediate 90 75 58 86 82 74 102 92 84 82.5 
 
 Logical memory, delayed. .... 65 82 95 93 107 84 89 100 75 87.7 
 
 Rote memory, concrete 57 69 58 60 74 79 86 104 102 76.5 
 
 Rote memory, abseract 34 36 42 45 62 64 67 94 96 62.7 
 
 Stubtstition digit-symbol 18 27 28 36 32 35 53 63 68 44.5 
 
 Opposites 17 21 21 37 46 47 43 64 61 44.6 
 
 Genus-species 17 16 16 23 28 21 27 32 34 25.3 
 
 Part-whole 25 33 26 32 32 45 54 50 47 41.5 
 
 Word building 7.5.. 50 66 48 45 50 21 51 77 48.1 
 
 When Negro children are compared with white 
 children v/ith reference to learning capacity alone, it 
 is found that only about seven and one-half per cent, 
 reach the median for whites, while ninety-two and one- 
 half per cent, of the whites exceed the Negro median. 
 In general mentality as shown in Table 33 about one- 
 fifth of the Negroes are equal or superior to the average 
 of the whites. Three fourths of the whites are equal 
 or superior to the average of the Negroes. 
 
206 THE PSYCHOLOGY OF LEARNING 
 
 Causes of Individual Differences. — The causes of 
 individual differences in learning capacity may be 
 grouped under two heads, (1) hereditary and (2) 
 environmental. The great influence of hereditary 
 factors is shov^n by such studies as Galton's study of 
 twins, Thorndike's study of twins, Goddard's study of 
 the causes of feeble-mindedness, Galton's studies of 
 hereditary genius, and by various studies of learning 
 capacity. 
 
 Galton found that twins that were alike remained 
 alike in spite of difference in treatment, and that twins 
 that were unlike remained unlike in spite of similarity 
 of treatment. Thorndike found the resemblance 
 among twins to be greater than the resemblance among 
 siblings. Goddard's extensive studies of feeble-mind- 
 edness show heredity to be the chief cause of mental 
 defect. In fact, Goddard's figures seem to show that 
 feeble-mindedness is Mendelian and recessive. In 
 Galton's studies, although there is always the queston 
 as to the part played by favorable family influences 
 and training, there seems no question of the prepon- 
 derant influence of heredity. In learning experiments, 
 it is found that practice does not eliminate differences. 
 In some cases, practice decreases, somewhat, individual 
 differences, but in no case does it eliminate them. All 
 careful experiments in learning show not only that 
 ability to learn is a definite characteristic, but that 
 final efficiency at any performance is definitely de- 
 pendent upon native, inherent factors, that seem as 
 definite and characteristic as are a person's height 
 and weight, or any other physical features. When a 
 group of subjects are started on a learning experiment, 
 after a short time, they assume characteristic, definite. 
 
DIFFERENCES IN LEARNING CAPACITY 207 
 
 relative positions, that remain constant except for 
 minor fluctuations, which are due to temporary causes. 
 
 In the pubUc schools, it is found that children early 
 in the grades assume a position in ability with refer- 
 ence to their fellows which remains fairly constant in 
 the later grades. For a pupil to be poor in one grade 
 and good in a later grade is the exception, and probably 
 has some clear explanation other than native ability. 
 As a rule, bright children continue to be bright and 
 dull children continue to be dull in spite of what may 
 happen to them. Special attention has now been given 
 to subnormal children for many years. It is found that 
 even when they are taken early and given the best 
 training that science is able to give, they remain sub- 
 normal in as true a sense as they retain their hair-color 
 or their facial features. 
 
 One of the strongest arguments for the influence of 
 hereditary factors comes from a theoretical considera- 
 tion. Biological studies in heredity show that the 
 structure of living tissues is dependent upon hereditary 
 forces. The brain in its structural aspects falls within 
 the general category of causality, is a product of hered- 
 ity. The differences in the nervous development of 
 different classes and species of animals are clearly 
 hereditary, so also are the smaller differences within 
 the human race. No one would probably deny that the 
 tremendous difference between the idiot and the genius 
 is due primarily to differences in brain structure. 
 There seems no doubt that smaller mental differences 
 are also due chiefly to differences in hereditary brain 
 structure. 
 
 An important fact pointing to the influence of here- 
 dity comes from studies concerned with training in 
 
208 THE PSYCHOLOGY OF LEARNING 
 
 mental functions. It seems that training has no very- 
 great influence on simple functions, such as reaction 
 time, mental span, speed of association, sensory dis- 
 crimination. The ordinary experiences of life bring 
 these and other simple mental functions to approxi- 
 mately their full possibilities. Much improvement is 
 possible in the case of complex mental functions, but 
 such improvements are not due to fundamental 
 changes in native ability but to tricks and schemes of 
 method and procedure; in most cases they are due to 
 the establishing and perfection of some bond or other. 
 Environmental Influences. — The relative importance 
 of heredity and environment in the life of an individual 
 is a question that has been much discussed, and on 
 which people have very different opinions. There is 
 really no room for difference of opinion. An appeal to 
 the facts shows that heredity and environment are 
 complementary, that each makes its own contribution, 
 and which the other can not make. The bodies that we 
 have, with their bones and muscles and nerves, come 
 from our ancestors ; they are the gift of the past. We 
 are tall or short, heavy or thin, light or dark, because 
 of heredity. Our nervous systems on which all educa- 
 tional influences must work, are the gifts of heredity. 
 The value and efficiencies of these nervous systems 
 range all the way from near zero to those having almost 
 infinite possibilities. Heredity gives us the raw 
 material on which educational influences work. This 
 raw material has its possibilities and its limitations. 
 Education cannot transcend these limitations. But 
 while social influences are limited to the material which 
 heredity gives, they are of very great importance. 
 What this importance is, we must now see. 
 
DIFFERENCES IN LEARNING CAPACITY 209 
 
 The statistical studies of Cattell show well the im- 
 portance of environment. He found that the Southern 
 states have contributed very few men of science as 
 compared to the New England states. The man of the 
 South is very much the same sort of man as his brother 
 of the North. He does not go into science because the 
 influences do not urge him in that direction. He be- 
 comes a land owner, a man of affairs. In New England 
 are our greatest universities. The influences there are 
 such as to invite to an educational career, to science, to 
 letters. Nature determines what is possible for us to 
 do. The kind of work in which we actually engage is 
 largely determined by the influences which surround us. 
 
 The relative importance of heredity and environment 
 is well illustrated by musical ability. Suppose that 
 somewhere in the "backwoods'* of the mountains a 
 child is born having in it the possibilities of a great 
 pianist. If the circumstances of life are such that this 
 child never even sees a piano, he will never become a 
 pianist. On the other hand, there are children who 
 can never become great pianists, although all the in- 
 fluences of wealth and science combine to make them. 
 The fact is that heredity and environment combine and 
 give us the outcome of every life. Every act of our 
 lives is the resultant of what we are and of the in- 
 fluences that work upon us. The same influences do not 
 get the same result from different people because the 
 people are different. If you lay side by side on an anvil, 
 a piece of glass and a piece of iron and strike them both 
 with equal force, the glass is crumbled to bits, the iron 
 is scarcely dented. The forces that act upon them are 
 the same, but the things are different. So it is with 
 people. The same influences that save one boy to good- 
 
210 THE PSYCHOLOGY OF LEARNING 
 
 ness and usefulness sometimes fail to save others, 
 becaitse the others are different and do not give the 
 same response to the same stimulus. In the same 
 school room the same educational influences work upon 
 all the pupils, but some learn fast and others slowly. 
 But what we must not lose sight of is that each can 
 (earn something. Education can make each different 
 from v/hat he would otherwise have been. It can not 
 make them all alike, it can not make them equally 
 efficient, but it can make each more efficient than he 
 would otherwise have been. It is well that we recog- 
 nise the truth; it is well that we know the facts and 
 face them squarely. Social reformers often make the 
 mistake of assuming that the same causes will always 
 produce the same effect. The same effect is produced 
 only when the same causes act upon the same or equal 
 things. And nothing in the world is more unequal 
 than human beings. On the one hand, we have the 
 idiot that can scarcely be taught to feed itself; on the 
 other, is a Newton, or a Shakespeare. 
 
 The School and Individual Differences. — Both scien- 
 tific studies and common observation show us that there 
 are great individual differences in learning capacity. 
 The practical question is : What are we to do about it ? 
 The school can not ignore them. It must take them 
 into account. By means of scientific measurements 
 and on the basis of actual achievements in the school, 
 the children should be divided into classes according to 
 their abihty to learn. A child should be put into a class 
 with other children having about the same ability. 
 When this is done, it is possible for us to do what is 
 best for each child. Some can go fast; others more 
 slowly. Not only will the different groups have differ- 
 
DIFFERENCES IN LEARNING CAPACITY 211 
 
 ent learning capacity, but they should learn, to some 
 extent, different things, in preparation for filling 
 widely different places in the world. The distinctly 
 sub-normal children should have a very different course 
 from that pursued by the great majority of normal 
 children. To make this possible, they should be separ- 
 ated from the other children. Their education should 
 be almost wholly manual. The unusually bright 
 children should also be put into classes by themselves 
 to make it possible for us to do the right thing for them. 
 They are to be the leaders of the coming generation. 
 In every case, we are to do for each child what is best 
 for each. The dull child has its claim upon us as well 
 as does the bright. We put them into different classes 
 only to enable us to do what is best for them. 
 
 EXPERIMENTS AND EXERCISES. 
 1. All the experiments so far performed may be 
 used to show individual differences. Frequency sur- 
 faces can be plotted for each learning test, and for the 
 combined scores worked out in the exercises of the 
 preceding chapter. What is the range betv/een the 
 best and poorest in the several tests ? In the combined 
 results ? 
 
 REFERENCES. 
 
 Bonser, F. G., The selective significance of reasoning ahiliiif 
 tests, J. E. P., VII, 1916, 187. 
 
 Brown, W., Some experimental results in the correlation of 
 mental abilities, B. J. P., Ill, 1910, 296. 
 Mental Measurement, 1911. 
 
 Dashiell, J. F., The effects of practice upon two ynental tests, 
 3. E. P., XI, 1920, 151. 
 
 Donovan, M. E. and Thorndike, E. L., Improvement in a irrac- 
 tice exiKriment under school conditions, A. J. P., XXIV, 1913, 
 426. 
 
 Gray, C. T., A comparisGn of two types of learning by means of 
 a stibstitution test, J. E. P., IX, 1918, 143. 
 
212 THE PSYCHOLOGY OF LEARNING 
 
 Kiug, I. and Homan, T. B., Logical memory and school grades, 
 J. E. P., IX, 1918, 262. 
 
 McCall, W. A., Correlation of some psychological and educa- 
 tional measurements, 1916. 
 
 Pyle, W. H., aucl Collings, P. E., The mental and physical develop- 
 ment of rural children. S. and S. VIII, 1918, 534. 
 
 Pyle. W. H.. A study of the mental and physical characteristics 
 of the Chinese, S. and S., VIII, 1918, 264. 
 
 The mentality of negro children S. and S. 1915, p. 357. Psych. 
 Bui. XIII, 191G, 82. 
 
 Simpson, B. R., Correlations of mental ahilities, 1912. 
 
 Stickland, G. I., The influejice of practice on the correlation of 
 ahilities, J. E. P., IX, 1918, 393. 
 
 Race, H. V., A study of a class of children of superior intelli- 
 gence, J. E. P., IX, 1918, 91. 
 
 Strong, E, K., An interesting sex difference, Ped. Sem., XXII, 
 1915, 521. 
 
 Wells, F. L. and Henmon, V. A. C, Concerning individual dif- 
 ferences in reactio?i time, P. R., XXI, 1914, 153. 
 
 Wells, F. L., The relation of practice to individual differences, 
 A. J. P., XXIII, 1912, 75. 
 
 Whitley, M. T., An iminrical study of certain tests for indi- 
 vidual differences. Arch, of Psych., No. 19, 1911. 
 
 Wyatt, S., The quantitative investigation of higher mental 
 lyrocesses, B. J. P., VI, 1913, 109. 
 
 Whipple, G. M., Classes for gifted children, 1919. 
 
Chapter XI. 
 TRANSFERENCE AND INTERFERENCE. 
 
 Nature of the Problem. — The problem of this chapter 
 may be stated in the following words : Will the forma- 
 tion of one habit either help or hinder in the formation 
 of another ? Will the acquiring of one bit of knowledge 
 help or hinder in acquiring other knowledge ? If I have 
 learned to operate a typewriter having a certain key- 
 board will it be either easier or more difficult to learn to 
 operate a typewriter having a different keybord? If 
 one has learned to run an automobile will it be either 
 easier or more difficult to learn to run a machine which 
 has different types of levers and pedals ? In the field of 
 knowledge, does the learning of one language make the 
 learning of a different language easier ? Will the study 
 of history make the acquisition of mathemathics easier ? 
 In general, are the results of learning narrow and 
 specific, or, are there general effects also ? 
 
 The problem raised here is of far-reaching theoretical 
 and practical significance. It involves the very founda- 
 tion principles of education, and we must face it at 
 every turn in practice. It involves our fundamental 
 conception of the nature of mind. Is the mind an entity, 
 a unity, of such a nature that it gains strength by exer- 
 cise ? If so, then the effects of all learning are general, 
 and it will make little if any difference v/hat the mind 
 does, for whatever it does gives it strength to do other 
 things. One can train the muscles of his arms by one 
 
 [213] 
 
214 THE PSYCHOLOGY OF LEARNING 
 
 sort of exercise and his arms will then have strength 
 to do other things. I can strengthen my arms by chop- 
 ping and they will have strenght for hoeing. Can I 
 likewise train my mind by studying mathematics so 
 that it will then have power to work in other fields? 
 Our answer to these questions will color our whole 
 scheme of education. If the results of training are spe- 
 cific, then we should learn those things which we most 
 need to know, without any reference to their general 
 effects. If the results of training are general, then we 
 should pick out as the studies for our curriculum those 
 branches which are best for the exercise of the mind. 
 
 Much of the misconception and misunderstanding 
 with reference to this problem have been due to a mis- 
 conception of the nature of mind. Modern psychology 
 does not look upon the mind as an entity that functions 
 as a unity and that can gain strength as a whole in 
 the same sense that is true of a muscle. Mind is a 
 complex, mosaic, of sensation, perception, idea, feeling, 
 and depends in every process upon an almost infinitely 
 complicated nervous structure. Modem conceptions of 
 the nature of learning also put a different aspect on the 
 problem. In the case of habit-formation, we couple a 
 muscular response to a sensory stimulus. In the case 
 of ideational learning, very definite brain processes 
 which underlie the ideas are coupled together. Whether 
 the formation of one such bond has any facilitating or 
 inhibiting effect on the formation of another is an ex- 
 perimental question which can not be answered on a 
 priori grounds. Let us turn to the experiments. 
 
 Historical and Critical. — Experimentation in this field 
 lies almost v/holly in the last twenty years. One 
 would think that in twenty years a question so simple 
 
TRANSFERENCE AND INTERFERENCE 215 
 
 and straightforward as this could have been definitely 
 settled. But such is not the case. It is not definitely- 
 settled. There is not agreement among psychologists 
 themselves, not to mention those who have no scientific 
 knowledge of the question. There has been much ex- 
 perimentation. The literature is large. But very few 
 experiments have been done with sufficient thorough- 
 ness and attention to scientific detail to merit the 
 respect of an impartial investigator. As one reads the 
 experimental literature, one seldom feels, with refer- 
 ence to any experiment, that it is final, that it settles 
 that aspect of the question with which it deals. Few 
 experimenters have repeated their experiments again 
 and again, to see if every result confirmed every other. 
 Too often the article reporting the experiments is only 
 a "preliminary report." One usually searches the lite- 
 rature in vain to find a **full report" of the "main stu- 
 dy." In more than one case, an experimenter has re- 
 ported his results and given his inferences, while ano- 
 ther psychologist would claim that different inferences 
 were warranted from the results. Thorndike's infer- 
 ences have been so questioned by Judd; Winch's, by 
 Sleight. If educational psychologists are to command 
 the respect of a scientific world, they must do their 
 v/ork with such thoroughness that it will stand the tests 
 of repetition and criticism. Too often a class exper- 
 iment that is scarcely worth anything as a mere demon- 
 stration is published as having scientific value. 
 
 A Statement of Method. — Before reviewing the ex- 
 perimental literature it will be well to make some en- 
 quiry as to the type of experiment required in solving 
 our problem. We must take a group of subjects and 
 have them form a habit which we shall call "X," and 
 
216 THE PSYCHOLOGY OF LEARNING 
 
 then have them form another habit, "Y," and determine 
 whether the formation of "Y" was easier because of 
 their having previously formed habit '*X." But how 
 are we to know whether the second habit is easier of 
 formation than the first ? The only way we can be sure 
 of this point is to have a second group form the second 
 habit. The only difference between the groups must 
 be that one has formed habit ''X," and the other has 
 not. Both are to form the second habit; only one is to 
 form the first. The groups must be equal in every 
 other respect; or, if not equal, the differences must be 
 accounted for. Of course, if there is any way of com- 
 paring the difficulty of the two habits, the control group 
 is not necessary. If, for example, we know that two 
 habits are of equal difficulty, then we can. have our 
 subjects form first one, then the other. If the second 
 is formed more easily than the first, then we can say 
 that the effects of the first are favorable in the forma- 
 tion of the second. In some cases, such procedure is 
 permissible. 
 
 Much of the earlier experimentation is worthless be- 
 cause the experimenter used no control group. Some of 
 the later experiments also suffer from this defect. 
 Ebert and Meumann's extensive memory experiments 
 are worthless because of the lack of a control group. 
 These experimenters gave their subjects practice in 
 one aspect of memorising, and found improvement in 
 other aspects. Later, Dearborn repeated the experi- 
 ments and found that the control group improved about 
 as much as Meumann's group that had the practice. 
 Further explanation will make the matter of method 
 clear. Suppose there are ten aspects of memory, and 
 we wish to learn whether training in aspect number 5 
 
TRANSFERENCE AND INTERFERENCE 217 
 
 will improve all the other nine aspects. We take a 
 group of people and measure all 10 aspects of memory ; 
 we then train the group in aspect number 5 until there 
 is great improvement and then measure again in the 
 other nine aspects. Suppose we find that there is im- 
 provement in all the other nine aspects ; we can not say- 
 that it is due to the practice in number 5, because, for 
 all we know to the contrary, the group might have made 
 this improvement without the practice. It may be that 
 if we give the ten tests and then wait a few weeks and 
 give ten similar tests again, there will be considerable 
 improvement. In fact, such is usually the case. We must, 
 therefore, in an experiment of this kind, take two 
 groups and give one group the initial and final tests and 
 give the other group the same initial and final tests and 
 the special practice besides. Then whatever differences 
 in the final tests are not otherwise accounted for, may 
 be considered to be due to the practice. 
 
 Evidence from Card-sorting. — A very simple experi- 
 ment in card-sorting will serve to bring the problem of 
 transfer clearly before us. Suppose we take a card- 
 sorting apparatus which has on each side six rows of 
 five boxes each, thirty boxes on each side, sixty in all. 
 Suppose we learn a row a day until we have learned the 
 12 rows. Will the mastery of each succeeding row be 
 easier because of our having learned the rows before? 
 The answer is yes. The following record is typical : A 
 subject spends six days with one side of the box. Each 
 row has different numbers from the preceding rows. 
 Each succeeding row is learned with greater ease. 
 After waiting two weeks, the other side of the box is 
 learned. This other side has the same numbering as 
 the first side, but the numbers are differently arranged. 
 
218 THE PSYCHOLOGY OF LEARNING 
 
 The average number of cards sorted per minute for the 
 first six rows were 53, 60, 64, 67, 75, 79. For the next 
 six rows, the records were 78, 80, 80, 84, 91, 90. These 
 records would raise a very strong suspicion in any mind 
 to the effect that learning to sort cards into a row of 
 numbered boxes would make learning to sort into a row 
 of boxes with different numbers easier. We wonder 
 whether all subjects will behave as did this one. Such 
 proves to be the case. The author has repeated the ex- 
 periment with different subjects and different groups 
 and in all cases it is found that learning to sort cards 
 into one row of numbered boxes makes learning to sort 
 into another roiv ivith different numbers, easier. 
 
 The experiment just described was repeated with a 
 group of 47 students. The group, however, sorted for 
 only five days, learning a new rov/ each day. Ten sort- 
 ings a day were made in each case. The average time 
 in seconds for each of the five successive rows was as 
 follov7s: 23.2, 20.5, 19.7, 18.7, 18.2. The average time 
 for the fifth row was 27 per cent, less than for the first 
 row. The single subject reported above reached an 
 efficiency on the fifth day 41 per cent, better than was 
 reached on the first day with the first row. 
 
 Still another card-sorting experiment was performed 
 as follows : A group of four university students sorted 
 cards for fifteen days, using one side of the box contain- 
 ing 30 compartments, then fifteen days using the other 
 side of the box having the same number of compart- 
 ments. The thirty boxes on one side had the same num- 
 bers as the boxes on the other side, but the arrange- 
 ment of the numbers v/as different. In the sorting of 
 the second period, therefore, the habits of the first 
 period had to be broken up. In spite of the inhibition 
 
TRANSFERENCE AND INTERFERENCE 
 
 219 
 
 from the first set of habits, on the fifth day of the sec- 
 ond experiment, a speed was reached as great as that 
 attained in fifteen days in the first part of the experi- 
 ment. The results of this experiment are shown 
 graphically in Figure 32. 
 
 Figure 32. Learning Curves, Card-sorting Experiment. The 
 lower curve shows the results of fifteen days sorting with one 
 scheme of numbering ; the upper curve shows the results of a second 
 sorting of fifteen days, with a different numbering of the boxes. 
 
220 THE PSYCHOLOGY OF LEARNING 
 
 These simple experiments leave no doubt that the 
 experience gained from sorting cards into one row of 
 boxes makes it easier to learn to sort into another row 
 of boxes having different numbers, or having the same 
 numbers differently arranged. Two important ques- 
 tions now arise. (1) What are the causes of this im- 
 provement in ability to learn? (2) Does the improve- 
 ment in ability to learn extend to other kinds of learn- 
 ing? The second question we shall leave for later dis- 
 cussion. The first question can be partially answered 
 now. (1) Facility was gained in manipulating the cards. 
 This increased facility improved the scores from day to 
 day. Proof that increased facility is a factor comes 
 from the following experiment: Two subjects were re- 
 quired to deal a set of playing cards into a compart- 
 ment. The time became less from day to day. The 
 records for the two subjects for 15 days were as fol- 
 lows : 
 
 Sub- 
 ject. Days 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 
 
 A 17.6 14.6 11.9 10.5 9.7 9.5 8.5 7.7 7.5 6.9 5.9 5.4 5.3 4.9 4.9 
 
 B 20.0 16.8 16.2 16.0 13.2 12.9 10.3 8.8 9.2 7.7 6.3 6.2 6.0 5.9 6.9 
 
 The records are in terms of seconds required to deal 
 out the 52 cards, all into the same compartment. It will 
 be seen that the time required the 12th day was about 
 70 per cent, less than that required the first day. See 
 figure 6, page 25. 
 
 (2) A second probable factor was increased ability to 
 recognize the numbers of the cards. (3) A third factor 
 was the learning of schemes by which the locations of 
 numbers were associated and fixed. It is possible that 
 subjects learn how to hold themselves to the task and 
 resist distracting influences. One also learns how to 
 hold the body, how to use the hands to best advantage. 
 
TRANSFERENCE AND INTERFERENCE 221 
 
 how to get the cards out of the pack in the quickest 
 way. All of these factors and probably others operate 
 to reduce the time of learning the successive rows of 
 boxes. 
 
 Other Similar Experiments. — Coover and Angell find 
 that practice in card-sorting increases efficiency in 
 typewriting. Bair's experiments are identical with the 
 card-sorting experiments of the author's reported in the 
 paragraphs above, except that he used the typewriter, 
 requiring his subjects to master successive tjrpes of 
 keyboard. He changed the keys from one experiment 
 to another by putting caps on them. Six labeled keys 
 constituted a series, and 20 series were learned. The 
 average number of seconds required by four subjects 
 for the first series was 73.4. For the twentieth series 
 the time was 57.5, or 21 per cent. less. 
 
 Experiments in Memorising. — ^We shall pass over the 
 early experiments of James and their repetition by 
 Peterson without comment. As is well known, James 
 took the position that memory in. the sense of retention 
 could not be improved. The most extensive experi- 
 ments on the transfer of memory training have been 
 performed by Fracker, Winch, and Sleight. The exten- 
 sive experiments of Ebert and Meumann were repeated 
 by Dearborn. As was mentioned above, Ebert and 
 Meumann did not use a control group. Dearborn gave 
 a group the initial and final tests to determine how 
 much of the improvement which the German experi- 
 menters attributed to the practice was due to improve- 
 ment from the initial to the final test. He found that 
 the improvement of the final test over the first was 
 about as much as the improvement of Ebert and Meu- 
 mann^s training group. In some cases, the improve- 
 ment was more. 
 
222 THE PSYCHOLOGY OF LEARNING 
 
 Fracker trained a group in memorising the order of 
 four different intensities of the same tuning fork, and 
 determined the spread of improvement to similar per- 
 formances such as memory for four grays, and to dis- 
 similar performances, such as mem^ory for poetry, geo- 
 metrical figures, etc. In the similar memorising, the 
 improvement was 16 per cent, more than the control 
 group. In the dissimilar memories, the improvement 
 was only 3.75 per cent, more than the control group. 
 
 Winch performed two series of experiments on Lon- 
 don school children. One group was trained in memo- 
 rising poetry, and the transfer to memory for prose 
 material in geography and history was determined. In 
 one case the improvement was 10 per cent, more than 
 the control group, and in another, 5 per cent. In a sec- 
 ond series of experiments the training group was prac- 
 ticed in the rote memorising of meaningful and also of 
 meaningless material. The transfer of the training to 
 substance memory was determined. The transfer was 
 very slight; the residual gain over the control group 
 was about 3.3 per cent, on the initial ability. 
 
 The most extensive and careful experiments on the 
 transfer of memory training are those of Sleight. He 
 trained one group in memorising poetry, another in 
 memorising tables, and another in memorising prose. 
 He gave initial and final tests in many different aspects 
 of memorising. His results show that training in one 
 aspect of memorising gives little or no increased effi- 
 ciency in other types of memorising. Sleight conducted 
 two independent experiments, one with school children 
 and one with two classes of v/omen students 18 to 19 
 years old. The results obtained from the latter are 
 shown in Table 34. The column to the left shows the 
 
TRANSFERENCE AND INTERFERENCE 223 
 
 aspect of memory tested. Group 1 was the unpracticed 
 group; group 2 was practiced in poetry; group 3, in 
 learning tables; group 4, in learning the substance of 
 prose. The figures indicate the improvement divided by 
 the standard deviation. The practice covered a period 
 of twelve days, one-half hour a day. A study of the table 
 shows that there is no general improvement of mem- 
 ory. In nine cases the unpracticed group improved more 
 in the aspects of memory tested than did the practiced 
 group. In three cases the practiced group made a poorer 
 record in the final tests than they had made in the ini- 
 tial tests. 
 
 TABLE 34. 
 Aspect of Memory 
 
 Tested. Group. Gain. 
 
 Dates Group 1 Unpracticed 3 
 
 Group 2 Practiced in poetry S5 
 
 Group 3 Practiced in tables _ 63 
 
 Group 4 Practiced in prose substance... — 1 
 
 Nonsense syllables Group 1 Unpracticed - 66 
 
 Group 2 Poetry 100 
 
 Group 3 Tables 75 
 
 Group 4 Prose substance 4 
 
 Poetry Group 1 Unpracticed 14 
 
 Group 2 Poetry 47 
 
 Group 3 Tables —12 
 
 Group 4 Prose substance 7 
 
 Prose, literal Group 1 Unpracticed 35 
 
 Group 2 Poetry 43 
 
 Group 3 Tables —1 
 
 Group 4 Prose substance 18 
 
 Prose, substance Group 1 Unpracticed 16 
 
 Group 2 Poetry 8 
 
 Group 3 Tables 65 
 
 Group 4 Prose Substance 68 
 
 Letters Group 1 Unpracticed 34 
 
 Group 2 Poetry 9 
 
 Group 3 Tables 30 
 
 Group 4 Prose substance 7 
 
 A study of Table 34 shows that practice in learning 
 poetry and tables gives facility in learning dates but 
 
224 THE PSYCHOLOGY OF LEARNING 
 
 not in learning prose substance. Practice in learning 
 poetry gives more facility in learning nonsense syllables, 
 but very little more in learning tables than the unprac- 
 ticed group gained, and not nearly so much gain in 
 learning prose substance as was gained by the unprac- 
 ticed group. These examples are sufficient to enable the 
 reader to interpret the table. 
 
 Interpretation of Memory Experiments.— -The mem- 
 ory experiments are essentially the same as the card- 
 sorting experiments first discussed. In card-sorting, the 
 subjects spent a certain period of time in fixing the 
 association between a certain series of numbers and 
 their corresponding boxes. They then built up another 
 series of connections between other numbers and other 
 boxes. We found that the first experience facilitated 
 the later habit-formation. In the memory experiments, 
 the subjects were given practice in building up the 
 bonds necessary to enable them to recite, verbatim, 
 poetry. It was found that this experience gave in- 
 creased facility in learning dates and nonsense syllables, 
 slightly increased facility in learning prose verbatim, 
 but none in learning prose substance or in learning let- 
 ters. Practice in learning tables gave increased facility 
 in learning dates and prose substance, slight increase 
 in learning nonsense syllables, but none in learning 
 poetry or prose verbatim or letters. Practice in learn- 
 ing prose substance did not give increased facility in 
 learning any of the other types of material. All the 
 memory experiments are in substantial agreement. 
 They show that experience in one type of learning may 
 either facilitate or hinder another type. 
 
 Discriminative Judgments. — The experiments to be 
 reported under this head are quite different from those 
 
TRANSFERENCE AND INTERFERENCE 225 
 
 above discussed. They consisted in giving subjects prac- 
 tice in making a certain type of judgments, and then 
 determining whether this practice facihtated or hin- 
 dered the making of other types of judgements. In one 
 type of experiment the judgment of experimenters is 
 unanimous. In reaction experiments, practice in react- 
 ing to stimuli in one sense department gives facility in 
 reaction to stimuli from other sense departments. Gil- 
 bert and Fracker, for example, found that simple and 
 choice reactions to sound stimuli reduced the reaction 
 time to touch and visual stimuli. He did not, however, 
 use a control group. Coover and Angell did in similar 
 experiments, and obtained a similar result. 
 
 Thomdike and Woodworth gave subjects practice in 
 estimating various types of magnitude, such as lines, 
 and surfaces, and weights. They found the effects of 
 practice very narrow. Training in the estimation of 
 the lengths of lines gave facility in estimating similar 
 lengths but not in the case of lines very much longer 
 or shorter. Judd however finds that practice in judg- 
 ments of location of lines in one position facilitated 
 similar judgments when the lines were in different posi- 
 tions, and that the influence of practice in making judg- 
 ments concerning the Mliller-Lyer illusion figure, was 
 effective when judgments were made with the figure 
 under different conditions. The situation with refer- 
 ence to this type of experiment is not so clear as in the 
 case of the other experiments so far considered. There 
 has not been sufficient repetition of experiments in the 
 field of discriminative judgments to settle the question 
 of the extent of transfer. 
 
 Experiments in Cross Education. — Experimenters 
 are in agreement here. We need not go into a detailed 
 
226 THE PSYCHOLOGY OF LEARNING 
 
 description of the experiments. A general statement 
 will suffice. The most obvious fact here is that after 
 we have learned to write with our right hand, we can 
 write with some degree of legibility with the left hand. 
 Further illustrations are as follows : Practice with the 
 right hand in tapping improves the left. Practice in 
 touch discrimation of one side increases the sensi- 
 tivity on the other. Experiments with one eye in cer- 
 tain experiments with perspective are effective on the 
 other eye. In general, practice which directly affects 
 one side of the body has an indirect influence on the 
 other side. 
 
 Transfer of Knowledge. — A few typical experiments 
 will show the results of the transfer of knowledge to 
 the mastery of a new situation. Judd tested subjects in 
 shooting at a target under water. The subjects who 
 knew the physical principle of refraction involved were 
 able more readily to adjust themselves to a change in 
 the depth of the target under water than was the case 
 with those subjects who did not know this principle. 
 
 Hyde and Leuba found that practice in reading Ger- 
 man script made the mastery of writing it somewhat 
 easier, although this effect was limited. The author 
 performed a similar experiment in card-sorting. If one 
 subject sorts cards for two days, while another sub- 
 ject instead of sorting the first day, merely studies the 
 location of the boxes and then sorts on the second day, 
 the result is that the person who studied the first day 
 makes a better score the next day than would have been 
 the case without the study. However, the score is not 
 so good as if the person instead of studying the first 
 day, had begun at once to sort the cards. It is clear 
 that knowledge helps in habit-formation, but nothing 
 can fully take the place of direct, actual practice. 
 
TRANSFERENCE AND INTERFERENCE 227 
 
 There have been many other experiments on transfer, 
 that of Ruger with puzzles, and those of Bagley and 
 Ruediger in neatness, as well as others which need not 
 be described here. We have yet, however, to describe 
 the most extensive experiment of all, most extensive in 
 point of time and numbers involved. 
 
 Rugg's Experiment. — Rugg gave a group of 326 sub- 
 jects a semester of training in descriptive geometry. 
 He compared them with a control group of 87 subjects 
 who did not have the training in descriptive geometry. 
 The 326 subjects were freshman engineers. The con- 
 trol group was made up of 72 students in the school of 
 education, and 15 in the college of engineering. It is 
 unfortunate that the control group was not similar to 
 the training group. The control and practice groups 
 were given initial and final tests in non-geometrical 
 material, quasi-geometrical material, and geometrical 
 material. The non-geometrical material consisted in 
 mental division, two tests, and in making words using 
 the letters in the word M-A-T-E-R-I-A-L. The quasi- 
 geometrical test was a test in straight-line lettering. 
 There were two geometrical tests, one which Rugg calls 
 the painted cube test and the other a test in imaging 
 objects and determining how many lines would be neces- 
 sary to construct them in space. The painted cube test 
 was this: A three-inch cube is painted on all sides, 
 how many one-inch cubes have paint on three sides? 
 On tv/o sides? On one side? On no side? 
 
 Results.^ — In one of the division tests, the practice 
 group lacked .9 per cent, doing as well in the final ex- 
 periment as did the control group. In the other division 
 test, the residual gain of the practice group over the 
 control group was 15.78 per cent. The average residual 
 
228 THE PSYCHOLOGY OF LEARNING 
 
 gain in these two experiments was 7.44 per cent. In the 
 word-building test, the residual gain of the practice 
 group was 13.37 per cent. As this test was given, it 
 was subject to direct influence from the practice in 
 descriptive geometry. Holding the word and the let- 
 ters in mind would be easier by virtue of the practice in 
 visualising in descriptive geometry. The improvement 
 in the quasi-geometrical test was 20.4 per cent. ; while 
 the improvement in geometrical material was 31.25 per 
 cent. In each case, the gain is the residual gain over 
 the control group. These, then, are the residual gains 
 for the three types of test, — 7.44 per cent., 20.4 per 
 cent., and 31.25 per cent. Or, if we average the word- 
 building test in with the two division tests, we have for 
 the non-geometrical tests, a residual gain of 9.42 per 
 cent. 
 
 This criticism should be passed on the experiment. 
 While the division tests are non-geometrical, they are 
 mathematical and the engineers would have more train- 
 ing in mathematical calculations during the semester 
 than would the 72 education students. All the tests are 
 too closely allied to the practice in descriptive geometry. 
 Why did not Rugg test his two groups in something 
 further removed from the practice, say in learning the 
 English equivalents of Latin words or Chinese words. 
 As the experiment stands, however, it falls in line with 
 the other experiments already reported. Some transfer 
 is shown. While Rugg tries to account for all the vari- 
 ous factors that are likely to influence the results of his 
 experiment, the impartial reader is likely to feel that 
 the statistical method after all can not solve this par- 
 ticular problem. The ultimate solution will probably 
 
TRANSFERENCE AND INTERFERENCE 229 
 
 come from extensive studies of a few subjects, with 
 every factor under control or accounted for. 
 
 Starch and Hewins have also used regular school 
 room studies in investigating the question of transfer. 
 Starch uses eight students in his practice group and 
 seven in the control group. He gave the former 14 
 days practice in mental multiplication and determined 
 the effect of the practice on other mathematical opera- 
 tions and on immediate memory span. The drill group 
 made a residual gain of 29 per cent, in the arithmetical 
 operations but practically no gain in immediate memory 
 span. 
 
 Summary of the Evidence. — We have now reviewed 
 the results of the more important studies. We have not 
 discussed several of the minor studies, but reference to 
 all is made in the bibliography at the end of the chapter. 
 What is the outcome? From all the important studies 
 we get evidence that an experience may affect a later 
 experience. What we do today determines and limits in 
 some measure what we can do tomorrow. The forma- 
 tion of habit X today can facilitate the formation of 
 habit Y tomorrow, while it may interfere with the for- 
 mation of habit Z. Knowledge which I acquire today 
 may facilitate what I undertake to do tomorrow, but 
 it may also hinder by giving an attitude or a mode of 
 attack not applicable. Into the interpretation or mas- 
 tery of a new experience, we carry our old experience. 
 Some aspects of the old experience will be available in 
 mastering the new; other aspects will not, and may 
 actually hinder in mastering the new. 
 
 Interpretation and Explanation of the Results. — The 
 following factors probably explain most of the phe- 
 nomena of transfer: 
 
230 THE PSYCHOLOGY OF LEARNING 
 
 (1) Identical elements, (a) In habit-formation. If 
 habit X is formed, and afterward, a complex habit Y is 
 formed — a habit in which X is a constituent part — 
 then Y is more easily formed because of the previous 
 formation of habit X. To illustrate: Addition is a 
 part of multiplication. Multiplication is more easily 
 mastered if addition is previously mastered. In general, 
 mastery of any process will facilitate the mastery of 
 any other process in which the one first mastered is a 
 constituent part, (b) In knowledge-getting. What is 
 true in habit-formation is also true in knowledge-get- 
 ting. Acquiring one bit of knowledge facilitates the 
 acquisition of other related knowledge. The mastery 
 of zoology is easier after the mastery of botany, for 
 many facts are common to the two sciences, and their 
 methods are largely the same, the tools of study and 
 investigation, largely the same. For example, one 
 learns to use the microscope in one study and this skill 
 is profitable in the other. One learns many facts about 
 cells, development, and heredity in the one that con- 
 tribute to an understanding of the other. Similarly, 
 mastery of Latin assists in the study of French because 
 of common elements in the vocabularies. Higher math- 
 ematics involves algebra ; physics and chemistry involve 
 mathematics. In general, any study will be more easily 
 mastered if the learner has previously acquired knowl- 
 edge that contributes to an understanding or explana- 
 tion of the principles of this study, or a method that can 
 be successfully used in it. 
 
 (2) Attitudes, and Methods of Attack. — In learning 
 one thing, a person not only forms a definite habit or 
 gets definite knowledge, but gets a general scheme of 
 attack, a point of view or a method that will be carried 
 
TRANSFERENCE AND INTERFERENCE 231 
 
 into other situations. In sorting cards into one row of 
 boxes, one not only fonns the definite bonds involved, 
 but forms schemes of association which assist the mem- 
 ory and are available in learning to sort into a different 
 row of boxes. In mastering any study, one not only 
 masters the content of that study, but learns a certain 
 method of mastery, and this method he carries into the 
 mastery of another study. Some aspects of the method 
 may be applicable in that other study, some may not. 
 In solving puzzles, as Ruger found, subjects learn cer- 
 tain principles, certain schemes of solving puzzles which 
 are applied to the solution of new puzzles. They may 
 work, they may not, it depends on the puzzle. In 
 science one learns always to look for causes. One comes 
 to take the attitude that every phenomenon has a cause. 
 This attitude of looking for causes, the scientist carries 
 into the solution of new problems. The mathematician 
 forms the general attitude of trying to get a quantita- 
 tive statement to apply to every fact or condition. He 
 always wants to get some sort of graph or curve to see 
 what light it may throw on the nature of his data. The 
 lawyer, the doctor, the minister, form attitudes peculiar 
 to their profession. Out of most studies, there come, 
 therefore, not only specific content of habit and knowl- 
 edge, but by-products of method, attitude, and definite 
 schemes of orientation. 
 
 (3) Ideals. Related to attitudes are ideals of accu- 
 racy, ideals of thoroughness, ideals of intellectual hon- 
 esty, ideals of perseverance, ideals of doing one's best 
 under all circumstances, and these ideals may carry 
 over into new experiences. Of course, ideals may have 
 all degrees of generality, but whatever degree of gener- 
 ality they have, they carry over from one experience to 
 
232 THE PSYCHOLOGY OF LEARNING 
 
 others. Honesty, for example, can be general or can 
 apply only to certain types of situation. It is important 
 to know that it can be general, and usually is general. 
 A human being is very much of a piece, and general 
 principles color our whole life. It may be, as Bagley 
 holds, that ideals of neatness may be either specific or 
 general, but it is important to know that they can have 
 generality. All ideals have some degree of generality. 
 Honesty and truthfulness, if we possess them at all, to 
 some extent permeate our being and enter into all our 
 acts. A person can form an ideal which will not allow 
 him to do a poor piece of work. Such an ideal then af- 
 fects all he does. He is unwilling to do anything poorly. 
 On the other hand, one can have such an ideal which 
 applies only to certain work. In doing other work, he 
 will not care. But we do get ideals and we do carry 
 them over to other work. 
 
 (4) Confidence. Successful mastery of a problem or 
 a process gives confidence in attacking similar prob- 
 lems, and this confidence is no small factor in leading to 
 success with these problems. It enables one to put 
 forth all his energy and to persist till success comes. 
 This factor is of unusual importance in education. For 
 example, if a child is fortunate in his early experience 
 with mathmetics, he succeeds. He likes it because he 
 succeeds. He studies mathemathics more and more be- 
 cause he likes it. The more he studies it, the more 
 power he has in it. On the other hand if he is unfor- 
 tunate, he does not like mathematics, he has not faith 
 in his ability in it. As time goes by, he has less and 
 less power in it because of neglect of it. And in general, 
 success or failure has a great effect on our lives because 
 
TRANSFERENCE AND INTERFERENCE 233 
 
 of the effects they have on our future efforts. The man 
 who succeeds, has more and more confidence in his abili- 
 ties. He comes to attack all problems with great confi- 
 dence, and this attitude has a great deal to do with 
 continued success. While failure makes one distrust 
 his powers, he attacks problems with a faint heart and 
 little courage, and is doomed to failure. Success breeds 
 success, and failure breeds failure. 
 
 (5) Attention. There is no question that certain 
 aspects of what we call attention can be trained and 
 have some degree of generality. In the first place one 
 can learn to stick to a task. One can learn in studying 
 Latin, for example, to set apart certain hours for study, 
 to take precautions not to be interrupted, to keep in 
 good condition for study, etc. In taking up another 
 study, all these habits, ideals and attitudes will be help- 
 ful. There is a certain training in what may be called 
 concentration, that is not a myth, but a very real thing. 
 One can learn to gather himself together and devote 
 himself to the task in hand, and this attitude of atten- 
 tion will be helpful in every thing which he undertakes. 
 The evidence seems pretty clear that, at least in the 
 early life of children, we can somewhat extend the range 
 of attention. One can learn through practice in so-called 
 mental arithmetic, to hold things in mind and manipu- 
 late them. Such training has, to some extent, general 
 effects. The work of Aiken and Dallenbach seems to 
 make it evident that we can, to some extent, improve 
 certain aspects of attention, and that this improvement 
 has some degree of permanence and generality. We 
 can, then, be trained to stick to a task, to work with 
 all our power while we are at it, and to hold facts in 
 
234 THE PSYCHOLOGY OF LEARNING 
 
 mind for mental manipulation. The influence of this 
 training is not wholly narrow and specific. 
 
 Other factors of transfer have been given by various 
 writers, but they are merely other names for the fac- 
 tors which we have enumerated. The fact is, all these 
 factors might well be called, as done by Thomdike, 
 identical elements. Habits, knowledge, ideals, and atti- 
 tudes that result from our experience are carried into 
 new experiences, and color and affect those new experi- 
 ences. It could not be otherwise. We have nothing to 
 carry into a new experience but our old experience. But 
 it will be a mistake if we suppose that our old experi- 
 ence will always be helpful, will always be adequate. 
 We shall make a mistake if we assume that certain 
 kinds of training will give the mind strength in general 
 and make it able to cope with all kinds of situations. We 
 form certain habits and acquire certain knowledge. 
 This is all we have to help us in a new situation. Some 
 knowledge and some habits will be much more helpful in 
 a given situation than other habits and other knowledge. 
 Furthermore, a habit is a specific thing. A certain 
 stimulus touches off a definite response. The new situa- 
 tion must have something corresponding to the definite 
 stimulus or the response will not come. A certain bit 
 of knowledge might help in a given situation, but it 
 v/ill be of no avail unless something in the new situation 
 brings the bit of knowledge to mind!. We sometimes 
 speak of the mind having power and of increasing that 
 power. But the only sense in which the mind can have 
 power is in the person's possessing available habits and 
 knowledge and methods applicable to the situation in 
 question, and in the habits and knowledge being 
 brought to bear through properly organised association. 
 
TRANSFERENCE AND INTERFERENCE 235 
 
 There is nothing mythical or mysterious about the 
 whole matter. Either all experience transfers or none 
 of it does ; it all depends on how we look at the matter. 
 A given muscular response must have its adequate 
 stimulus or it does not come. A needed idea does not 
 come to consciousness except by some organised asso- 
 ciative route. The mind has power if it has available 
 and usable habits, knowledge, and attitudes. 
 
 Generalised Experience. — The discussion of the pre- 
 ceding paragraph leads us to consider a matter clearly 
 presented by Judd, namely the importance of generalis- 
 ing our experience. Whether and to what extent habits, 
 knowledge, ideals, and attitudes function in a new ex- 
 perience depend to a large extent upon their organisa- 
 tion. In the various branches of study, one gets various 
 facts about the world. These facts are abstract and 
 isolated. To be useful in the future they must have an 
 interorganisation that will make them available when 
 they can be useful. One of the great dangers of school 
 and college instruction is abstraction and isolation. We 
 too often learn facts out of their natural setting, and 
 when such setting comes, the facts do not come to mind 
 because it has never really been coupled to that situa- 
 tion. It is the duty of teachers and parents to assist 
 children in learning things in as nearly as possible their 
 natural setting and in making helpful and profitable 
 associative connections, so that the knowledge will be 
 usefully available in the future. Book-learning, as com- 
 pared to learning from actual life-situations, is poor. To 
 make effective a fact learned from a book, the fact must 
 be coupled to our actual experience with the world of 
 things, the world of forces, the world of people. It must 
 be extended in its connections to all imagined or re- 
 
236 THE PSYCHOLOGY OF LEARNING 
 
 membered situations in which it is likely to be helpful. 
 All isolated facts must be organised not only with ref- 
 erence to our experience, but with reference to other 
 facts, laws and principles, so that we can go in thought 
 from one fact to other related facts. The mere having 
 an experience is no guarantee that the experience will 
 come up in memory and help us when it might be help- 
 ful. Whether it comes or not depends on organization. 
 Every phenomenon has intricate and far-reaching re- 
 lations to other phenomena. For us to profit from our 
 experience with a phenomenon, we must know its rela- 
 tions. Truly knowing the phenomenon means knowing 
 these relations. We do not know a fact in any helpful 
 way until we have generalised it and carried the gen- 
 eralisation over to its multitude of applications. This 
 discussion carries us back to our previous discussion of 
 meaning. The important thing about a fact is its mean- 
 ing, its relation to the world of other facts. 
 
 It turns out that those psychologists who hold to the 
 specific nature of learning are right. All learning con- 
 sists in bonds. But these bonds can be of such nature 
 as to make our learning more or less generally avail- 
 able. Indeed, our solution of a new problem will de- 
 pend upon such general availability. If I can not solve 
 a problem by using remembered facts and remembered 
 methods or other methods which they suggest I am 
 not likely to solve the problem, unless by fortunate acci- 
 dent. Even the accident is likely to be the result of 
 persistent attack which previous experience has taught 
 me. 
 
 Chemistry, physics, geology, history and all other 
 subjects are useless taught merely as such. All their 
 facts must be related by me to the life I live to be of 
 
TRANSFERENCE AND INTERFERENCE 237 
 
 use to me. Briefly, we are always to ask in the face of 
 a new fact, what does it mean? What new Hght does 
 it throw on my world? What are its consequences? 
 What facts are related to it? In what generalisation 
 does it find a place? What are the uses to be made 
 of it? 
 
 Formal Discipline. — What bearing have the facts 
 which we have considered and the inferences we have 
 drawn from them on the question of formal discipline ? 
 To make our answer perfectly clear we must define for- 
 mal discipline. The doctrine of formal discipline is that 
 the mind gains strength through use and that this 
 strength is generally available in whatever the mind 
 undertakes to do. For example, one could spend many 
 years studying mathematics, through this study the 
 mind acquires strength to solve all its problems of 
 v/hatever nature. There is no evidence that formal 
 discipline in this sense has any foundation. Years spent 
 in studying mathematics would, help us to solve prob- 
 lems where mathematical knowledge would be helpful 
 and where its method would be available, but would not 
 help us to solve other problems. It is true that nearly 
 all problems have a quantitative aspect. Mathematical 
 facts and methods, therefore, have a wide application. 
 But no mathematical knowledge would enable me to 
 know whether to give a child, for a certain sickness, 
 quinine or calomel. Mathematics will help me in solv- 
 ing problems of quantity but not questions of quality. 
 Of course, in the study of mathematics, I could acquire 
 habits of application, habits of care, habits of honesty, 
 that would be very useful in other studies. But the 
 only sense in which mathematics, or any other study, 
 can give me mental power is through the habits, knowl- 
 
238 THE PSYCHOLOGY OF LEARNING 
 
 edge, ideals, and attitudes which they give me. There- 
 fore, every study must have in itself its only justifica- 
 tion. Every study must stand on its own legs. I must 
 study neither Latin nor zoology for any general myth- 
 ical powers which they are supposed to give me, but for 
 the habits, knowledge, and ideals which they do give 
 me. If Latin has anything in itself worth while which 
 I want, then I must study Latin, unless, perchance, I 
 choose to study something else which will give me 
 something which I think more worth while and want 
 more. What, then, does Latin give me if I study it? If 
 I study it long enough, it gives me the ability to read an 
 old literature in its native tongue, which means to get 
 certain ideas through other symbols than our ordinary 
 English words. Is this literature, or any literature any 
 better got through one form of symbols than through 
 another? No, in so far as literary value depends on 
 ideas, it makes no difference through what symbols we 
 get the ideas. If the value of a piece of literature de- 
 pends largely on its form, most of this value will be lost 
 in translation. The Aeneid is a story which can be told 
 in our language as well as in the dead Latin language. 
 The same is true of all literature. 
 
 What about algebra? Shall I study algebra? "Yes," 
 if I want the tools which algebra gives me. "No," if 
 I do not. To all studies, we should put the same ques- 
 tion, What do you have for me? What is the specific 
 contribution which you can make to my life? What 
 knowledge can you give me which will help me to solve 
 the problems which I shall have to solve ? What needed 
 skill will you give me? Unless there is some specific 
 contribution to be got from a study, I am not justified 
 
TRANSFERENCE AND INTERFERENCE 239 
 
 in studying it for such by-products as it may give me, 
 for there are other studies which are valuable in them- 
 selves and have the same by-products. I need not study 
 Latin, nor Greek, nor Sanscrit, nor anything else to 
 leam concentration when I can study something else 
 more valuable in itself to my life in this present day, 
 and which will serve just as well as a means of ac- 
 quiring habits of application and concentration. 
 
 Interference. — The problem to be considered here is 
 the result of trying to form two mutually interfering 
 habits at the same time. The most extensive experi- 
 ments bearing directly upon this problem have been 
 with card-sorting. Brown used playing cards, having 
 his subjects sort them into four piles according to suit. 
 They used one scheme one day and a different scheme 
 the next day, and so continued to alternate from day to 
 day. Brown concludes: "The inference to be drawn 
 from these experiments is that learning to do a thing 
 in two different ways is not detrimental. It may be 
 helpful." 
 
 I have performed extensive experiments similar to 
 Brown's. My subjects sorted 150 cards into 30 boxes. 
 The boxes were numbered from 11 to 40. The cards 
 were numbered correspondingly, five to each number. 
 One group of four subjects sorted an hour a day for 15 
 days with one arrangement of the boxes and then for 15 
 days with a different arrangement of the boxes. The 
 boxes had the same numbers in both cases, but a differ- 
 ent arrangement. Another group of subjects sorted 
 for 30 days alternating from day to day from one ar- 
 rangement to the other. The first method of forming 
 the two mutually inhibiting sets of habits was clearly 
 
240 
 
 THE PSYCHOLOGY OF LEARNING 
 
 Figure 88. Learning Curves, Card-sorting Experiment. The 
 upper curve shows the results of sorting 150 cards into 30 boxes 
 with a certain scheme of numbering, practice one hour a day for 
 15 days. The lower curve shows the results obtained from a dif- 
 ferent group of subjects who sorted for 30 days alternating one 
 arrangement with a different arrangement. The lower curve shows 
 the effects of interference. The scores are cards sorted per minute. 
 
TRANSFERENCE AND INTERFERENCE 241 
 
 the best. My experiment is essentially different from 
 Brown's. Sorting cards into four piles is a different 
 sort of thing from sorting them into 30 piles. When 
 one sorts playing cards into four piles, he very quickly 
 gets the pattern of the arrangement, and after only a 
 few minutes, sorts with great speed. But this is not 
 true when the number of bonds to be formed is great. 
 Although my four subjects using the alternating 
 method were slightly better learners than the other 
 group, the results of their method were poorer. On 
 the basis of my experiments, I should say that if one 
 has two complicated sets of mutually interfering motor 
 habits to form, the most economical procedure is to 
 form one set, then the other. 
 
 That two mutually interfering sets of habits can exist 
 side by side in the same individual has been estab- 
 lished by many investigators. In the card-sorting 
 work of Miss Howe, already mentioned in previous 
 chapters, practice with two interfering habits was car- 
 ried further than has ever been done before. Her ex- 
 periment extended over a period of seven months. She 
 practiced on one scheme till she acquired great speed, 
 then acquired speed in the other scheme. She alter- 
 nated from one scheme to the other during the same 
 hour, continuing with one scheme, however, till she 
 reached maximum speed, then turning to the other. 
 Finally, she alternated directly from a sorting with one 
 scheme to the other. At her last sitting, she made the 
 following records, which are the number of seconds re- 
 quired for sorting the 150 cards: (The first number 
 represents one scheme; the second, the other, and so 
 on) 112, 112, 112, 112, 110, 107, 116, 120, 112, 117. 
 
242 THE PSYCHOLOGY OF LEARNING 
 
 This represents the placing of the cards at the rate of 
 about two-thirds of a second to each card, which is a 
 very great speed. 
 
 An interesting question is: Is habit interference 
 greater with fast learners or with slow learners? 
 Brown found that interference was greater with fast 
 learners but that they overcame it quickly. My own 
 experiments seem to confirm Brown's conclusion, but 
 the matter is not definitely settled, and forms an inter- 
 esting problem for the future. Interference is certainly 
 very different v/ith different individuals. In the mirror 
 writing experiment, which involves the interference of 
 old habits in the formation of new, some subjects can 
 write the whole alphabet in two or three minutes, while 
 some subjects can not do it in a whole afternoon. Just 
 what causes this great difference has not yet been 
 determined. We do not yet know just what type of 
 person suffers most from interference, nor do we know 
 whether it is an advantage or a disadvantage to suffer 
 from interference. If one's old habits resist change, 
 that makes for stability; on the other hand, it works 
 against progress when progress means the breaking 
 up of old habits. 
 
 EXPERIMENTS AND EXERCISES. 
 
 1. Only a simple experiment in transfer can be un- 
 dertaken as a matter of demonstration and illustration. 
 If the following experiment has not already been done 
 it should nov/ be done. Use five rows of the card-sort- 
 ing boxes. Make two sortings for each row. This can 
 be done in an hour. Note that the scores for each suc- 
 ceeding row are better than the preceding scores. How 
 many causes for the improved scores can be discover- 
 
TRANSFERENCE AND INTERFERENCE 243 
 
 ed ? If time is available, the instructor can repeat any 
 of the shorter experiments reported in the chapter or 
 discussed in the references, but transfer experiments, 
 as a rule, are too difficult to be undertaken as a class 
 exercise. 
 
 REFERENCES. 
 
 Bagley, W. C. and Squire, C. R., ExperiTnents on transfer of 
 ideals of neatness, in Bagley's Educational Values, 1905, 188, 
 
 Bair, J. H., The practice curve. P. R. INIon. Sup., No. 19, 1902. 
 
 Bergstrom, J. A., Experiments upon physiological memory hy 
 means of the interference of associations. A. J. P., V, 1893, 356. 
 
 The relation of the interference to the practice effects of an 
 association, A. J. P., VI, 1894, 433. 
 
 Brown, W. Habit interference in card-sorting, Univ. of Cali- 
 fornia Publications in Psychology, Vol. 1, No. 4, 1914. 
 
 Coover, J. E. and Angeil, F., General practice effect of special 
 exercise, A. J. P., XVIII, 1907, 328. 
 
 Cowan, E. A., An experiment on the influence of training on 
 memory, J. E. P., VII, 1916, 31. 
 
 Culler, A. J., Interference and adaptability. Arch, of Psych. 
 No. 24, 1912. 
 
 Cummins, R. A., A study of the effect of hasket-hall practice on 
 7}iotor reaction, attention a7id suggestihiUty, P. R., XXI, 1914. 
 356. 
 
 Dashiell, J. F., A comparison of complete vs. alternate methods 
 of learning tivo haUts, P. R., XXVII, 1920, 112. 
 
 Dearborn, W. F., The general effects of special practice in 
 memory, P. Bui., VI. 1909, 44. 
 
 Ebert, E. and Meumann, E., TJeher einige Grundfragen der 
 Psychologic der Vehnngsphanomene im Bereiche des Geddcht- 
 nisses. Arch. f. d. gesamte Psych., IV, 1904, 1. 
 
 Evans, J. E., The effect of distraction on reaction time with 
 special reference to practice and the transfer of training, Arch, 
 of Psych. No. 37, 1916. 
 
 Fracker, G. C, 07i the traiisference of traini7ig in memory. 
 P. R. Mon. Sup., No. 38, 1908, 56. 
 
 Gilbert, J. A. and Fracker, G. C, The effects of practice in 
 reaction and discrimi7iation and discrimination for sound upon 
 the time of reaction and discri7nination for other forms of stimuli, 
 Univ. of Iowa Studies in Psychology, I, 1897, 62. 
 
 Hewins, N. P., The doctrine of formal discipline in the light of 
 experitnental investigatio7i, 1916. 
 
 Judd, C. H., Practice and its effects on the perception of illu- 
 si07is. P. R., IX, 1902, 27. 
 
 The relation of special trairmig to ge7ieral intelligence (with 
 Scholkow), Ed. Rev., XXXVI, 1908, 28. 
 
 Kline, L. W., Some experimental evidence m regard to doctrine 
 of f07'mal discipline, J. E. P., V, 1914. 
 
244 THE PSYCHOLOGY OF LEARNING 
 
 Leuba, J. H. and Hyde, W., A71 experiment in learning to make 
 hand movements, P. R., XII, 1905, 351. 
 
 Lyans, C. K., The doctrine of formal discipline, Ped. Sem., 
 XXI, 1914, 343. 
 
 Miles, W. R., A study in the correlation of elementary and high 
 school grades, J. E. P., I, 1910, 537. 
 
 Pofifenberger, A. T. The influence of improvement in one 
 simple mental j}rocess upon other related processes, J. B. P., VI, 
 
 1915, 459. 
 
 Pyle, W. H., Transfer and interference in card-sorting, 3. E. P., 
 X, 1919, 107. 
 
 Reed, H. B., A repetition of Ehert and Meumann's practice 
 experiment in memory, J. Exp. Psych,, II, 1917, 315. 
 
 Ruediger, W. C, The indirect improvement of mental functions 
 through ideals, Ed. Rev., XXXVI, 1908, 364. 
 
 Ruger, H. A., The psychology of efficiency. Arch, of Psych. No. 
 15, 1910. 
 
 Kiigg, H. O., The experimental determination of mental dis- 
 cipl'me in school studies, 1916. 
 
 Sleight, W. G., Memory and formal discipline. B. J. P., IV, 1911. 
 386. 
 
 Starch, D., Transfer of training in arithmetical operations, 
 
 J. E. P., II, 1911, 306. 
 
 Thorndike, E. L. and Woodworth, R. S., The influence of 
 improvement in one mental function upon the efficiency of other 
 functions, P. R., VIII, 1901, 247, 384 and 553. 
 
 Wallin, J. E. W., The doctrine of formal discipline; two neglected 
 instances of transfer of training, J. E. P., I, 1910, 168. 
 
 Washburn, M. F. and McMein, M., The effect of mental type on 
 the interference of motor hahits, A. J. P., XX, 1909, 282. 
 
 Winch, W. H., The transfer of improvement in memory in 
 school children, B. J. P., Ill, 1910, 386. 
 
 Additional researches in learning to spell, J. E. P., VII, 
 
 1916, 93. 
 
 Further work on numerical accuracy in school children, 
 J. E. P., II, 1911, 262. See also same journal, 1910, 557. 
 
 The transfer of improvement in memory in school children, 
 B. J. P., II, 1908, 284. 
 
 Woodrow, H., Practice and transference in normal and feeble- 
 minded children, J. E. P., VIII, 1917, 85 and 151. 
 
Chapter XII. 
 FATIGUE AND LEARNING. 
 
 Nature of Fatigue. — The subject of fatigue has be^^T^ 
 studied by both psychologists and physiologists. Many 
 experiments have been performed. The literature of 
 the subject is voluminous. It is difficult to sift the mass 
 of results and determine what is established and what 
 is not. The results, in many cases, are conflicting and 
 inconclusive. The phenomena of fatigue are so com- 
 plex, and so combined with all sorts of other phenomena 
 that the experimenter finds it difficult to assign results 
 to their proper causes. The phenomena nearly always 
 appear in connection with practice effects, the one off- 
 setting the other. Moreover, an adequate measure of 
 fatigue, universally applicable, has not yet been dis- 
 covered. Many of the physiological phenomena due to 
 fatigue are also produced by other causes. Fatigue 
 produces inability to work, but the work curve also falls 
 because of many other causes, such as loss of interest, 
 monotony, etc. It shall be our purpose in this chapter 
 to present such facts of practical importance as seem 
 to be definitely established. And, in spite of the fact 
 that the psychology of fatigue can not be finally and 
 definitely written, there is a respectable body of facts 
 fairly well established. 
 
 The fatigue of an organ is merely a reduced capacity 
 to function brought on by work. In the functioning of 
 the various organs of the body, their tissues are torn 
 
 [245] 
 
246 THE PSYCHOLOGY OF LEARNING 
 
 down and the products of the katabolic processes are 
 eliminated by various excreting organs. When work is 
 of such amount that the waste products can not be 
 eliminated as fast as produced, they are scattered over 
 the body by the circulation and produce the effect 
 known as fatigue. This effect is essentially incapacity 
 ^^ work. That the toxic substances produced in our 
 bodies by the functioning of its organs lower their ca- 
 pacity to function is definitely known, but just how 
 they produce this incapacity is not known. It seems 
 pretty clear that it is a nervous phenomenon. The 
 fatigue poisons seem to lessen the conductivity of the 
 nerves, thereby lowering their functioning capacity. 
 Fatigue is therefore loss of ability to work, not so much 
 because of lack of energy as because the energy is not 
 available. 
 
 The recent discoveries of a German physiologist 
 seem to make conclusive the evidence that fatigue is 
 due to certain toxic substances produced in the body. 
 He found that certain antitoxins which he introduced 
 into the body of his subjects counteracted the toxins 
 produced by work. His experiments were conducted 
 with guinea pigs and school children. By injecting the 
 antitoxin into guinea pigs, he lessened fatigue in them. 
 On school children, he performed the following experi- 
 ment. A number of school children were divided into 
 groups. The children of each group were required to 
 do hard work in arithmetic. One group worked in a 
 room the air of which contained the antitoxin. The 
 pupils of this room suffered less from fatigue than the 
 children who did not breathe air containing the anti- 
 toxin. 
 
FATIGUE AND LEARNING 247 
 
 Measures of Fatigue. — Nearly every sort of mental 
 and physical performance has been used as a measure 
 of fatigue. In the earlier work some kind of physical 
 test was usually used, especially aesthesiometry and 
 dynamometry. The attempt was also made to deter- 
 mine the presence of fatigue by changes in the circu- 
 lation and respiration. While fatigue does produce defi- 
 nite changes in both circulation and respiration, these 
 changes are also produced by other causes, and are inti- 
 mately connected with the emotions. Therefore when 
 these changes occur, it is difficult to know what factor 
 or what combination of factors has produced them. In 
 practice, it is difficult to use any of the physiological 
 methods with any degree of certainty. The school 
 teacher is concerned with the fatigue of children due to 
 mental v/ork. As a measure of mental fatigue, certain 
 mental tests have proved far more successful than any 
 of the physiological methods. A fairly successful 
 method of measuring general mental fatigue is to take 
 the measure of some simple mental function after first 
 bringing the function to maximum efficiency through 
 practice. For example, we can use auditory memory 
 span for digits. We first bring this span up to its maxi- 
 mum through practice. Then we apply the test as a 
 measure of fatigue on the assumption that fatigue v/ill 
 narrow the memory span. This is typical of the more 
 successful fatigue experiments. The purpose of bring- 
 ing the function up to its maximum efficiency before 
 using a measure of its efficiency as a criterion of fatigue 
 is to prevent practice effects from offsetting the effects 
 of fatigue. 
 
 Another successful method of studying fatigue is to 
 measure the decrease in efficiency of some task continu- 
 
248 THE PSYCHOLOGY OF LEARNING 
 
 ously performed. School children can be kept at work 
 continuously on some definite school task, as an exercise 
 in arithmetic, and measures be taken of their efficiency 
 at successive intervals of time. If practice effects are 
 eliminated or accounted for, the amount of fatigue is 
 indicated by the loss in efficiency. Such a method was 
 used by Thorndike, Arai, and others. 
 
 Kinds of Fatigue. — When fatigue is occasioned by 
 mental work, it is called mental fatigue, when it is occa- 
 sioned by muscular work it is called bodily or physical 
 fatigue. Its nature is the same however caused. If 
 fatigue is very great, its effects are general ; if not very 
 great, its effects may be to some extent only local or 
 specific. Owing to the effects of loss of interest in 
 work, it is difficult to determine to what extent fatigue 
 is general. Some experiments seem to indicate that 
 in the case of mental fatigue, there is still ability to 
 do physical work. And when one is so fatigued that he 
 can not pursue the same kind of mental work profitably, 
 it is often found that he can do some other kind of 
 mental work with profit. This is probably not because 
 mental fatigue is to a very great extent specific, but 
 because it is relative. One can pursue one kind of 
 mental work till the fatigue is so great for the interest 
 in that subject that further work is not profitable, while 
 work will still be profitable in another field where inter- 
 est is fresh. This is not because there is more energy 
 for the other field but because interest makes the en- 
 ergy available. Common experience as well as experi- 
 ments seem to indicate that when mental fatigue is 
 very great, one is incapacitated for any kind of mental 
 work. On purely a priori grounds, it would seem that 
 fatigue must necessarily be rather general, for as the 
 
FATIGUE AND LEARNING 249 
 
 fatigue poisons accomulate, the circulation will scatter 
 them till they affect the entire nervous system. In- 
 capacity to work due to fatigue seems not to be because 
 of depletion of the nervous or muscular tissue but is due 
 to the effects of the poisons upon the nervous system. 
 Certainly, very great fatigue due to muscular work in- 
 capacitates one for serious mental work. Winch's ex- 
 periments with the students in evening schools would 
 seem to corroborate this view. These students were for 
 the most part clerks who during the day were engag^'d 
 in work chiefly physical. They fatigued so readily in 
 the work of the night school, that Winch drev/ the con- 
 clusion that such work was not profitable. 
 
 Fatigue of School Children. — The first important 
 practical question to raise is concerning the amount of 
 fatigue occasioned in school children by the ordinary 
 work of the day. Careful experiments have been per- 
 formed by several investigators v/hich throw light on 
 this question. The experiments of Winch, Gates, 
 Thorndike, Heck and others are fairly unanimous to the 
 effect that the regular work of the school occasions very 
 little fatigue. Winch found that children 6 and 7 years 
 old were more fatigued in the afternoon than v/ere 
 children 11 years old, and those 11 years old more than 
 those 13 years old. The extensive studies of Heck on 
 fatigue of school children led him to the following con- 
 clusion, which I give in his own words: ''Mental fa- 
 tigue in relation to the daily school program is far less 
 than is generally believed. The small amount of fatigue 
 noticeable during the school day was more probably 
 caused by improper conditions of ventilation, lighting, 
 etc., than by the school work itself. Unhygienic con- 
 ditions in the school and physical defects, however 
 
250 THE PSYCHOLOGY OF LEARNING 
 
 slight, in the children are undoubtedly the great causes 
 of fatigue in most schools. The decrease in quality of 
 work of children as the day advances, supposed to be 
 more or less general in schools, is due less to a using 
 up of the energy-producing materials in the nerve cells 
 of the body and an autopoisoning of the nervous sys- 
 tem by the waste products from this process, than to 
 a loss of interest in the school work with its lack of 
 vital and varied appeal and its monotony of instruction 
 and environment. The bored child unconsciously or 
 consciously, rebels and does a less correct amount of 
 work. Continued work produces boredom and continued 
 boredom decreases efficiency, on account of the close 
 mutual relation betw^een physiological processes and 
 mental attitudes. With sound bodies, a hygienic school, 
 proper classification, frequent relaxation, a vital and 
 varied curriculum, and live teachers, most children will 
 show no problem of fatigue in relation to the daily 
 school program. However, the individual variations 
 in fatigue in children of the same class are so great 
 that the teacher is under constant obligation to watch 
 the easily fatigued child and decrease his work when- 
 ever necessary belov/ the requirement for the class as 
 a whole." The work of Thorndike leads to a similar 
 conclusion. Under the author's direction extensive ex- 
 periments have been carried out in several city school 
 systems in three states to determine the learning capa- 
 city at the end of the day as compared with the learn- 
 ing capacity at the beginning of the school day. I did 
 not attempt to measure fatigue, but to find an answer 
 to this question: Is the capacity of school children to 
 learn in the late afternoon any less than it is at the 
 beginning of the school day ? The Whipple digit-symbol 
 
FATIGUE AND LEARNING 251 
 
 substitution test was used in these experiments. The 
 results of the various experiments were in general 
 agreement. The ability to learn just before dismission 
 in the afternoon was only about 2 per cent, less than in 
 the morning. Our results are therefore in agreement 
 with those of Heck and Thorndike. 
 
 The Course of Daily Efficiency. — Similar to the ques- 
 tion just discussed is the problem of the change of 
 working efficiency during the course of the day. But 
 this is not precisely the same question, although it is 
 related to the question of fatigue. After we awake in 
 the morning from sleep, we have recuperated from the 
 fatigue of the preceding day, but are not able to do our 
 most efficient work of the day. During sleep the func- 
 tioning of the various organs of the body is suspended 
 or partially suspended. After we awake, it takes some 
 time for them to rise to their maximum efficiency. 
 Winch, using arithmetical reasoning as a measure of 
 efficiency, studies children of both sexes and various 
 ages and finds that at 11 :30 they do on the average 5.6 
 per cent, better than at 9 :40. But in one school Winch 
 obtained different results. This school v/as in a poor 
 neighborhood. The pupils were the children of laborers 
 and got up earlier in the morning than did the children 
 of the other schools. The pupils in this school reached 
 maximum efficiency earlier than the children who got 
 up later. 
 
 The extensive studies of Gates are in agreement with 
 those of Winch. Gates* first study was with fifth and 
 sixth grade school children. He used as tests of effi- 
 ciency, addition, multiplication, auditory memory, visual 
 memory, recognition, completion, cancellation, and 
 speed and accuracy of movement. His results are re- 
 
252 THE PSYCHOLOGY OF LEARNING 
 
 produced in Figure 34. In the more purely mental func- 
 tions there was an improvement up till nearly twelve 
 o'clock, a decline at one o'clock, with some improve- 
 ment up to three o'clock. The motor tests show an 
 improvement throughout the day. There is better 
 muscular control and better speed the latter part of 
 the day. 
 
 Gates performed similar experiments with 165 college 
 students. With these students he used as tests, audi- 
 tory memory, visual memory, substitution, recognition, 
 and logical memory. The combined results of all the 
 
 TABLE 35 (FROM GATES). 
 
 Hour 8:00 9:00 10:0011:00 1:00 2:00 3:00 4:00 5:00 
 Auditory 
 memory : 
 
 Mean 100.0 97.5 98.8 100.3 97.4 94.5 98.2 95.5 93.8 
 
 Median 100.0 103.0 103.0 107.0 105.8 103.9 105.8 101.4 101.0 
 
 Visual 
 memory: 
 
 Mean 100.0 99.3 101.5 107.0 105.8 103.9 105.8 101.4 101.0 
 
 Median 100.0 100.0 103.0 101.5 96.7 100.0 104.0 101.5 100.0 
 
 Substitu- 
 tion: 
 
 Mean 100.0 102.7 105.2 104.3 96.0 102.6 101.5 101.2 94.3 
 
 Median 100.0 101.2 104.0 103.4 95.5 97.2 97.7 97.7 95.3 
 
 Recogni- 
 tion: 
 
 Mean 100.0 115.7 122.2 115.7 106.5 111.0 120.0 120.0 118.5 
 
 Median 100.0 108.0 117.0 112.0 96.4 99.0 117.0 121.0 117.0 
 
 Logical 
 memory : 
 
 Mean 100.0 109.0 107.7 103.0 95.5 99.3 101.4 102.2 91.3 
 
 Median 100.0 107.3 103.7 105.1 100.0 100.0 103.7 100.0 93.3 
 
 Average 100.0 104.3 103.6 105.6 98.7 100.6 105.1 104.2 100.4 
 
 In the above table from Gates, the eight o'clock records are 
 taken as the standards of comparison and called 100. The records 
 for the other hours are expressed in terms of per cent, of the 
 eight o'clock records. 
 
FATIGUE AND LEARNING 
 
 253 
 
 Sp.fA 
 
 Figure 34 From Gates. The graphs show the course of efficiency 
 during the day. Add.=addition ; Mult.=multiplication ; Aud.= 
 auditory memory ; Vis.=visual memory ; Recog.=recognition • 
 Compl.— completion ; Cancel.=cancellation ; Sp. and Ae.=speed and 
 accuracy. 
 
254 THE PSYCHOLOGY OF LEARNING 
 
 tests are shown in Figure 35. The results are much 
 the same as those obtained from the children. There is 
 improved efficiency till nearly noon, lowest efficiency 
 about one o'clock, an improvement till about three, a 
 slight falling off at four and still more at five. In Table 
 35 are shown in tabular form Gates' results from the 
 students. The first record of the day is called 100, and 
 the other records are expressed in terms of per cents, 
 of this initial record. It is interesting to compare his 
 results from the substitution test with those of the au- 
 thor, mentioned above, obtained from school children of 
 all ages. At 9 :25, the adults made a score of 179 digits, 
 and at 3:25 a score of 177 digits, only about 1.3 per 
 cent. less. The difference in the case of the children in 
 our study was just a trifle more, i. e., the children were 
 not quite so efficient in the late afternoon as compared 
 with morning efficiency as was the case with the older 
 students. It will be remembered that Winch found the 
 younger children showing more fatigue as the day 
 progressed than did the older children. From many 
 sources the evidence is clear that children fatigue more 
 readily than do older people. Children are not capable 
 of such long sustained effort as is the case with adults. 
 This is doubtless partly, though not wholly, due to lack 
 of habituation. 
 
 The practical significance of these studies on the 
 course of diurnal efficiency is as follows : Neither chil- 
 dren nor adults are capable of the best work early in 
 the morning. They gradually come into the full swing 
 of their power, and just before noon are able to do their 
 best work of the day. Just after noon, their powers 
 are at the lowest ebb for mental work, with some in- 
 
FATIGUE AND LEARNING 255 
 
 crease in efficiency toward the end of the school day. In 
 arranging the school work of the day, then, the most 
 difficult work should be placed at the last morning 
 period; the least difficult just after noon. It has been 
 suggested that motor work, such as drawing or writing 
 might well come in the first afternoon period. While 
 mental efficiency is lowest at this time, motor efficiency 
 is higher than at any earlier period, though not so high 
 
 Figure 35. The graph shows the course of eflScieucy during the daj 
 in the case of college students. The horizontal line represents the 
 efficiency at eight o'clock. The efficiencies for the other hours are 
 represented in percents of the eight o'clock efficiency. 
 
 as later in the afternoon. While these differences are 
 not very great, they are great enough to be taken into 
 practical account. If, for example, we wished to get the 
 maximal effort from children in a mental performance, 
 v/e should have them work from about 10:30 to 11:30. 
 If we wish to get the maximal physical effort, we should 
 have the work from about three to four o'clock. The 
 accompanying conditions must be taken into account, 
 however. The above facts are approximately true if 
 the subjects are doing the ordinary school work, but are 
 probably not true under other conditions. If the chil- 
 dren should work at a much higher level of efficiency 
 
256 THE PSYCHOLOGY OF LEARNING 
 
 than is ordinarily the case in school, the point of highest 
 efficiency would doubtless be shifted. It would probably 
 come earlier. And if instead of doing the ordinary 
 school work, they should be engaged in hard manual 
 work, the point of highest efficiency at such work 
 would doubtless come earlier than four or five o'clock. 
 
 The work of several intestigators makes it appear 
 that mental fatigue comes earlier than physical fatigue, 
 under the same conditions of mental and physical work. 
 If one works at hard mental work, fatigue comes earlier 
 than if one works at hard physical work in a field where 
 he is habituated. As mentioned above, fatigue is largely 
 a nervous phenomenon. In mental work, the nervous 
 system is much more fully involved than it is in physical 
 work. 
 
 Continuous Work. — From the laboratory of Professor 
 Thorndike have come some interesting studies of long 
 continued work of the same kind. The most notable 
 work is that of Arai. This investigator practiced her- 
 self at mental multiplication of five place numbers, and 
 then worked for a long period. The notable thing in 
 the results is that efficiency continued for so long a 
 time with little falling off. 
 
 Painter performed a similar experiment. Painter 
 first reached stability through practice, then worked 
 from 11:00 p. m. till 3:07 a. m., multiplying four place 
 numbers. Inability to work seemed to come rather 
 suddenly. He concludes: 'There exists a definite and 
 relatively abruptly appearing point beyond which men- 
 tal work becomes impossible/' It seemed to Painter 
 that there was no "tailing off." He was not only unable 
 to multiply four place numbers but was unable to do 
 
FATIGUE AND LEARNING 257 
 
 anything. Such a conclusion can not be accepted with- 
 out further confirmation. Other experiments indicate 
 that inabihty to work comes on by degrees. Memory 
 span, for example, does not suddenly jump from the 
 normal span to zero. Painter was working with an 
 habituated process. It is possible that in the case of 
 habituated processes, even though they are complex, 
 inability to work may come rather suddenly, and that 
 when one is unable to carry on such a process he is 
 practically unable to do anything. 
 
 In accepting Thorndike's conclusions in regard to our 
 ability to do long-continued work without much fatigue 
 and without much diminution in our ability to perform 
 the work, a certain distinction must be kept in mind, 
 namely, the difference between habituated and non- 
 habituated work. Fatigue comes much more quickly 
 in non-habituated work. The author once undertook 
 to learn the point alphabet of the blind. Fatigue was 
 so great that rest periods had to be allowed every fif- 
 teen minutes. When children are learning to write, or 
 trying to master any difficult feat of muscular co- 
 ordination, fatigue comes very quickly, and rest periods 
 should be frequent. In establishing new bonds in men- 
 tal learning, fatigue also comes very quickly. 
 
 For still another reason the distinction between 
 habituated and non-habituated work must be kept in 
 mind. As has been shown, one can perform a task in 
 habituated work for a long period of time with little 
 diminution in effectiveness. If one starts out to do a 
 non-habituated task, one's efficiency rises for a time and 
 then fails to rise any more during the work period, even 
 though work continue for hours. Efficiency may remain 
 at very near the same level, just as in the habituated 
 
258 THE PSYCHOLOGY OF LEARNING 
 
 performance it remains at nearly the same level. But 
 two other things besides efficiency are to be taken into 
 account. In the case of the non-habituated perform- 
 ance, although the level of work remains about the 
 same, the effect of the work in fixing the habit after 
 the first three or four half -hours is practically nothing. 
 Effective learning can not go on unless fatigue is at a 
 minimum. And in the case of habituated work, al- 
 though one can maintain the same high level of effi- 
 ciency for a long period of time, that the after effects 
 of a short period of work are the same as for a long 
 period has not been demonstrated. The evidence points 
 in the other direction. For long continued work, one 
 pays the penalty afterward, as is indicated by the work 
 of Smith which is discussed below. 
 
 Fatigue Antitoxins. — In a preceding paragraph the 
 discovery of a fatigue antitoxin was mentioned. This 
 antitoxin introduced into the circulation of an animal 
 delayed the effects of fatigue. The work of Miss Smith 
 makes it look as if the body itself produces an anti- 
 toxin which in some degree neutralises the effects of 
 fatigue. Smith withheld sleep and then noted the effect 
 on work. At first, the effect of loss of sleep seemed to 
 enhance work, but later, after a day, or two or three or 
 four days, work suffered from the loss of sleep. We can 
 not consider such a notion established, but many phe- 
 nomena seem to support Smith's inference as to the 
 bodily antitoxins. Such a theory would explain the re- 
 sults of Thomdike, Arai, and Painter, and the common 
 observation of every day life of people who work for 
 long periods at a high tension and then suffer from 
 complete exhaustion. If as soon as waste products are 
 produced in the body, the body sets to work not only to 
 
FATIGUE AND LEARNING 259 
 
 eliminate them but to neutralise them, one ought to be 
 able, under sufficient incitement to work to the point of 
 exhaustion. In such case exhaustion would be due not 
 only to an accumulation of the fatigue toxins above 
 elimination and neutralization, but also to at least a 
 partial depletion of the tissues. Our attitude toward 
 Smith's speculation must be that of waiting for fur- 
 ther facts. 
 
 Several experiments have been performed to deter- 
 mine the fluctuations in efficiency during a work period. 
 Phillips found that in the first minute of a ten minutes 
 practice period, 6 to 12 per cent, more work was done 
 in the fundamentals of arithmetic than for the average 
 of the remaining nine minutes. Poffenberger compared 
 the two halves of a very short work period. He gave 
 subjects tests in opposites, adding, color-naming, and 
 cancelling, of less than a minute in length. The first 
 half of the work was done in less time than the second 
 half. Poffenberger's inference was that fatigue had set 
 in in this short time, and thought fatigue not so rare 
 as sometimes supposed. Smith's hypothesis comes to 
 mind in this connection. It is possible that when we 
 start to work the waste products can not be immedi- 
 ately taken care of either by elimination or neutraliza- 
 tion. It doubtless takes some time for the production 
 of the bodily antitoxins, if such a thing happens. 
 Physiological processes can not be instantaneous. When 
 waste products are produced anywhere in the body, it 
 takes a measurable time for the body to establish a new 
 equilibrium. It has been demonstrated that the func- 
 tioning of the nerve fibres as well as the nerve cells re- 
 quires the presence of oxygen. When a group of neu- 
 rones begins to function, their supply of oxygen is tem- 
 
260 THE PSYCHOLOGY OF LEARNING 
 
 porarily depleted, a new balance must be established. It 
 can not be established immediately, for the stimulus for 
 the increased supply of oxygen is the need of it. This 
 explains why it takes some time for one to come into 
 the full swing of his power. When one starts a certain 
 piece of work, the part of the nei'vous system involved 
 is rested and in full functioning condition and can do 
 well immediately, but can not do well continuously until 
 a new balance of performance is established among all 
 the processes involved, elimination, neutralization, oxy- 
 gen supply, or whatever they may be. The author has 
 extensive records which show that in card-sorting, even 
 after considerable habituation is reached, the best work 
 of an hour's sitting is not done at the beginning, but 
 only after the sorting has gone on for a while. The 
 records of four subjects are shown in Figure 36. (See 
 also Fig. 8) . While there is some individual difference, 
 all subjects agree in showing a drop in the curve or in- 
 creased efficiency after the initial performance of the 
 hour. This seems to be universally true in motor work. 
 Sleep and Fatigue. — The best cure for fatigue is 
 sleep. During sleep, the body does only the work neces- 
 sary to keep life going. The waste products of work 
 are eliminated during sleep. The waste of the tissues 
 is repaired. A stable equilibrium is established. Sleep 
 is absolutely necessary. The body by one device or 
 another can keep at work for some hours, but eventu- 
 ally it must have sleep. There must be complete cessa- 
 tion from all work except what is necessary to keep the 
 vital processes going. Several authorities have pub- 
 lished statements giving the amount of sleep required 
 by people of different ages. One of the latest state- 
 ments is from Terman and Hocking. They give not the 
 time people should sleep but the average amount of 
 
FATIGUE AND LEARNING 
 
 261 
 
 rff36. 
 
 / Z J 'f^ ^ . 
 
 Figure 36. The graphs show the course of efficiency during an hour 
 after habituation in card-sorting, four different subjects. 
 
262 THE PSYCHOLOGY OF LEARNING 
 
 time they actually do sleep. Their data are from sev- 
 eral places in western United States, and are shown in 
 the accompanying table. 
 
 TABLE 30. 
 
 SHOWING THE AVERAGE TIME IN HOURS AND MINUTES SPENT IN 
 SLEEP BY PEOPLE OF VARIOUS AGES, 
 
 Age 6-7 7-8 8-9 9-10 10-11 11-12 12-13 
 
 Amount of 
 
 sleep 11:14 10:41 10:42 10:13 9:56 10:00 9:36 
 
 Univ. Stu- 
 
 Ages 13-14 14-15 15-16 16-17 17-18 18-19 dents. 
 
 Amount of 
 
 sleep 9:31 9:06 8:54 8:30 8:46 8:46 7:47 
 
 The hours of sleep shown in the table may be taken 
 as fairly representative of the amount of sleep required 
 at the different ages. Rather, they should be taken as 
 the minimum requirements, for it is doubtful that any 
 of the figures used in computing the averages would be 
 too high, and it is quite likely that in some cases, they 
 would be too low. This would make the averages a lit- 
 tle lov/er than they should be. 
 
 It seems clear that some people need more sleep than 
 do others, and doubtless the work one is doing makes 
 some difference in the amount of sleep needed. Older 
 people should determine by experiment hov/ much sleep 
 they need, and take that amount. Parents should de- 
 termine how much sleep their children need and see 
 that the children get that amount. Present social con- 
 ditions in the cities make it difficult for children to get 
 enough sleep, for there are so many attractions to keep 
 them up and awake in the evening. They consequently 
 go to bed too late to get enough sleep before they have 
 to get up to get to school on time. Teachers should 
 make a study to determine the amount of sleep taken 
 
FATIGUE AND LEARNING 263 
 
 by their pupils, and should take such steps as are neces- 
 sary to bring about the proper amount of sleep. 
 
 Terman and Hocking found no relation between 
 amount of sleep and intelligence or success in school. 
 This need not mean that sleep is not a factor in school 
 success but that other factors are of more importance 
 than the slight variation in sleep. The children gifted 
 by nature with superior endowment are able to hold 
 their high ranks in school with varying amounts of 
 sleep. The children dull by nature are not able to excel 
 by any amount of sleep. But children whether bright 
 or dull will do their best work v/hen they have plenty of 
 sleep. 
 
 Experiments have shown that the most value comes 
 from the early sleep of the night. Sleep is then deeper. 
 After only two or three hours of sleep, a subject shows 
 as great working capacity as if allov/ed to sleep the re- 
 quired amount. The experiments on which this state- 
 ment is based did not determine how long the subjects 
 could work, but only their efficiency for a short time. 
 It is quite probable that an hour or two of sleep suffices 
 for a removal of the fatigue poisons and a renev/al of 
 the oxygen supply to the neurones, but is not sufficient 
 for a renewal of the w^asted tissues. One's im.mediate 
 capacity to work doubtless depends upon the amount 
 of fatigue poisons present and on the oxygen supply in 
 the central nervous system, but one's capacity for pro- 
 longed work depends on the amount of available energy. 
 
 If one wishes to keep his body in a high state of 
 working efficiency, he should have enough rest during 
 the day and sleep at night to enable the body to repair 
 the waste of work. If one work in such a way that for 
 a considerable time the outgo of energy exceeds the re- 
 
264 THE PSYCHOLOGY OF LEARNING 
 
 pair, then he must eventually pay the penalty. In sick- 
 ness the body has to make such a fight to overcome the 
 disease that it becomes very much weakened. There is 
 often great loss of tissue shown by decrease in weight. 
 After sickness, children therefore fatigue very easily, 
 and when they return to school, their work should be 
 lightened instead of increased. 
 
 The Feeling of Fatigue, — It is of practical importance 
 to know that the feeling of fatigue is no reliable cri- 
 terion of the presence of fatigue. Fatigue may be pres- 
 ent when one does not feel fatigued; and on the other 
 hand one may feel fatigued when the body is not fa- 
 tigued. The only sure test is the ability to work. Some- 
 times one feels tired and is disinclined to work, but 
 finds that if he will start in and try, after a little time 
 he is working with great efficiency. A very interesting 
 phenomenon is that of "second breath," a matter not 
 yet fully explained. One may work for a time, fatigue 
 comes on and there is lessened capacity for work, but 
 it is often found that if one keeps on at the work, the 
 capacity increases again and efficiency goes up perhaps 
 even to a higher point than at first. Such rhythms may 
 be repeated several times in the course of continuous 
 work. Their explanation, while not fully known, doubt- 
 less is to be found in the various physiological processes 
 already discussed. An excess of waste products pro- 
 duces the temporary incapacity, and at the same time 
 stimulates the removal and renewal processes, which 
 presently restore the original capacity. These rhythms 
 keep up till there is probably temporary depletion of 
 the tissues, which only rest and sleep can cure. 
 
 Practice and Fatigue. — Several times we have called 
 attention to the relation of practice to fatigue. Ha- 
 bituation reduces fatigue. When we are learning a 
 
FATIGUE AND LEARNING 265 
 
 process fatigue is great, after we have acquired effici- 
 ency, the work can be carried on for a long period of 
 time with relatively little fatigue. These facts should be 
 borne in mind in school work, especially in that of 
 young children. In trying to master the move- 
 ments of writing, the young child fatigues very 
 easily; after the process is mastered, one can 
 write for hours with little fatigue. The same 
 is true in all work. When the organism be- 
 comes adapted through the strengthening of the tissues 
 and the establishment of the necessary neuro-muscular 
 bonds work can go on for a long period of time with 
 very much less fatigue than is occasioned earlier in the 
 work. When a process is thoroughly habituated, it par- 
 takes of the nature of a reflex, and can be carried on all 
 day at a high point of efficiency. The organs and tis- 
 sues concerned become so adapted and adjusted that 
 they can maintain the processes for long periods of 
 time. 
 
 A related phenomenon is the fact that the best work- 
 ers fatigue less easily than do the poor workers. The 
 fastest workers usually make fewer mistakes than do 
 the poor and slow workers. In the slow worker, the 
 elements involved in the work are not so well co-ordi- 
 nated; there is usually more waste of energy through 
 interfering processes. Even in the same worker there 
 is often found a direct relation between speed and 
 accuracy, fewer mistakes being made in connection 
 with the fastest speed. Some investigators have found 
 an inverse relation between improvability and fatigue, 
 those who improve the fastest suffering least from fa- 
 tigue. These several phenomena have their chief ex- 
 planation in facts of adjustment and adaptation. 
 
266 THE PSYCHOLOGY OF LEARNING 
 
 Work and Oxygen Supply.— Physiological experi- 
 ments have shown the necessity for the presence of 
 oxygen in the functioning of the neurones. An isolated 
 nerve fibre functions when oxygen is present. If it is 
 placed in some other medium, it ceases to function. 
 When oxygen is supplied, it functions again. These 
 facts show the necessity for hygienic conditions proper 
 for work. They show the importance of proper ventila- 
 tion of school rooms, of the necessity for the whole 
 breathing apparatus of the child being in perfect condi- 
 tion. The open-air school has fully confirmed this con- 
 clusion. One should also have constant pure air while 
 asleep in order to facilitate the restoration of all the 
 organs of the body to their full working power; Food, 
 in proper quality and quantity is also necessary if chil- 
 dren are to be in condition to profit to the full extent 
 from instruction. It is folly to maintain schools at 
 great expense and then go to no trouble to see that 
 children are in proper condition to profit from the work 
 of the schools. It is the business of a teacher not only 
 to hold recitations, but to know that all the conditions 
 necessary for effective work have been supplied. 
 
 The Work of Students. — University students can 
 profit from a consideration of the facts discussed in 
 this chapter. Their energies are often dissipated. They 
 too often have all sorts of activities and interests which 
 consume both their time and energy. In some cases, 
 the night is half gone before they get down to the work 
 which is supposed to be the proper work at a univer- 
 sity. But the assignments must be prepared, so they 
 work, or attempt to work, when their bodies are in no 
 condition for work. The energy of our body is like the 
 money in our purse. When it is gone, it is gone. If it 
 
FATIGUE AND LEARNING 267 
 
 is spent in one way, then it is not available for spend- 
 ing in another way. We have to decide; we have to 
 choose. If we spend all of our energy performing in 
 the side shows, there is none available for the main 
 circus. 
 
 A very serious trouble among students is that they 
 do not keep their bodies physically fit, to be able to 
 learn effectively. The body, after all, is a motor ma- 
 chine. To be maintained in a high state of effective- 
 ness, a large amount of motor work is necessary. Few 
 people break down on account of too much work, but 
 they often break down on account of too much work 
 ^or the conditions under which they live. 
 
 EXPERIMENTS AND EXERCISES. 
 
 1. Simple experiments showing the relation of fa- 
 tigue to learning can be undertaken. The instructor 
 and students should work out a plan for comparing the 
 learning capacity of students under various degrees of 
 fatigue. Nonsense syllables should be used for learning 
 material. Various plans can be devised for inducing fa- 
 tigue. 
 
 2. The course of daily efficiency can be studied after 
 the manner of Gates. Determine the learning capacity 
 for nonsense syllables for various hours of the day and 
 night. The work can be much simplified by dividing 
 the class into groups, measuring their learning capa- 
 city by identical procedure, then let different groups 
 try out learning at different hours of the day and night. 
 
 REFERENCES. 
 
 Avai, T., Mental fatigue, Columbia Univ. Cont. to Ed., No. 54, 
 1912. 
 
 Ash, I .E., Fatigue and its effect upon control. Arch, of Psych., 
 IV, No. 31, 1914. 
 
 Burnham, W. H., Oxygen supply as a condition of efficient 
 hrain activity, J. E. P., II, 1911, 421. 
 
268 THE PSYCHOLOGY OF LEARNING 
 
 Burnham, W. H., The hygiene of sleep, Ped. Sem., XXVII, 
 1920, 1. 
 
 Dodge, R., The laws of relative fatigue, P. R., XXIV, 1917, 89. 
 
 Garth, T. R., Mental fatigue. Arch, of Psych. No. 41, 1918. 
 
 Gates, A. I., Diurnal variations in mcmorii and assoeiation. Univ. 
 of Cal. Pub. in Psych., I, No. 5, 1916, 323. 
 
 Heck, W. H., A study of mental fatigue, 1913. 
 
 Holliii^worrh, H. L., Variations in ejflcicncii during the icorJx- 
 ing day,^P. R., XXI, 1914, 473. 
 
 Lee, P. S., The nature of fatigue, P. S. M., LXXVI, 182. 
 
 Marsh, H. D., The diurnal course of efficiency, 1916. 
 
 Mosso, A., Fatigue, Eng. tr., 1904. 
 
 McDougal, W., On a new method for the study of concurrent 
 me7ital operations and of mental fatigue, B. J. P., I, 1905, 435. 
 
 Offner, M. M., Mental fatigue, Eng. tr. by Whipple, 1911. 
 
 Painter, W. S., Efficiency in mental multiplication under 
 extreme fatigue, J. E. P., VII, 1916, 25. 
 
 Phillips, F. M., A comparison of the work done in the successive 
 minutes oj a ten-minute practice period in the fundamentals of 
 Arithmetic, J. E. P., VII, 1916, 271. 
 
 Poffenberger, A. T. and Tallman, G. G., VariaMUty in per- 
 formance during hrief periods of work, P. R., XXII, 1915, 371. 
 
 Root, W. T., Jr., A study of mental fatigtie with a group of five 
 hoys, J. E. P., VII, 1916, 533. 
 
 Seashore, C. E. and Kent, G. II., Periodicity and progressive 
 change in continuous mental toork, P. R. Mon. Sup., VI, No. 
 28, 1905. 
 
 Smith, M., A contribution to the study of fatigue, B. J. P., VIII, 
 
 1916, 327. 
 
 Starch, D. and Ash, I. E., The oriental work curve, P. R., XXIV, 
 
 1917, 391. 
 
 Terman, L. INI. and Hocking, A., The sleep of school children, its 
 distribution according to age and its relation to physical and 
 mental efficiency, J. E. P., IV, 1913, 138, 199 and 269. 
 Thorndike, E. L.. Mental Fatigue. P. K.. YII, 1900, 400 and 547. 
 Mental Fatigue, J. E. P., II, 1911, 61. 
 The curve of work, P. R., XIX. 1912, 165. 
 The effect of continuous exercise and of rest upon difficult 
 mental multiplication, J. E. P., V, 1914, 597. 
 
 Fatigue in a complex function, P. R., XXI, 1914, 402. 
 Wells, F. L., Practice and the ivork curve, A. J. P., XXIV, 1913, 
 35. 
 
 Normal performance in the tapping test before and during 
 practice, tvith special reference to fatigue, A. J. P., XIX, 1908, 437. 
 Wimms, .T. H., The relative effects of fatigue and practice pro- 
 duced by different kinds of mental work, B. J. P., II, 1907, 153. 
 
 Winch, W. H., Mental adaptation during the school day as 
 measiired by arithmetical reasoning, J. E. P., IV, 1913, 17. 
 
FATIGUE AND LEARNING 269 
 
 Some measurements of mental fatigue in adolescent pupils in 
 evening schools, J. E. P., I, 1910, 13 and 83. 
 
 Mental fatigue in clay school children as measured hy imme- 
 diate memory, J. B. P., Ill, 1912, 18 and 75. 
 
 Mental fatigue in day school children as measured hy arith- 
 metical reasoning, B. J. P., IV, 1911, 315. 
 
 Wright, W. R., Some effects of incentives on work and fatigue, 
 P. R., XIII, 1906, 23. 
 
 Yoakum, C. S., An experimental study of fatigue, P. R. Mon. 
 Sup., No. 46, 1909. 
 
Chapter XIII. 
 INBORN NATURE IN RELATION TO LEARNING. 
 
 Habits and knowledge are the results of established 
 bonds, instincts are the result of inherited bonds. In 
 the strict sense, an instinct is a response connected to 
 its stimulus by an inherited neural bond. This inherited 
 bond may have all degrees of definiteness, from zero on 
 the one hand to absolute definiteness on the other. It 
 is our purpose in this chapter to enquire into the rela- 
 tionship between the inherited bonds and the acquired 
 bonds, particularly into the use that can be made of 
 the instincts and emotions in the work of forming 
 habits and acquiring knowledge. 
 
 In the first place it is necessary to correct an er- 
 roneous notion in regard to the instincts. Some writers 
 state that the instincts are the basis of all our acquire- 
 ments, that every habit is merely the modification of 
 some instinct. This is not true. We are bom v/ith a 
 nervous system with some differences in resistance to 
 various types of stimuli already established. But any 
 conceivable connection of stimuli Vvith responses is pos- 
 sible. We have the nerves. They go on the one hand 
 to sense organs ; on the other, to muscles. The stimulus- 
 response couplings that are possible are practically in- 
 finite, regardless of the inherited couplings. Not only 
 are inherited bonds not the basis of all acquirements, 
 but we can overcome and reverse inherited differential 
 resistances. The inherited bonds are colorless or indif- 
 
 [270] 
 
INBORN NATURE AND LEARNING 271 
 
 ferent to most of our acquirements. For example, lan- 
 guage, saying the multiplication table, playing a piano, 
 are in no sense modifications of instincts. Some chil- 
 dren learn to call a certain animal a "dog" ; others, learn 
 to call it "ein Hund"; the ancient Pwoman called it 
 "canis." And we could just as well learn to call it any- 
 thing else, for there is no inherited bias in the matter. 
 In playing the piano, the position of the notes on the 
 staff determine what piano key we shall strike. It is 
 of course absurd to believe that there is any inherited 
 bias for striking one key for a certain note rather than 
 some other key. The part which the instincts play in 
 our learning is other than forming the basis of our 
 acquirements. 
 
 Imitation. — -Imitation is not an instinct in the strict 
 sense. If it were, then the perception of an act would, 
 without training, be an adequate stimulus for the act. 
 But such is not the case. However, the tendency to imi- 
 tate, the desire which children have to do what they 
 see others do, is an inborn human characteristic. This 
 tendency is universal in the human race, and from the 
 point of view of education, is one of the most important 
 of all human traits. The tendency to do what is seen 
 done probably has more to do in leading a child to a 
 mastery of the v/orld which immediately surrounds him 
 than any other single factor with the possible excep- 
 tion of play. One has only to observe children in the 
 home to be convinced of the importance of imitation. 
 The little girl wants to do and tries to do all that she 
 sees her mother do. She attempts to do all that she 
 sees done. The acts which she thus attempts, she can 
 not perform at first, unless these acts have already been 
 mastered in some other connection, but she tries to do 
 
 / 
 
272 THE PSYCHOLOGY OF LEARNING 
 
 them, and she keeps on trying till they are mastered. 
 As a rule, when such an act is mastered, it loses its at- 
 traction, and the child passes on to the mastery of some 
 other perceived action. The result is mastery of all 
 the mother does; and in acquiring this mastery, the 
 girl has learned a great many things about the world. 
 The case is similar with the boy, only his world is 
 larger. In mastering what he sees his father and other 
 older people do, he not only learns to do many different 
 things, but learns much about the world in all its varied 
 aspects. He learns much about forces and energy, the 
 principles of machines, etc. One can hardly overesti- 
 mate the importance which imitation has in adjusting 
 and adapting the child to the physical and social world. 
 Through imitation, children acquire their language, 
 manners and customs, traditions, morals, religion, and 
 their attitude toward nearly all important aspects of 
 life. It is doubtful whether any other factor is so im- 
 portant in developing habits of honesty, industry, per- 
 severance, and the various important ideals of life and 
 character. 
 
 Imitation in School Work.— Imitation can be used in 
 school as a method of mastering mechanical processes, 
 such as arithmetical processes, writing, drawing, and 
 music. The first step in the process of habit-formation 
 should be demonstration by the teacher. When any 
 new process is taken up, say long division, the teacher 
 should demonstrate clearly every step in the process. 
 This is a matter of economy. Such a procedure leads to 
 quick mastery. The child should not have originality in 
 formal processes. He needs merely to master them 
 and do them as older people do them, and master them 
 
INBORN NATXJRE AND LEARNING 273 
 
 in the quickest possible way. To imitate the perform- 
 ance of the teacher is the quickest way. 
 
 Imitation and Ideals, — During- the high school period, 
 which covers roughly the period of adolescence, boys 
 and girls acquire what, in most cases, prove to be their 
 final life ideals. These ideals are copied from the acts 
 of their fellows, primarily from their elders, and from 
 the acts which they read about in literature and his- 
 tory. It seems probable that the attitudes and ideals 
 which adolescents acquire are of far more importance 
 to their lives afterward than all the facts they acquire 
 during this period. If this is true, and the author be- 
 lieves it is, then the influences which the school, home 
 and all the other social forces bring to bear to develop 
 these ideals and attitudes, are of more importance than 
 the ordinary formal high school curriculum. The actual 
 amount of the content of the high school curriculum 
 that gets into the life of the boys and girls, and becomes 
 an important factor there, is doubtless small, but there 
 can be no doubt of the profound influence on their lives 
 of the moulding of character that goes on during this 
 period. The character-formation of these years is cru- 
 cial and determining for the future. All educational 
 forces that concern this period should take these facts 
 into account. In careful, scientific fashion the sources 
 of adolescent ideals should be determined and measured. 
 Of great use here is literature. The high school teacher 
 of literature has great things within her power. The 
 whole literature of the world, of all ages and all coun- 
 tries, should be studied with reference to its value in 
 bringing before adolescents'^the highest ideals of action. 
 In spite of apparent indications to the contrary, adoles- 
 cents are teachable, susceptible, easily swayed one way 
 
274 THE PSYCHOLOGY OF LEARNING 
 
 or another. The teachings of the world's great teachers 
 should be brought before them. Under the sympathetic 
 guidance of teachers and parents these forces mould 
 the character of the boy and girl. 
 
 Imitation in Adult Life,— Imitation is an important 
 factor not only in the lives of children, but in the lives 
 of men and women. Throughout adult life, imitation is 
 one of the great social forces. In childhood, as shown 
 above, we become adapted and adjusted largely through 
 the process of imitation. As long as we live imitation 
 holds us to the forms to v/hich we have become accus- 
 tomed. Of course, other factors are also at work, the 
 desire of public approval and the fear of public con- 
 demnation. These three forces are the whip which 
 makes us conform, makes each do what all the others 
 are doing. What others are wearing, I must wear; 
 what others are saying, I must say; what others are 
 doing, I must do; what others are learning, I must 
 leam; what others are thinking, I must think. The 
 press comes to the aid of conformity. We read the 
 same newspaper at breakfast, the same magazine after 
 supper, the same ''best seller" at the week-end. Educa- 
 tion itself is a great force for conformity. Education is 
 conservative ; it moulds us in the same mould. It must 
 necessarily be organised and systematised. In the 
 process there is a danger that the child will lose origi- 
 nality and initiative. Organisation and systematisa- 
 tion are but means to an end. The end is the teaching 
 of children. The machinery of education has its proper 
 function, and principals and superintendents who direct 
 the machinery have their proper functions, namely to 
 secure all those conditions that will enable the best 
 teachers to do their work in the best way. We must 
 
INBORN NATURE AND LEARNING 275 
 
 not, however, lose sight of the fact that the final goal 
 which all must strive to reach is good teaching. The 
 school must preserve a proper balance between con- 
 servatism and progress. While the school is an institu- 
 tion of society one of whose great functions has been 
 to hand down to a new generation that which the old 
 generations thought good, it must also be a means of 
 progress by encouraging and stimulating individuality. 
 In impressing the past on the young, we must not de- 
 stroy the future. Imitation is the aspect of inborn na- 
 ture which makes possible the great influence which the 
 old generation has upon the new. But there are also in 
 human nature inborn tendencies which work for varia- 
 tion and individuality. The wise teacher will seek a 
 proper balance between the two sets of tendencies. 
 There must be a certain degree of conformity if we are 
 to live together with any considerable contentment and 
 happiness, but the conformity must not be such as to 
 destroy all initiative, individuality, and progress. 
 
 Rivalry and Learning.— One of the oldest and strong- 
 est inherited tendencies in man is the tendency to fight. 
 For ages unnumbered, our ancestors have been fighters. 
 Each human individual has had to look out for himself. 
 Natural selection has picked out the fighter. Our an- 
 cestors were those who overcame their adversaries. 
 Man's progress has been one great fight. Every step 
 has been marked in blood. As a result of the constant 
 action of natural selection, there have been established 
 in us many strong inherited tendencies connected with 
 individual survival, that may be called the individual- 
 istic instincts. Examples of these instincts are the acts 
 connected with rivalry, competition, fighting, and with 
 the emotions of anger, envy, and jealousy. The stimuli 
 
276 THE PSYCHOLOGY OF LEARNING 
 
 that evoke these various tendencies and responses are 
 attacks upon ourselves. This attack may be an actual 
 physical attack upon our person, or a threatened 
 attack. It may, however, be merely a verbal attack 
 upon our belief, or our looks, or our friend or relative. 
 Whatever the nature of the attack, the physiological 
 response is much the same. There are characteristic 
 changes in the heart-beat, circulation, respiration, and 
 in various other muscles and organs. A strong emotion 
 is aroused, as anger, or envy, or jealousy. 
 
 When children start to school no tendency is stronger 
 or more marked than individualistic tendencies. The 
 self is then the biggest thing in the v^orld. Children 
 are constantly measuring themselves with one another. 
 They not only compare their personal selves but every- 
 thing that belongs to them, their parents, their homes, 
 their clothes. I have heard children boast of how many 
 times they had been to the dentist! Eternal competi- 
 tion is the rule of child life, as it has been the rule of 
 all life, all business, all trade, in fact almost everything 
 in the life of man. What should be the attitude of the 
 teacher toward this characteristic of childhood? We 
 can not get rid of it. It permeates almost every fibre 
 in the structure of life, and every work and institution 
 of man. We must take it into account if we are going 
 to deal with hum^an nature because it is a large part of 
 human nature. Fortunately it has its good aspect. Self 
 respect depends upon this aspect of our nature. When 
 a person has no fight in him, he is just about hopeless. 
 As society and human nature are now constituted, we 
 must make a limited use of competition in learning. 
 The desire to excel must continue to be a motive. But 
 the use we make of rivalry and competition must be 
 
INBORN NATURE AND LEARNING 277 
 
 limited. As the child grows older we can develop other 
 motives. We can create desire for things in them- 
 selves, desire for absolute good, apart from what others 
 may have. Education must certainly take this attitude, 
 for nearly all the pain and sorrow and trouble in the 
 world are due to some aspect of selfishness. This selfish- 
 ness is in human nature, it is true, but our only hope for 
 any sort of decent life consists in suppressing certain 
 aspects of selfishness, and in developing sympathy and 
 co-operation. The hope of the world lies in building 
 school houses rather than battleships. As long as peace 
 is maintained by the sword, it is not peace, it is merely 
 a truce. 
 
 Roving and Collecting Tendencies. — Man's wild life in 
 the past has left him with two tendencies which are 
 educational assets. The child, by nature, dislikes con- 
 finement, restriction of liberty, and likes freedom. Chil- 
 dren like to be going, to see new things, to have new 
 experiences in other places. A closely related tendency 
 is collecting. A child, by nature, picks up and takes 
 along with him everything that attracts attention and 
 is loose. Education can profit greatly from both these 
 tendencies. Part of the work of education is learning 
 the world. The world is, for the most part, outside of 
 the school house. The child must go out to study it. 
 The river, the mountain, the animals, the plants, nearly 
 all the world, is outside, and the child must go out to 
 study it. He can bring back to school not only a report 
 but specimens. From the specimens should be made 
 a school museum, which should be a reproduction of 
 the world outside. 
 
 Play and Education. — From the point of viev/ of edu- 
 cation, imitation and play are the most important 
 
278 THE PSYCHOLOGY OF LEARNING 
 
 aspects of child life for they lead to nearly all acquire- 
 ments. The importance of imitation has already been 
 pointed out. Play is the spontaneous activity of the 
 individual. It is doing what we want to do, it is the 
 expression of our inmost desires, our most real self. 
 Play activities are those intrinsically pleasurable, those 
 which we seek for the pleasure which they give and not 
 for any good that results. The child learns most suc- 
 cessfully when his practice has in it the spirit of play. 
 This is at once evident to us when we remember what 
 was said in preceding chapters about attention and in- 
 terest. We learn best when we are attentive and when 
 the learning or its result is pleasurable. To put it in 
 another way : we learn best when we put our whole self 
 into the learning. In no other activities are these con- 
 ditions so well met as in play activities. But it is not 
 always easy to realise these conditions. We who are old 
 set the tasks for those who are young. We know what 
 is best for the young to learn. We put them to work 
 to do what w^e impose upon them. One is not likely to 
 put forth all his energy except in a self-imposed task. 
 The learning of children will progress at a much faster 
 rate if we can enlist their native desires and their own 
 ambitions. Many helpful improvements can be made 
 in this direction. Goals and objectives can be selected 
 that are near instead of remote. While the ultimate 
 end of education is adult efficiency, closer ends can al- 
 ways be made out. A child can not be expected to work 
 on indefinitely in the mere hope that some time, some- 
 where, he will reap some return for his labors. In a 
 large measure a child's present needs can be used as 
 incentives which lead to performances that prepare for 
 later life. About all this amounts to is this : In as far 
 
INBORN NATURE AND LEARNING 279 
 
 as possible, we are to make appeal to the child's native 
 equipment, native desires, and native ambitions. The 
 danger is that the teacher, in making this appeal, will 
 go to an extreme that will cause education to lose more 
 than it gains. While we learn best when we do what 
 v/e want to do, an education that permits the child to do 
 only what it wants to do is no education at all. For life 
 is such and this world is such that we all have to do 
 many hard things, many disagreeable things. No one 
 ever amounts to anything who has not learned to do 
 what is hard, who has not learned to stick to a task 
 undertaken. No use of play should be made in educa- 
 tion that in any way hinders us in giving the child a 
 discipline and a training that will lead to worthwhile 
 achievement. The proper solution of the problem con- 
 sists in leading the child to form ambitions and aims 
 and desires in the attainment of which he is willing to 
 work to the uttermost. After all, we never succeed in 
 education until our aims have become the child's aims, 
 until our purposes have become his purposes. 
 
 Manipulation. — A native tendency that is of use in 
 early education and all later education as well, is that 
 of manipulation. The child, by virtue of its inborn 
 nature, wishes to touch, handle, manipulate all that it 
 sees. In the early years of infancy this tendency leads 
 the child to a mastery of many things in its environ- 
 ment. Of course, it is a rather expensive tendency, and 
 parents usually suppress it in the interest of preserving 
 the furniture and other household implements, such as 
 clocks, sewing machines and door bells. In suppressing 
 this tendency, parents kill or partially kill one of the 
 most important characteristics of human nature. True 
 and accurate knowledge comes only from contact with 
 
280 THE PSYCHOLOGY OF LEARNING 
 
 things, only from manipulation. But in opposition to 
 all that is natural, we take the child to school, put it 
 into a seat, tell it to keep its hands off of things, and 
 give it a book. As a result, the child's education is a 
 sort of second hand education, one or more steps re- 
 moved from reality. The proper procedure would be 
 to use in every step of education the inborn tendency 
 to manipulate. In getting a knowledge of the world, 
 the child should manipulate the various objects of the 
 world. 
 
 The Emotions and Learning. — Our discussion of the 
 inborn nature of children leads us finally to a considera- 
 tion of the emotions and their relation to learning. Our 
 emotions are the most intimate parts of us; they are 
 back of nearly all that we voluntarily do. At the bot- 
 tom of nearly every act is love, or hate, or envy, or 
 jealousy, or anger, or fear. Nothing of very great con- 
 sequence is ever undertaken that does not have back of 
 it some emotion. The great teacher is he who has a 
 profound understanding of human life, and knows how 
 to mould and change it ; knows how to use what he finds 
 as a means of making the child different. 
 
 In the preceding chapters of this book we have dis- 
 cussed many factors and elements of learning, we have 
 tried to find the most economical ways of learning. And 
 this is well for we must use every known means 
 to facilitate learning. But no factor is of more signifi- 
 cance than the emotions. It is worth while that a child 
 practice for the proper length of period at whatever he 
 is trying to learn, but unless he is in the proper emo- 
 tional attitude it does not make much difference 
 whether he practice at all. The great teacher is he who 
 can profoundly influence the child, who can inspire, and 
 
INBORN NATURE AND LEARNING 281 
 
 command respect and reverence. The teacher is not 
 merely to find out what the child wants and help him 
 to attain his wants, but should change the wants. The 
 child is to be changed into a being that wants the higher 
 things. What gives us pleasure is in a large measure 
 the result of experience and training. It is the business 
 of the teacher to train the child to get pleasure from the 
 higher things of eternal value. All knowledge, all 
 habits are merely a means. The end is the satisfaction 
 of the higher nature of man. The teacher in the detail 
 of teaching geography, or arithmetic, or history, should 
 always have in mind the higher aim of education as a 
 whole; should always remember that education is to 
 lead to knowledge, and habits, and ideals that are the 
 means of living a life that is worthy of man. 
 
 EXPERIMENTS AND EXEPvCISES. 
 
 1. Make a rough study of the effects of emotion on 
 learning by learning nonsense syllables under different 
 emotional conditions. Within the course of a few v/eeks, 
 oportunity for the different experiments may be offered. 
 
 2. Make a study of the different treatises of instincts, 
 such as are found in Kirkpatrick's Fundamentals of 
 Child Study or the author's Outlines of Educational 
 Psychology, and select all those aspects of original 
 nature that can be used in learning. 
 
 REFERENCES. 
 
 Barnes, E., Studies in education, Vol. I, 1896, Children's Col- 
 lections, p. 144. 
 
 Burk, C. F., The collecting instinct, Fed. Sem. VII, 1900, 179. 
 
 Carver, A., The generation and control of emotions, B. J. P., 
 X, 1919, 51. 
 
 Dearborn, G. V. N., The influence of joy, 1916. 
 
 Hall, G. S., A Study of anger, A. J. P., X, 1898-99, 516. 
 
282 THE PSYCHOLOGY OF LEARNING 
 
 Henry, T. S., The education and control of emotions, J. E. P., 
 VIII, 1917, 407. 
 
 Hunter, W. S., The modificatio7i of instinct from the standpoint 
 of social lihychology, P. R., XXVII, 1920, 247. 
 
 Johnson, G. E., Education by plays and games, 1907. 
 
 Kantor, J. R., A functional interpretation of human instincis, 
 P. R., XXVII, 1920, 50. 
 
 Kline, L. W., Truancy as related to the migrating instinct, 
 Ped. Sem., V, 1897-98, 381. 
 
 The migratory impulse vs. love of home, A. J. P., X, 1898- 
 99, 1. 
 
 Martin, L. J., The training of emotion, 1918. 
 
 Ordahl, G., Rivalry, its genetic development and pedagogy, 
 Ped. Sem., XV, 1908, 492. 
 
 Richardson, R. F., The jysychology a7id pedagogy of anger, Ed. 
 Psych. Mon., No. 19, 1918. 
 
 Saxby, I. B., Some conditions affecting the groioth and perma- 
 nence of desire, B. J. P., IX, 1917, 93. 
 
 Tolman, E. C., Instinct and purpose, P. R., XXVII, 1920, 217. 
 For fuller treatment of the instincts and for more extensive 
 references, see the chapters on the instincts in the author's Out- 
 lines of educational psychology. 
 
Chapter XIV. 
 MEASURES. 
 
 Central Tendency. — After measuring the learning 
 capacity of a group of students, we must have some 
 method of expressing and interpreting the results. The 
 first important determination is the central tendency. 
 There are three commonly used measures of central 
 tendency, the average, the median, and the mode. To 
 find the average, we simply add the scores made by the 
 various individuals tested and divide the sum by the 
 number of cases. The median is the middle measure. 
 To find the median it is necessary to rank the subjects 
 tested on the basis of efficiency in the test, giving the 
 person making the highest score the rank 1; the next 
 highest, the rank 2; the next highest, the rank 3, and 
 so on to the poorest. We then find that rank that has 
 as many ranks above it as below it. The score which 
 this rank represents is the median score. It often hap- 
 pens that the median represents no actual score, as 
 when the number of cases is an even number, and when 
 the median falls within a group of cases having the 
 same score. In such cases, the median must be calcu- 
 lated as shown in the illustration. 
 
 The three measures of central tendency may be illus- 
 trated from the following data, which represent the 
 speed in seconds attained by each of 64 university stu- 
 dents, sorting 25 cards into five boxes. The numbers 
 
 [283] 
 
284 THE PSYCHOLOGY OF LEARNING 
 
 given represent the 12th score. The actual scores were 
 as follows: 20, 19, 21, 11, 19, 16, 17, 21, 19, 27, 16, 15, 
 13, 16, 21, 16, 15, 20, 22, 17, 16, 22, 15, 20, 15, 17, 18, 
 22, 18, 16, 20, 15, 17, 25, 20, 16, 19, 20, 22, 17, 25, 16, 
 20, 24, 24, 21, 21, 22, 24, 19, 16, 13, 21, 24, 18, 18, 19, 
 16, 15, 16, 18, 20, 15, 13. 
 
 Score in Number of 
 
 Seconds. Cases. Sum. 
 
 11 1 
 
 13 3 
 
 15 7 
 
 16 11 
 
 17 5 
 
 18 5 32 
 
 19 6 32 
 
 20 8 
 
 21 6 
 
 22 5 
 
 24 4 
 
 25 2 
 27 1 
 
 The sum of the scores is 1189; the number of cases 
 is 64. 1189-^-64=18.56, the average. By arranging the 
 scores as shown in the accompanying table, it can be 
 seen that 32 scores are 18 seconds or less, and that 32 
 cases are 19 seconds or more. The median therefore 
 lies midway between 18 and 19, and is 18.5 seconds. 
 Suppose we omit the slowest person, whose time was 
 27 seconds, and find the median of the 63 remaining 
 scores. There are 31 scores of 19 seconds and slower 
 and 32 scores of 18 and faster. The median lies in the 
 group of five making a score of 18 seconds. In such a 
 case we find the percentage of this group that must be 
 added or subtracted to make the number of cases equal, 
 and add or subtract, according to which way we are 
 counting, this percentage to the last score preceding 
 
MEASURES 285 
 
 the score of the group in which the median lies. The 
 number of scores up to and including 17 seconds is 27. 
 Half the number of scores is 31V2- 311/2—27=41/2. 
 There are 5 scores in the group 18. 41/2 is .9 of 5. We 
 therefore add .9 to 17, which gives a median of 17.9. 
 
 The mode is the most frequent measure. Reference 
 to the table shows that 16 seconds is the mode. 
 
 Measures of Variability. — After we have measured 
 the central tendency of a group, we need to know the 
 variability, i. e., how much the members of the group 
 vary from the central tendency. There are three com- 
 mon measures of variability, the average deviation, the 
 standard deviation, and the probable error. The aver- 
 age deviation is simply the average amount of deviation 
 from the central tendency. It is found by finding each 
 individual's deviation from the central tendency, adding 
 these amounts without regard to whether they are posi- 
 tive or negative, and dividing by the number of cases. 
 The standard deviation is the square root of the aver- 
 age of the squares of the individual deviations. The 
 probable error is that measure above and below the 
 central tendency that includes half the cases. All of 
 these measures can be illustrated from the scores from 
 card-sorting given above. For computing them, it is 
 customary to use either the median or the average as 
 the measure of the central tendency. We shall use the 
 measure 18.5, which is practically both the median and 
 the average. In the table are shown the individual 
 scores, the deviations, the squares of the individual 
 deviations. 
 
286 THE PSYCHOLOGY OF LEARNING 
 
 Score 
 
 d 
 
 d"" 
 
 22 
 
 3.5 
 
 12.25 
 
 24 
 
 5.5 
 
 30.25 
 
 20 
 
 1.5 
 
 2.25 
 
 15 
 
 3.5 
 
 12.25 
 
 24 
 
 5.5 
 
 30.25 
 
 19 
 
 .5 
 
 .25 
 
 20 
 
 1.5 
 
 2.25 
 
 21 
 
 2.5 
 
 6.25 
 
 21 
 
 2.5 
 
 6.25 
 
 15 
 
 3.5 
 
 12.25 
 
 21 
 
 2.5 
 
 6.25 
 
 11 
 
 7.5 
 
 56.25 
 
 17 
 
 1.5 
 
 2.25 
 
 22 
 
 3.5 
 
 12.25 
 
 19 
 
 .5 
 
 .25 
 
 18 
 
 .5 
 
 .25 
 
 24 
 
 5.5 
 
 30.25 
 
 16 
 
 2.5 
 
 6.25 
 
 22 
 
 3.5 
 
 12.25 
 
 19 
 
 .5 
 
 .25 
 
 17 
 
 1.5 
 
 2.25 
 
 18 
 
 .5 
 
 .25 
 
 16 
 
 2.5 
 
 6.25 
 
 21 
 
 2.5 
 
 6.25 
 
 10 
 
 2.5 
 
 6.25 
 
 13 
 
 5.5 
 
 30.25 
 
 19 
 
 .5 
 
 .25 
 
 20 
 
 1.5 
 
 2.25 
 
 21 
 
 2.5 
 
 6.25 
 
 27 
 
 8.5 
 
 72.25 
 
 15 
 
 3.5 
 
 12.25 
 
 24 
 
 5.5 
 
 30.25 
 
 16 
 
 2.5 
 
 6.25 
 
 17 
 
 L5 
 
 2.25 
 
 18 
 
 .5 
 
 .25 
 
 15 
 
 3.5 
 
 12.25 
 
 25 
 
 6.5 
 
 42.25 
 
 18 
 
 .5 
 
 .25 
 
 13 
 
 5.5 
 
 30.25 
 
 20 
 
 1.5 
 
 2.25 
 
 19 
 
 1.5 
 
 2.25 
 
 16 
 
 2.5 
 
 6.25 
 
 16 
 
 2.5 
 
 6.25 
 
 IG 
 
 2.5 
 
 6.25 
 
 21 
 
 2.5 
 
 6.25 
 
 19 
 
 .5 
 
 .25 
 
 15 
 
 3.5 
 
 12.25 
 
 16 
 
 2.5 
 
 6.25 
 
 20 
 
 1.5 
 
 2.25 
 
 16 
 
 2.5 
 
 6.25 
 
 15 
 
 3.5 
 
 12.25 
 
 22 
 
 3.5 
 
 12.25 
 
 18 
 
 .5 
 
 .25 
 
 20 
 
 1.5 
 
 2.25 
 
 17 
 
 1.5 
 
 2.25 
 
 20 
 
 1.5 
 
 2.25 
 
 22 
 
 3.5 
 
 12.25 
 
 25 
 
 6.5 
 
 42.25 
 
 15 
 
 3.5 
 
 12.25 
 
 17 
 
 1.5 
 
 2.25 
 
 16 
 
 2.5 
 
 6.25 
 
 13 
 
 5.5 
 
 30.25 
 
 16 
 
 2.5 
 
 6.25 
 
 20 
 
 1.5 
 
 2.25 
 
 
 
 
 Slim 179.0 712.00 
 
 The sum of the individual deviations is 179. 
 179_^64=2.8, A. D. The sum of the squares of the 
 individual deviations is 712. 712-^64=11.12. The 
 square root of 11.12=3.3, the S. D. or cj. 
 
 To determine the probable error, we count back each 
 way from the average or median till we have 32 cases. 
 Taking two seconds above and below gives us 24 cases. 
 In the 16-second group are 11 cases, and in the 21-sec- 
 ond group are 6 cases. In the two groups are 17 cases. 
 Only eight of them can be counted to make up the re- 
 quired 32. Eight is .47 of 17, the probable error is 
 therefore 2.47 seconds. In a normal distribution, the 
 probable error is .6745 of the standard deviation. If 
 the scores were distributed in accordance with the nor- 
 mal frequency curve, the probable error would be .6745 
 of 3.3, or 2.23. 
 
MEASURES 
 
 287 
 
 16 
 
 24 
 13 
 12 
 
 11 
 10 
 
 8 
 
 6 
 
 S 
 4 
 3 
 ^ 
 1 
 
 /=>?. jj 
 
 n 
 
 /J 
 
 
 /^ /<^ /^ ^^ <-2- 
 /^ // ^^ //. '-^ 
 
 
 FiGUEE 37. Frequency Surface or Curve of Distribution. The 
 scores are represented on the horizontal axis, and the number of 
 subjects making the respective scores, on the vertical axis. 
 
288 
 
 THE PSYCHOLOGY OF LEARNING 
 
 The relationships among the different measures of 
 variability are as follows: 
 
 S. D.==1.2533 A D., 1.4825 P. E. 
 
 P. E.= .6745 S.D., .8453 A. D. 
 
 A.D.==1.1843P.E., .7979 S.D. 
 
 Frequency Surfaces. — The frequency surface or 
 curve of distribution is a graphical means of represent- 
 ing the distribution of a group with reference to some 
 measure. The distribution of the scores given above is 
 shown in Figure 37, constructed as follows : We deter- 
 mine the number of cases included in the 10 and 11 
 minute groups, the 12 and 13 minute groups, the 14 
 and 15 minute groups, and so on. The scores are rep- 
 resented on the horizontal axis and the number of 
 cases for each score, on the vertical axis. We arbi- 
 trarily select some unit to represent one case, and then 
 simply take those distances above the base that repre- 
 sent the several numbers of cases. The score-groups 
 with their respective numbers of cases are as follows : 
 10-11, 1; 12-13, 3; 14-15, 7; 16-17, 16; 18-19, 11; 20-21, 
 
 Fi^.n 
 
 Figure 38. Normal Frequency Curve. a=the probable error 
 b=the ahe average deviation ; c=the standard deviation. 
 
MEASURES 289 
 
 14; 22-23, 5; 24-25, 6; 26-27, 1. A normal distribution 
 or symmetrical curve is shown in Fig. 38. Human 
 traits are probably distributed in accordance with the 
 normal curve, and in measuring human traits, if we 
 have a large number of cases, we may expect our re- 
 sults to approximate the normal curve. 
 
 Correlation. — By correlation we mean the relation 
 that exists between two functions. For example, sup- 
 pose we have measured the several abilities of the 
 members of a group in one kind of learning and wish 
 to compare those abilities with the corresponding 
 abilities in another kind of learning, we resort to the 
 correlation formula. In a small group, mere 
 inspection of the results of the experiments would 
 enable us to make a rough comparison of the two differ- 
 ent functions, but from mere inspection we could not 
 make an accurate quantitative statement. There are 
 several formulas for computing correlation. The one in 
 
 2XY 
 
 more general use is the Pearson formula, r= , 
 
 n 0^ 02 
 in which 
 
 r==correlation, 
 
 2= the algebraic sum of 
 
 X=the individual deviations in one function, 
 
 Y=the individual deviations in the other function, 
 
 n=the number of cases, 
 
 o-i=the standard deviation in one function, 
 
 cr2=the standard deviation in the other function. 
 
 We give below an illustrative problem. In column one 
 are given the scores of a group of 15 students in visual 
 substance learning, in column 4 are given their corre- 
 sponding scores in visual verbatim learning. 
 
290 THE PSYCHOLOGY OF LEARNING 
 
 12345678 
 
 41 _ 9 81 44—6 36 54 
 
 36 —14 196 35 —15 225 210 
 
 48—2 4 53+3 9 — 6 
 
 52+2 4 57+7 49 14 
 
 64 +14 196 61 +11 121 154 
 
 62 +12 144 62 +12 144 144 
 
 60 +10 100 02 +12 144 120 
 
 62 +12 144 52+2 4 24 
 
 54+4 16 51+1 1 4 
 
 47—3 9 30 —20 400 60 
 
 45—5 25 27 —23 529 115 
 
 39 —11 121 61 +11 121 —121 
 
 51+1 1 56+6 36 6 
 
 47—3 9 55+5 25 — 15 
 
 45 — 5 25 45—5 25 25 
 
 1075 1869 930 142 
 
 The scores are shown in columns 1 and 4, the devia- 
 tions in columns 2 and 5, the squares of the deviations 
 in columns 3 and 6, the plus products are shown in 
 column 7 and the minus products in column 8. The 
 
 1075 
 
 standard deviation for the first column is V =8.46. 
 
 15 
 1869 
 
 The standard deviation for column 4 is V =11.16. 
 
 15 
 The numerator of the fraction in the Pearson formula 
 is the algebraic sum of the products of the individual 
 deviations. The sum of the plus products as shown in 
 column 7 is 930. The sum of the minus products as 
 shown in column 8 is 142. 930—142=788. We then have 
 788 
 
 r= =.556. 
 
 15X8.46X11.16 
 A simple method of computing correlation by using 
 rank differences instead of the absolute deviation is 
 what is known as Spearman's "Foot-Rule" formula: 
 
MEASURES 291 
 
 R = l- 
 
 n"~l 
 
 in which R==correlation, 
 
 :§=:the sum of 
 
 g=an individual's gain in rank in the second function 
 over the rank in the first function, 
 
 n=-.the number of cases. 
 
 The Spearman formula gives a lower correlation than 
 the Pearson formula. 
 
 The procedure in using the Spearman formula is as 
 follows : Rank the subjects with reference to standing 
 in one function, then find their corresponding ranks in 
 the other function. The gains in rank are then added 
 and multiplied by 6. This product is the numerator of 
 the fraction. The denominator is one less than the 
 square of the number of cases. The fraction converted 
 into decimal form is subtracted from 1. The result is 
 the correlation. The result is not so accurate as that 
 obtained from the Pearson formula because no account 
 is taken of the absolute amount of individual differ- 
 ences. It is a useful formula for rough determination 
 of correlation when there are only a few cases. 
 
 The use of the Spearman formula is illustrated in 
 the following example from the same data used in 
 illustrating the Pearson formula. In the first and third 
 columns are given the scores in the two functions. In 
 the second column are the ranks in the first function, 
 and in the fourth column are the ranks for the second 
 function. In the fifth column are indicated the gains in 
 rank. 
 
292 THE PSYCHOLOGY OF LEARNING 
 
 12 3 4 5 
 
 64 1 61 3.5 
 
 62 2.5 62 1.5 1.0 
 
 62 2.5 52 9 
 
 60 4 62 1.5 2.5 
 
 54 5 51 10 
 
 52 6 57 5 1 
 
 51 7 56 6 1 
 
 48 8 53 8 
 
 47 9.5 30 14 
 
 47 9.5 55 7 2.5 
 
 45 11.5 45 11 .5 
 
 45 11.5 27 15 
 
 41 13 44 12 1 
 
 39 14 61 3.5 10.5 
 
 36 15 35 13 2 
 
 Sum of gains=22 
 6X22 
 
 1 =.41=R. 
 
 224 
 N=15, W = 22o, N^— 1=224. 
 
 Probable Error of Correlation. — The formula for de- 
 termining the probable error of correlation is, 
 
 1— r' 
 
 P. E.=.6745 
 
 Vn 
 
 In the illustration of the Pearson formula above, the 
 correlation was found to be .556. (.556) •=.309136. 
 
 1_.309136=.690864, 
 
 V15=3.87, 
 
 .690864-f-3.87=:.18=P. E. 
 
 A correlation is not of much significance unless it is 
 at least three times its P. E. 
 
 Conversion of Grades and Scores. — It often happens 
 that we wish to convert scores or grades made in dif- 
 ferent experiments or different subjects to the same 
 basis for purpose of comparison, or to be able to add 
 the scores and give a single rating for all the scores 
 
MEASURES 293 
 
 combined. The object of reducing to the same basis is 
 to be able to give the same weight to the different 
 measures. For example: Suppose we have given a 
 number of different mental tests to students and wish 
 to express in one number their respective abilities. If 
 we simply add the scores more weight is given to those 
 scores that have a high average, but if we reduce the 
 scores in each test to a common average, each test has 
 the same weight in this average. It makes no differ- 
 ence what average is chosen, but 50 is a convenient 
 number. The process of conversion is as follows : Find 
 the average in each test, and then give each subject a 
 score that has the same ratio to his actual score that 
 50 has to the actual average. By the use of a slide rule 
 this conversion can be quickly and accurately done. In 
 tests that have a work limit, efficiency depending upon 
 time required to do the work, a low score means high 
 efficiency. In such cases, first determine the reciprocals 
 of each score by means of a slide rule or by using tables, 
 such as Barlow's, then make tiie conversion as explained 
 above. 
 
 REFERENCES. 
 
 Rugg, H. 0., Statistical Methods ApjJlied to Education, 1917. 
 Thorndlke, E. L., Mental and Social Measurements, 1904, espe- 
 cially Chs. 4, 5, 6 and 9. 
 
 Whipple, G. M., Manual of Mental and Physical Tests, 1915 
 especially Ch. 3. 
 
BIBLIOGRAPHY 
 
 Explanation of abbreviations. — A, J. P., American Journal o* 
 Psychology ; Arch, of Psych., Archives of Psychology; B. J. P.. 
 British Journal of Psychology; Ed. Rev., Educational Review r 
 J. E. P., Journal of Educational Psychology ; J. Exp. Psych., Journal 
 of Expcriynental Psychology ; Peel. Sem., Pedagogical Seminary; 
 P. R., Psychological Revieio; P. R. Mon. Sup., Psychological Review 
 Monograph Supplement; P. R. Mon., Psychological Review 
 Monograph; S. and S., School and Society; T. C. Cont. to Ei 
 Teachers' College ColumMa University Contributions to Educa- 
 tion. 
 
 Aikens. H. A., and Thorndike, E. L., Correlations in the per- 
 ceptive and associative processes, P. R., IX, 1902, 374. 
 
 Arps, G. F., A Preliminary report on work with knowledge vs 
 work without knoivledge of results, P. R., XXIV, 1917, 449. 
 
 Ash, I. E., Fatigue and its effects upon control, Arch, of Psych., 
 IV, 1914, No. 31. 
 
 Bagley, W. C. and Squire, C. R., Experiments on transfer of 
 ideals of neatness, Educational Values, 1905, 188. 
 
 Bair, J. H., The Practice curve, P. R. Mon. Sup. No. 19, 1902. 
 
 Barnes, E., Studies in education, Vol. I, 1896-7, 144. 
 
 Bean, C. H., The curve of forgetting, Arch, of Psych., XX, No. 3, 
 1912. 
 
 Bennett, F., The correlation between different memories, J. 
 Exp. Psych., I, 1916, 404. 
 
 Bergstrom, J. A., Experiments upon physiological memory by 
 means of the interference of associations, A. J. P. V., 1893, 356. 
 
 The relation of the interference to the practice effect of an 
 association, A. J. P., VI, 1894, 433. 
 
 Bonser, F. G., The selective significance of reasoyiing ability 
 tests, J. E. P., VII, 1916, 187. 
 
 Book, W. F,, The role of the teacher in the most expeditious 
 and economic learning, J. E. P., I, 1910, 183. 
 
 Boswell, F. P. and Foster, W. S., On memorising with the inten- 
 tion permanently to retain, A. J. P., XXVII, 1916, 420. 
 
 Bradford, C. J., An experiment in typewriting, Ped. Sem., 
 XXII, 1915, 445. 
 
 Brown, D. E., Browning, M. and Washburn, M. F., The effect 
 of the interval between repetitions on the speed of learning a 
 series of movements, A. J. P., XXIV, 1913, 580. 
 
 Brown, J. C, An investigation on the value of drill work in 
 [294] 
 
BIBLIOGRAPHY 295 
 
 fundamental operations of arithmetic, J. E. P., II, 1911, 81, also 
 III, 1912, 485, 561. 
 
 Brown, W., Some experimental results in the correlation of 
 mental aUlities, B. J. P., Ill, 1910, 296. 
 Mental measurement, 1911. 
 
 Practice in associating color-names ivith colors. P. R., XXII, 
 1915, 45. 
 
 Practice in associating numl)er names ivith numher symbols, 
 P. R., XXII, 1915, 77. 
 
 Bryan, W. L. and Harter, Noble, Studies on the telegraphic 
 language; the acquistion of a hierarchy of habits, P. R., VI, 1899, 
 346. 
 
 Burk, C. F., The collecting instinct, Ped. Sem., VII, 1900, 179. 
 
 Burnham, W. H., Oxygen supply as a condition of efficient 
 brain activity, J. E. P., II, 1911, 421. 
 
 The hygiene of sleep, Ped. Sem., XXVII, 1920, 1. 
 
 Burt, C, Experimental tests of general intelligence, B. J. P. Ill, 
 1908, 94. 
 
 Carver, A., The generation and control of eynotions, B. .J. P., 
 X, 1919, 51. 
 
 Chamberlain. A. H., A memory test loith school children. P. R., 
 XXII, 1915, 71. 
 
 Chapman, J. C, Individual differences, in ability and improve- 
 ment and their correlations, T. C. Cont. to Ed., No. 63. 
 
 Chapman, J. C. and Nolan, W. J., Initial spurt in a simple men- 
 tal function, A. J. P., XXVII, 1916, 256. 
 
 Chapman, J. C. and Feder, R. B., The effect of external incen- 
 tives on improvement, J. E. P., VIII, 1917, 469. 
 
 Cogan, L. G. and Conklin, A. M. and Hollingworth, H. L., An 
 experimental study of self analysis, estimates of associates and 
 psychological tests, S. and S., II, 1915, 171. 
 
 Conrad, H. E. and Arps, G. F.. An experimental study of eco- 
 nomical learning, A. J. P., XXVII, 1916, 507. 
 
 Coover, J. E. and Angell, F., General practice effects of special 
 exercise, A. J. P., XVIII, 1907, 328. 
 
 Cowan, E. A., An experiment on the influence of training on 
 memory, J. E. P., VII, 1916, 31. 
 
 Culler, A. J., Interference and adaptability. Arch. Psych., No. 24, 
 1912. 
 
 Cummins, R. A., A study of the effect of basket-ball practice on 
 motor reaction, attention and suggestibility, P. R., XXI, 1914, 
 356. 
 
 Improvement and the distribution of practice. T. C. Cont. to 
 Ed., No. 97, 1919. 
 
 Dallenbach. K. M., The effect of practice on visual apprelien- 
 sion in school children. J. E. P., V, 1914, 321, 387. 
 
 The effect of practice upon visual apprehension in the feeble- 
 minded, J. E. P., X, 1919, 61. 
 
296 THE PSYCHOLOGY OF LEARNING 
 
 Dashiell, J. F., The effects of practice upon two mental tests, 
 J. E. P., XI, 1920, 151. 
 
 A comparison of complete vs. alternate methods of learning 
 two haUts, P. R., XXVII, 1920, 112. 
 
 Dearborn, G. V. N., The influence of joy, 1916. 
 
 Dearborn, W. F., The general effects of special practice in 
 memory, P. Bui., VI, 1909, 44. 
 
 Dell, J. A., ^ome observations on the learning of sensible 
 material, J. E. P., Ill, 1912, 401. 
 
 Dewey, J., Hoiv we think, 1910. 
 
 Dodge, R., The laws of relative fatigue, P. R., XXIV, 1917, 89. 
 
 Donovan, M. E. and Thorndike, B. L., Improvement in a prac- 
 tice experiment under school conditions, A. J. P., XXIV, 1913, 
 426. 
 
 Downey, J. E. and Anderson, J. E., Retension of skill after lapse 
 of practice, A. J. P., XXVIII, 1917, 396. 
 
 Dunlap, K., The biological basis of the association of ideas and 
 the development of perception, Psychobiology, II, 1920, 29. 
 Experiments in learning, J. E. P., I, 1910, 373. 
 
 Ebbinghaus, H., Memory, 1885, Tr. by H. A. Ruger, 1913. 
 
 Ebert, E. and Meumann, E., Ueber einige Grundfragen der 
 Psychologle der Uebnngsphdnomene im Bereiche des Geddcht- 
 nisses, Arch. f. d. gesamte Psych., IV, 1904, 1. 
 
 Elderton, E. M., On the association of draiving with other 
 capacities m school children, Biometrica, 1909-10, III, 222. 
 
 Evans, J. E., The effect of distraction on reaction time toith 
 special reference to practice and the transfer of training, Arch, 
 of Psych., No. 37, 1916. 
 
 Finkenbinder, E. 0.. The curve of forgetting, A. J. P., XXIV, 
 1913, 8. 
 
 Fracker, G. C, On the transference of training in memory, 
 P. R. Mon. Sup., No. 38, 1908, 56. 
 
 Fulton, M. .!., An experiment in teaching spelling, Ped. Sem., 
 XXI, 1914, 287. 
 
 Garth, T. R., Mental fatigue. Arch, of Psych., No. 41, 1918. 
 
 Gates, A. I., Diurnal variations in memory and association, 
 Univ. of Cal. Pub. in Psych., I, No. 5, 1916, 323. 
 
 Experiments on the relative efficiency of men and women in 
 memory and reasoning, P. R., XXIV, 1917, 139. 
 
 Recitation as a factor in memorising. Arch, of Psych., No. 
 40, 1917. 
 
 Correlation of immediate and delayed recall, J. E. P., IX, 
 1918, 489. 
 
 The mnemonic span for visual and auditory digits, J. Exp. 
 Psych., I, 1916, 393. 
 
 Gates, A. I., Variations in efficiency during the day, together 
 with practice effects, sex differences, and correlations, Univ. of 
 Cal. Pub. in Psych., II, No. 1, 1916. 
 
 Gilbert, J. A. and Fracker, G. C, The effects of practice in re- 
 action and discrimination for sound upon the time of reaction 
 and discrimAnation of other forms of stimuli, Univ. of Iowa 
 Studies in Psych., I, 1897, 62. 
 
BIBLIOGRAPHY 297 
 
 Gray, T. C, A co^nimrison of two types of learning hy means 
 of a substitution test, J. E. P., IX, 1918, 143. 
 
 Guillet, C, Retentiveness in child and adult, A. J. P., XX, 1909, 
 318. 
 
 A study of memory of young women, J. B. P., VIII, 1917, 65. 
 
 Haggerty, M. E., The laws of learning, P. R., XX, 1913, 411. 
 
 Hall, G. S., Note on early memories, Ped. Sem., VI, 1899, 485. 
 A study of anger, A. J. P., X, 1898-9, 516. 
 Some psychological aspects of language teaching, Ped. Sem., 
 XXI, 1914, 256. 
 
 Hart, B. and Spearman, C, General ability, its existence and 
 nature, B. J, P., V, 1912, 51. 
 
 Hawkins, C. J., Experiments on memory types, P. R., IV, 1897, 
 289. 
 
 Hayden, E. A., Memory for lifted weights, A. J. P., XVII, 1906. 
 497. 
 
 Heck, W. H., A study of mental fatigue, 1913. 
 
 Henmon, V. A. C, The relation between mode of presenta- 
 tion and retention, P. R., XIX, 1912, 79. 
 
 The relation between learning and retention and amount to 
 be learned, J. Exp. Psych., II, 1917, 476. 
 
 Henry T. S., The education and control of emotions, J. E. P., 
 
 VIII, 1917, 407. 
 
 Hewins, N. P., The doctrine of formal discipline in the light of 
 experimental investigation, 1916. 
 
 Hill, D. S., Minor studies in learning and relearning, J. E. P., 
 V, 1914, 375. 
 
 Hollingworth, H. L.. Variations in effliciency during the work- 
 ing day, P. R., XXI, 1914, 473. 
 
 Characteristic differences between recall and recognition. 
 A. J. P., XXIV, 1913, 532. 
 
 Correlation of abilities as affected by practice, J. E. P., IV, 
 1913, 405. 
 
 Individual differences before, during and after practice, P. R., 
 XXI, 1914, 1. 
 
 Hunter, W. S., The modification of instinct from the standpoint 
 of social psychology. P. R., XXVII, 1920. 247. 
 
 Johnson, G. E., Education by plays and games, 1907. 
 Judd, C. H., The relation of special training to general intelli- 
 gence, Ed. Rev., XXXVI, 1908, 28. 
 
 Practice and its effects on the perception of illusions, P. R., 
 
 IX, 1902, 27. 
 
 The psychology of high school subjects, 1915, Ch. 17. 
 
 Kennedy, P., On the experimental investigation of memory, 
 P. R., V, 1898, 477. 
 
 Kentor, J. R., A functional interpretation of human instincts, 
 P. R., XXVII, 1920, 50. 
 
 King, I., Relationship of abilities in certain mental tests to 
 abilities estimated by teachers, S. and S., V, 1917, 204. 
 
 King, I. and Homan, J. B., Logical memory and school grades, 
 J. E. P., IX, 1918, 262. 
 
298 THE PSYCHOLOGY OF LEARNING 
 
 Kirkpatrick, E. A., An experimental study of memory, P. R., 
 I, 1894, 602. 
 
 An experiment in memorising vs. incidental learning, J. E, P., 
 V, 1914, 405. 
 
 Kline, L. W., Some experimental evidence in regard to formal 
 discipline, J. E. P., V, 1914, 259. 
 
 The migratory impulse vs. love of home, A. J. P., 1898-9, 1. 
 Truancy as related to the migratory instinct, Ped. Sem., V, 
 1897-8, 381. 
 
 Kline, L. W. and Owens, W. A., Preliminary report of a study 
 in the learnirig process involving feeling tone, transference and 
 interference, P. R., XX, 1913, 206. 
 
 Krueger, F. and Spearman, C, Die correlation zivischen ver- 
 schiedenen geisiiqen LeistungsfliMg'keiten. Zeit, f. Psych. Bd., 
 XLIV, 1906, 50. 
 
 Kuhlman, F., The place of mental imagery and memory among 
 mental functions, A. J. P., XVI, 1905, 337. 
 
 Lakenan, M. E., The whole and imrt methods of memorising 
 iwetry and prose, J. E. P., IV, 1913, 189. 
 
 Lashly, K. S., Studies of cerehral function in learning, Psycho- 
 biology, II, 1920, 55. 
 
 Lee, F. S., The nature of fatigue, P. S. M., LXXVI, 1910, 182. 
 
 Leuba, J. H. and Hyde, W., An experiment in learning to make 
 hand-movements, P. R., XII, 1905, 351. 
 
 Lyons, C. K., The doctrine of formal discipline, Ped. Sem., 
 XXI, 1914, 343. 
 
 Lyon, D. O., The relation of quickness of learning to retentive- 
 ness. Arch, of Psych., No. 34, 1916. 
 
 Memory and the learning process, 1917. 
 
 McOall, W. A., Correlation of some psychological and educa- 
 tional measurements, 1916. 
 
 Also article same title, S. and S., V, 1917, 24. 
 
 McDoiigal. W., On a nciv method for the study of concurrent 
 mental operations and of mental fatigue, B. J. P., I, 1905, 435. 
 
 Marsh, II, D., The diurnal course of eJJiHcncy, 190G. 
 
 Martin, L. J., The training of emotions, 1918. 
 
 Miles, W. R„ A study in the correlation of elementary and high 
 school grades, J. E. P., I, 1910, 537. 
 
 Mosso, A., Fatigue, Eng. Tr., 1904. 
 
 Mulhall, E. F., Tests of the memories of school children, J. E. P., 
 VII, 1917, 294. 
 
 Miinsterberg. II.. and Bigham. J., Memory. P. R., I, 1894, 34. 
 
 Murphy, H. H., Distribution of practice periods in learning, 
 J. E. P., VII, 1916, 150. 
 
 Myers, G. C, A comparative study of recognition and recall, 
 P. R., XXI, 1914, 442. 
 
 Recall in relation to retention. J. E. P., V, 1914, 119, 
 Confusion in recall, J, E, P., VIII, 1917, 166. 
 
 Norsworthy, N., Acquisition as related to retention, J. E. P., 
 Ill, 1912, 214. 
 
BIBLIOGRAPHY 299 
 
 Offner, M., Mental Fatigue, Eng. Tr., 1911. 
 
 Ordahl, L. E., Consciousness in relation to learning, A. J. P., 
 XXII, 1911, 154. 
 
 Rivalry ; its genetic development and pedagogy, Ped. Sem., 
 XV, 1908, 492. 
 
 Otis, A. S., Do tve think in words?, P. R., XXVII, 1920, 399. 
 
 Painter, W. S., Efficiency^ iti mental multiplication under ex- 
 treme fatigue, J. E. P., VII, 1916, 25. 
 
 Patterson, T. L., Pedogogical suggestions from memory tests, 
 J. E. P., IX, 1918, 497. 
 
 Pechstein, L. A., Alleged elements of waste, in learning a motor 
 proUem 'by the "i^arr' method, J. E. P., VIII, 1917, 303. 
 
 Whole vs. part methods in learning nonsensical syllables, 
 J. E. P., IX, 1918, 381. 
 
 ^ Perkins, N. L., The value of distributed repetitions in rote 
 learning, B. J. P., VII, 1914, 253. 
 ^ Peters, C. C, The influence of speed drills upon the rate and 
 the effectiveness of silent reading, J. E. P., VIII, 1917, 350. 
 
 Peterson, J., TJie effect of attitude on immediate and delayed 
 repetition, J. E. P., VII, 1916, 523. 
 
 Peterson, A., Correlation of certain mental tests in normal 
 school students, P. R., XV, 1908, 323. 
 
 Phillips, F. M., A comparison of the loork done in successive 
 7ninutes of a ten-minute practice period in the fundamentals of 
 arithmetic, J. E. P., VII, 1916, 271. 
 
 Phillips, P. M., Yalue of daily drill in arithmetic, J. E. P., IV, 
 1913, 159. 
 
 Poffenberger, A. T., The influence of ifiprovement in one men- 
 tal process xipon other related processes, J. E. P., VI, 1915, 459. 
 Poffenberger, A. T. and Tallman, G. G., Variability in perform- 
 ance during brief periods of work, P. R., XXII, 1915, 371. 
 
 Pohlmann, A., Experimentelle Bcitrdge zur Lehre vom 
 Gedachtnis. 1906. 
 
 Potwin, E. B., Study of early memories, P. R., VIII, 1901, 596. 
 
 Pyle, W. H., Retention as related to repetition, J. B. P., II, 
 1911, 311. 
 
 Economical Learning, J. E. P., IV, 1913, 148. 
 Concentrated vs. distributed i>ractice, J. E. P., V, 1914, 247. 
 Standards of mental efficiency, J. E. P., IV, 1913, 61. 
 and Snyder, 
 
 , The most economical unit for committing to memory, J. E. P., 
 II, 1911, 133. 
 
 Is learning capacity constant for different types of material?, 
 J. E. P., X. 1919, 121. 
 
 A study of the mental and physical characteristics of the 
 Chinese, S. and S., VIII, 1918, 264. 
 
 Pyle, W. H. and Collings. P. E., The mental and physical devel- 
 opment of rural children, S. and S., VIII, 1918, 534, 
 
 Ruediger, W. C, The indirect improvement of mental functions 
 throuoh ideals, Ed. Rev., XXXVI, 1908, 364. 
 
800 THE PSYCHOLOGY OF LEARNING 
 
 Ruger, H, A., The psychology of efficiency, Arch, of Psych., 
 No. 15, 1910. 
 
 Reed, H. B., A repetition of Ehert and Meumann's practice 
 experiment in memory, J. Exp. Psych., II, 1917, 315. 
 
 Race, H. V., A study of a class of children of superior intelli- 
 gence, J. E. P., IX, 1918, 91. 
 
 Rich, G. J., Directed attention and learning, J. E. P., VIII, 1917, 
 239. 
 
 Richardson, R. F., The psychology and pedagogy of anger, 
 Ed. Psych. Mon. No. 19, 1918. 
 
 Rietz, H. L., On the correlation of the marks of students in 
 mathematics, and in law, J. E. P., VII, 1916, 87. 
 
 Root, W. T., A study of mental fatigue loith a group of five 
 hoys, J. E. P., VII, 1916, 533. 
 
 Ruckmich, C. O., The psychology of piano instruction, J. E. P., 
 V, 1914, 185. 
 
 Rugg, H. O., The experimental determination of mental dis- 
 cipline in school studies, 1916. 
 
 Seashore, C. E. and Kent, G. H., Periodicity and progressive 
 change in continuous mental work, P. R., Mon. Sup., VI, No. 28, 
 1905. 
 
 Shaw, J. C, A test of memory in school children, Ped. Sem., 
 IV, 1896, 61. 
 
 Simpson, B. R., Correlations of mental aUlities, T. C. Cont. to 
 Ed., No. 53, 1912. 
 
 Sleight, G. W., Memory and formal discipline, B. J. P., IV, 1911, 
 386. 
 
 Smith, W. G., The) place of repetition in memory, P. R., Ill, 
 1896, 21. 
 
 Smith, F. 0., The effect of training in pitch discrimination, 
 P. R. Mon., No. 69. 
 
 Smith, M. and McDougal, W., Some experiments in learning 
 and retention, B. J. P., X, 1920, 199. 
 
 Smith, T. L., On muscular memory, A. J. P., VII, 1896, 453. 
 
 Smith, M., A contribution to the study of fatigue, B. J. P., 
 VIII, 1916, 327. 
 
 Snoddy, G. S., An experimental analysis of a case of trial and 
 error learning in the human suhject, P. Mon., No. 124, 1920. 
 
 Spearman, C., General intelligence objectively determined and 
 measured, A. J. P., XV, 1904, 201. 
 
 Manifold suh-theories of the ''two factors'', P. R., XXVII, 
 1920, 159. 
 
 Starch, D., Transfer of training in arithmetical operations, 
 J. E. P., II, 1911, 306. 
 
 Periods of work in learning, J. E. P., Ill, 1912, 209. 
 Correlations among abilities in school studies. J. E. P., IV, 
 1913, 41L 
 
 Starch, D. and Ash, I. E.. The mental icork curve, P. R., 
 XXIV, 1917, 391. 
 
BIBLIOGRAPHY 301 
 
 Stetson, G. R., Some memory tests of whites and blacks, P. R., 
 
 IV, 1897, 285. 
 
 Stickland, G. I., The inftuence of practice in the correlation of 
 aUlities, J. E. P., IX, 1918, 393. 
 
 Strong, E. K., The effect of time interval upon recognition 
 memory, P. R., XX, 1913, 339. 
 
 Two factors which influence economical learning, Journal of 
 Philosophy and Scientific Method, XI, No. 5, 1914. 
 
 The effect of length of series upon recognition memory, 
 P. R., XIX, 1912, 447. 
 
 An i7iteresting sex difference, Ped. Sem., XXII, 1915, 521. 
 Swift, E. J., Studies in the psychology and physiology of learn- 
 ing, A. J. P., XIV, 1903, 201. 
 
 Memory of a complex skillful act, A. J. P., XVI., 1905, 131. 
 Talbot, E. B., An attempt to train the visual memory, A. J. P., 
 VIII, 1897, 414. 
 
 Terman, L. M. and Hocking, A., The sleep of school children, 
 its distribution according to age and its relation to physical and 
 mental efficiency, J. E. P., IV, 1913, 138, 199, 269. 
 
 Thompson, G. H., A hierarchy without a general factor, B. J. P., 
 VIII., 1916, 271. 
 
 Thorndike, E. L., Mental fatigue, P. R., VII, 1900, 466, 547. 
 Memory for paired associates, P. R., XV, 1908, 122. 
 The effect of practice in the case of a purely intellectual 
 function, A. J. P., XIX, 1908, 374. 
 
 The relation between memory for words and memory for 
 numbers, and the relation between memory over short and mem- 
 ory over long intervals, A. J. P., XXI, 1910, 487. 
 
 Practice in the case of addition. A. J. P., XXI, 1910, 483. 
 
 Mental Fatigue, J. E. P., II, 1911, 61. 
 
 The curve of work. P. R., XIX, 1912, 165. 
 
 Repetition vs. recall in memorising vocabularies. J. E. P., 
 
 V, 1914, 596. 
 
 Notes on practice. improvabiUtv, and the curve of work. 
 A. J. P., XXVII. 1916, 550. 
 
 The effect of continuous exercise and of rest upon different 
 mental multiplication. J. E. P., V, 1914. 597. 
 
 Fatigue in a complex function. P. R., XXI, 1914, 402. 
 
 and Lay, W. and Dean, P. R., The Relation of accttracy in 
 sensory discrimination to general intelligence, A.J.P., XX, 1909, 36-*. 
 
 and Woodworth, R. S., The influence of improvement in one 
 mental function upon the efficiency of other functions. P. R., VIII, 
 1901, 247, 384, 553. 
 
 Tolman, E. C, Retroactive inhibition as affected by conditions 
 of learning, P. Mon., No. 107. 1917. 
 
 Instinct and Purpose, P. R., XXVII, 1920, 217. 
 Wallin, J. E. W., Two neglected instances of the transfer of 
 training, J. E. P., I, 1910, 168. 
 
 Has the drill become obsolescent?, J. E. P.. I, 1910, 200. 
 
302 THE PSYCHOLOGY OF LEARNING 
 
 Washburn, M. F., The effect of the interval hetiveen repetition 
 on the speed of learning a series of movements, A. J. P., XXIV, 
 1913, 580. 
 
 The effect of mental type on the interference of motor haMts, 
 A. J. P., XX, 1909, 282. 
 
 Watkins, S. H., Immediate memory and its evaluation, B. J. P., 
 VII, 1914-15, 319. 
 
 Weglin. D. E., Correlation of abilities of high school pupils. 
 1917. 
 
 Wells, F. L., Normal performance in the tapping test before 
 and durinq practice, with special reference to fatigue phenomena, 
 
 A. J. P., XIX, 1908. 437. 
 
 The relation of practice to individual differences, A. J. P., 
 XXIII, 1912, 75. 
 
 Practice and the ivork curve, A. 3. P., XXIV, 1913, 35. 
 Practice effects of free association, A. J. P., XXII, 1911, 1. 
 and Henmon. V. A. C. Concerning individual differences in 
 reaction time. P. R.. XXI, 1914, 153. 
 
 Whipple, G. M. and Curtis, J. N., Preliminary investigation o) 
 skimming in reading, J. E. P., VIII, 1917, 333. 
 
 Whitehead, L. G., A study of visual and aural memory pro- 
 cesses, P. R., Ill, 1896, 258. 
 
 Whitley, M. T., An empirical study of certain tests for indi- 
 vidual differences. Arch, of Psych., No. 19, 1911. 
 
 Wimms, J. H., The relative effects of fatigue and practice pro- 
 duced by different kinds of mental work, B. J. P., II, 1907, 153. 
 
 Winch, W. H., Mental adaptation during the school day as 
 measured by arithynetical reasoning, J. E. P., IV, 1913, 17. 
 
 Some measurements of mental fatigue in adolescent pupils in 
 evening schools, J. E. P., I, 1910, 13, 83. 
 
 Mental fatigue in day school children as measured by imme- 
 diate memory, J. E. P., Ill, 1912, 18, 75. 
 
 Mental fatigue in day school children as measured by arith- 
 metical reasoning, B. J. P., IV, 1911, 315. 
 
 The transfer of improvement in memory in school children, 
 
 B. J. P., Ill, 1910, 386. 
 
 Additional researches on learning to spell — transfer and 
 direct vs. indirect methods. J. E. P., VII, 1916, 93. 
 
 Further loork on numerical accuracy in school children. 
 Does improvement in numerical accuracy transfer?, J. E. P., II, 
 1911, 262. 
 
 The transference of improvem^ent in memory of school chil- 
 dren. B. J. P., II, 1908, 284. 
 
 Immediate memory in school children, B. J. P., I, 1904, 127; 
 II, 1906, 52. 
 
 Some relations between substance mem.ory and productive 
 imagination in school childreji, B. J. P., 1911, 95. 
 
 Wissler, C, The correlatioyi of mental and physical tests, P. R. 
 Mon. Sup., No. 16, 1901, 62. 
 
BIBLIOGRAPHY 303 
 
 Woodrow, H., Practice and transference in normal and feeble 
 minded children, J. E. P., VIII, 1917, 85, 151. 
 
 Wright, W. R., Some effects of incentives on ivork end fatigue 
 P. R., XIII, 1906, 23. 
 
 Wyatt, S., The quantitive investigation of higher mental pro 
 cesses, B. J. P., VI, 1913, 109. 
 
 Yoakum, C. S. and Calfee, M., An analysis of the mirror draxo 
 ing experiment, J. B. P., IV, 1913. 283. 
 
 An experimental study of fatigue, P. R., Mon. Sup., No. 46 
 1909. 
 
INDEX 
 
 Abstraction aud isolation, 235. 
 Add, learning to, 24. 
 Adding, experiments in, 42. 
 Algebra, study of, 238. 
 Analysis of experience, 102. 
 Archery, experiments in, 42. 
 Arithmetic, drill in, 74. 
 Arps, 62. 
 
 Association, laws of, 54, 93. 
 y^ Attention and learning, 53. 
 Attention, range of, 81. 
 Attitude and learning, 57. 
 Attitude of school children, 59. 
 
 Bagley, 227. 
 
 Bean, 134. 
 
 Bergstrom, 62. 
 
 Bigham, 128. 
 
 Binet, 130. 
 
 Bolton, 121. 
 
 Book, 137. 
 
 Bryan, 32. 
 
 Bright and dull pupils, 195. 
 
 Brown, 14. 
 
 Burt, 166. 
 
 Cattell, 209. 
 
 Chapman, 47, 80. 
 
 Card-sorting experiment, 179. 
 
 Central tendency, 283. 
 
 Chamberlain, 119. 
 
 Chinese, learning capacity of, 
 203. 
 
 Collecting tendencies, 277. 
 
 Collings, 195. 
 
 Concentration, 54. 
 
 Continuous w^ork, 256. 
 
 Conversion of grades and scores. 
 292. 
 
 Correlations, 184, 289. 
 
 Coover and Angell's experi- 
 ments, 221. 
 
 Country children, learning capa- 
 city of, 201. 
 
 Cramming, 153. 
 
 Cross education, 225. 
 
 Curtis, 78. 
 
 Curve, different form of, 28. 
 
 Dallenbach, 81. 
 
 Daily efficiency, course of. 
 
 2.50, 251. 
 Dearborn, 38, 221. 
 Definiteness of bond, 64. 
 Definite procedure, 85. 
 DIFFERENCES IN 
 
 LEARNING CAPACITY, 193. 
 Disposition, 60. 
 Distribution of pupils, 20O. 
 Donovan, 75. 
 
 Drill, miscellaneous factors, 79. 
 Practical interference, 82. 
 
 Early life preparatory, 85. 
 Ebert and Meumann's experi- 
 ments, 216. 
 Ebbinghaus, 130, 133, 136. 
 ECONOMICAL LEARNING. 36. 
 ECONOMICAL LEARNING, 
 
 . GENERAL FACTORS, 53. 
 Education and conformity, 274. 
 Education, machinery of, 274. 
 Education and heredity, 210. 
 Education, higher aim of, 281. 
 Efficiency, maximum, 29. 
 Emotion and learning, 280. 
 English, indefinite, 66. 
 Environmental iufiuences, 208. 
 
 Fast learner, characteristics of, 
 
 55. 
 FATIGUE AND 
 
 LEARNING, 244. 
 Fatigue, antitoxins, 258. 
 feeling of, 263. 
 Gates' experiments, 254. 
 in habituated work, 257. 
 in non-habituated work. 
 257. 
 
 [305] 
 
306 
 
 THE PSYCHOLOGY OF LEARNING 
 
 kinds of, 247, 248. 
 
 and learning, 245. 
 
 measures of, 246, 247. 
 
 nature of, 244. 
 
 of school children, 248, 249 
 Feeling and learning, 6G. 
 Feeble-mindedness, Mendellian 
 
 206. 
 Finkinbinder, 133, 134. 
 Forgetting, 133. 
 Formal drill, 74. 
 Fraeker, 221, 222. 
 Franklin, 65. 
 
 Galton, 206. 
 
 Gates. 119, 248, 250. 
 
 Geissler's experiments, 57. 
 
 Generalised experience, 235. 
 
 General ability, 86. 
 
 General learning capacity, "uat 
 
 ure of, 169. 
 General mental factor, 171. 
 Getting knowledge, 101. 
 Goddard, 206. 
 Guillet, 152. 
 
 Habit, 1. 
 Hahn, 42. 
 Hawkins, 128. 
 Heck, 248. 
 Henderson, 123. 
 Henmon, 128, 130. 
 Henri, 130. 
 
 Hereditary factors, 207. 
 Hill, 139. 
 
 Hollingworth, 191. 
 Howe's experiments, 22, 29. 
 How to study, 108. 
 Hyde and Leuba's experiment;? 
 226. 
 
 Ideas, initial connection, 8. 
 Ideational disturbances, 56. 
 IDEATIONAL LEARNING, 91 
 Ideation learning, tests of, 181 
 Ideational method, 8. 
 Improvement, amount possible 
 
 22. 
 Improvement, limit of, 26. 
 
 (^ 
 
 mitation, 271. 
 
 in adult life, 274. 
 
 in school life, 272. 
 
 and ideals, 273. 
 Inattentive learning, 55. 
 Individual differences, 193. 
 
 cause of, 206. 
 INBORN NATURE IN RELA- 
 TION TO LEARNING, 270. 
 Incentives, 61, 80. 
 Initial connection, 6. 
 Interference, 239. 
 
 Brown's experiments, 239. 
 Interrelations of mental func- 
 tions, 164. 
 Instinct and habit, 270. 
 
 James. 121, 235. 
 -lost, 37. 
 Judd, 226. 
 
 Kemsies, 128. 
 Kirby, 42. 
 Kirkpatrick, 79. 
 Kline, 67. 
 Knowledge, 2. 
 
 and ideas, 91. 
 
 of end, 61. 
 
 of errors, 63. 
 
 function of. 11. 
 
 physiological basis of, 6. 
 
 of score, 62. 
 
 of when material is to he 
 used, 63. 
 Knowing relations. 236. 
 
 Latin, study of, 238. 
 Life-bents, 60. 
 LEARNING CAPACITY, 
 
 MEASURING. 175. 
 Learning capacity and flnal 
 
 efficiency, 189. 
 Learning, a matter of degree, 20.' 
 and memor.y, 113. 
 motor, 48. 
 
 objectively considered, 3. 
 physiologically considered, 
 
 4. 
 regularity of, 163. • 
 
INDEX 
 
 307 
 
 relation to retention, 122. 
 
 relation to other functions, 
 185. 
 
 specific, 236. 
 
 tests. 177. 
 LEARNING CURVES, 10. 
 Learning curves, construction 
 
 of, 16. 
 
 rise of, 19. 
 
 explanation of rise, 23. 
 
 smoothing of, 31. 
 Length of series, 129. 
 Lyon, 124, 130. 
 
 Man as result of learning, 13. 
 
 Manipulation, 279. 
 
 Marble-sorting experiment, 180. 
 
 Manner of presentation, 127. 
 
 Mathematics, definite, 06. 
 
 Meaning 108. 
 
 MEASURES, 283. 
 
 Measuring reasoning capacity. 
 106. 
 
 Measures, where made, 183. 
 
 Memorj^ and age, 114. 
 
 Memory, immediate and de- 
 layed, 149. 
 
 Memory and intelligence. 141. 
 
 Memory and practice, 120. 
 
 Memory and sex, 117. 
 
 Memory, different kinds of, 143. 
 
 jNIemory experiments, interpre- 
 tation of, 224. 
 
 Memory material, 126. 
 
 Memories, interrelation of, 162. 
 
 Meumann, 130. 
 
 Motor learning, tests of, 179. 
 
 Musical ability, 209. 
 
 Muller and Schumann, 121, 123. 
 
 Mulhall, 119. 
 
 Murphy, 42, 47. 
 
 NATURE OF LEARNING, 1. 
 NATURE OF LEARNING 
 
 CAPACITY, 157. 
 Negro children, learning capa- 
 city of, 182, 203, 205. 
 Norsworthy, 123. 
 
 Ogden, 123. 
 
 Ordahl, 59. 
 
 Organisation of experience, 100. 
 
 Painter, 256, 258. 
 
 Perkins, 37. 
 
 Peters, 77. 
 
 Peterson's experiments, 58, 221. 
 
 Phillips, 75. 
 
 Pitch discrimination, 81. 
 
 Plateaus, learning, 31. 
 
 Play and education, 227. 
 
 Plasticity, 12. 
 
 Pohlmann, 128. 
 
 Poor learners, 255. 
 
 Position in series, 153. 
 
 Practice, 44, 46. 
 
 Practice, direct, 79. 
 
 distribution of, 43. 
 
 in mental functions, 80. 
 
 periods for children, 48. 
 
 periods, length and distri- 
 bution of, 36. 
 
 vs. study, 226. 
 Preferential routes, 5. 
 Primary experience, 99. 
 Punishments, improper, 61. 
 Pyle, 38, 165. 
 
 Radical differences, 203. 
 
 Radossawljewitch, 133. 
 
 Reasoning, 96. 
 
 Re'asoning, specific, 105. 
 
 Recognition, 151. 
 
 Retentiveness, individual differ- 
 ences in, 144. 
 
 Retention, measurements of, 149. 
 
 RETENTION OF 
 
 EXPERIENCE, 113. 
 
 Retention over long periods. 
 136. 
 
 Rewards, 69. 
 
 Rivalry and learning, 275. 
 
 Roving tendencies, 277. 
 
 Rote memory, 148. 
 
 Ruediger, 227. 
 
 Ruger, 227. 
 
 Scholarship, stability of, 207. 
 School and individual differ- 
 ences, 210. 
 
f&^ 
 
 308 
 
 THE PSYCHOLOGY OF LEARNING 
 
 School room drill, 78. 
 
 Sensational disturbances, 50. 
 
 Simpson, 166, 167, 172. 
 
 Skill, how much desirable, 87. 
 
 Sleep and fatigue, 260. 
 
 Sleight, 221, 222. 
 
 Smedley, 128. 
 
 Snoddy, 67. 
 
 Spearman, 172. 
 
 Specific abilities, 168. 
 
 Spelling, drill in, 77. 
 
 Spurts, 29. 
 
 Strong, 47, 151. 
 
 Starch. 88, 39. 
 
 Substitution test, 177. 
 
 Swift, 138. 
 
 Teacher, function of, 89. 
 
 Thinking, 95. 
 
 Thorndike and Woodworth's ex- 
 periments, 225. 
 
 Thorndike, 75, 78, 140, 172, 206, 
 256, 258. 
 
 Transfer of knowledge, 226. 
 
 TRANSFERENCE AND 
 
 INTERFERENCE, 218. 
 
 Transference, discriminative 
 judgments, 224. 
 factors of, 280. 
 
 factors of, in card sorting, 
 
 220. 
 Gilbert and Fracker's ex- 
 periments, 225. 
 historical and critical, 214. 
 interpretation of results, 
 
 229. 
 memory exrteri|ments, 
 methods of studying, 215. 
 Rugg's experiments, 227. 
 summary of evidence, 229. 
 Trial and error, 7, 10. 
 Typewriting: experiments,, 
 Pyle's, 43. 
 
 University students, variations 
 in, 198. 
 
 Variability, measures of, 285. 
 Verbatim learning, 107. 
 Voluntary recall, 152. 
 Von Sybyl, 128. 
 
 Wallin, 77. 
 Whipple, 78, 81. 
 Winch, 221, 222, 248, 76. 
 Work of students, 265. 
 Wright, 02. 
 
 01