iri ee ee 
 
 SSabit oe 
 
 be e ee oe ys Sptstohgtsnetrtn : ° th (se tes z . ES dette ir pictae tity 
 TSowiscesgieeep raat : irnierereist in Safets nici OgSrt , : Sat any oStare ie sania 
 
 Cae ed) Soke na 
 
 abenteds, . 7 ad - 
 
 ‘ re lisbelar bak pula ee 
 
 = pete goiter Reb angady ELEY 
 
 Aw iete 
 
 OA amd et Be Safin 
 nis betel ehaheat ty 
 
 week 
 
 Pn: 
 
 > i ant - ety Date ja eters = aoe ves Pos De epee ins Heo het 
 e ig Fag Sai g Seite men : ras cane a heats 
 Fie f » > Fink M5 mek . : 
 ee ; x , : ; Sirians t 
 
 pe Say hare 
 ‘ peta 
 
 =e sift Caattansetaas 
 are eee eerrnes 
 
 . oP ® ar 
 
 3 a = pi HR NEN: 
 
 bal Be fag "ptt . 2 
 ahah setae ; oman ee . raha? ake 
 ap the ae Pete ey that etapa : twee stew oa 
 re Sacha ne leh 
 pain cates 
 
 Se ae :’ Se uedbthed hall lag he = Hag peda The 
 Nee! SS ee t <By fag $ a etinmacrteasetn o 
 ica » Bete eet of oe peer e ie) nee 
 mth : . 
 
 cdotetleigte 
 Ra detetal Ay 
 pimyteg hel 
 
 at 
 
 ka acta a a 
 EN Dd Wehaher oo 
 tare 
 
 
 
 484, 5 SEE RIYC tein + Lys SF, a aie ih a 
 ie ieee DERE EE Ye ig arene — 
 Gh is ee yte een enc ie tent a chs age a 
 ae ee TR PL RSA SS be oe* os Sth eb 
 

 
7 
 
 Me PSYCHOLOGICAL REVIEW PUBLICATIONS WHOLE No, ISI 
 NO. 4 192 
 
 Sea! eRnhy OF PRing 
 
 YY : 
 
 
 
 
 Psychological Monographs \,"" 
 
 A 
 % 
 LO o6tcAL SEM 
 
 EDITED BY 
 
 JAMES ROWLAND ANGELL, Yate University 
 HOWARD C. WARREN, Princeton University (Review) 
 JOHN B. WATSON, New York (J. of Exp. Psychol.) 
 MADISON BENTLEY, University or Itirnois (Jndexr) 
 S. W. FERNBERGER, University oF PENNSYLVANIA (Bulletin) 
 
 The Personalities of the Socially and 
 the Mechanically Inclined 
 
 A Study of the Differences in Personality Between 
 Men Whose Primary Interest is Social and Men 
 Whose Primary Interest isin Machines 
 
 BY 
 :, 
 MAX FREYD, Px.D. 
 
 PSYCHOLOGICAL REVIEW COMPANY 
 PRINCETON, N.J. 
 
 AcEnts: G.E.STECHERT & CO., Lonpon (2 Star Yard, Carey St., W. C.) 
 PARIS (16 rue de Condé) 
 

 
ACKNOWLEDGMENTS 
 
 Whatever merit this study may possess is due in great part to 
 the liberal cooperation and assistance given by a number of people, 
 and it is with pleasure that the writer acknowledges his indebted- 
 ness and gratitude for this help. The staff of the School of Life 
 Insurance Salesmanship, Dean T. M. Focke, of Case School of Ap- 
 plied Science, and Dean J. T. Morris, of Carnegie Institute of 
 Technology, cooperated in making their students available as sub- 
 jects for this research. Considerable assistance in handling the 
 data was given by Miss Mary Morrison, Miss Helen McKee, and 
 Miss Hortense Purdy. Numerous helpful suggestions and criti- 
 cisms were supplied by Dr. F. L. Wells, of the Boston Psycho- 
 pathic Hospital, and Dr. L. L. Thurstone, of the Department of 
 Psychology, and Dr. W. V. Bingham, Director of the Division of 
 ‘Cooperative Research of Carnegie Institute of Technology. 
 
 To the staff of the Bureau of Personnel Research must go a 
 large share of the credit for assistance. To acknowledge in detail 
 the indebtedness of the writer for constant aid which the members 
 of the Bureau staff stood ready to give, would be a lengthy though 
 not unpleasant task. Among others, Mr. L. D. Anderson cooper- 
 ated in allowing the use of unpublished data on the Will-Tempera- 
 ment Test which he had collected, and Dr. C. E. Hansen aided 
 materially in criticising the manuscript. To Dr. C. S. Yoakum, 
 Director of the Bureau of Personnel Research, without whose con- 
 stant advice and aid this study would have been almost impossible, 
 goes a large share of the credit for its completion and present 
 form. 
 
Digitized by the Internet Archive — 
 in 2022 with funding from , 
 Princeton Theological Seminary Library 
 
 https://archive.org/details/personalitiesofsOOfrey 
 
INTRODUCTION 
 
 This study was undertaken as a solution not to a practical prob- 
 lem, but to one of a general psychological nature. Its purpose is to 
 add to our knowledge of adult human personality. It was under- 
 taken with the conviction that the study of trends of personality 
 and agreements and disagreements in the reactions of adults to 
 their environment over a period of time, was fundamental to and 
 afforded richer and more valuable fields for psychological inquiry 
 than the analysis of momentary mental states or acts. 
 
 Specifically, this study seeks to determine the personality cor- 
 relates of interest and ability (1) in handling or motivating human 
 beings, and (2) in handling machines or inanimate objects. What 
 is the interrelationship of the differences in personality, and to 
 what extent do these differences connote types or extremes of per- 
 - sonality ? 
 
 In order to arrive at a solution of this problem, comparisons 
 were made between salesmen on the one hand and students of en- 
 gineering and mechanics on the other. Differences were studied by 
 the use of tests, interest and personal history records, and ratings. 
 Simple group comparisons were made in these measures. An ad- 
 ditional statistical method, involving the intercorrelation of these 
 measures, was adopted in order to arrive at a consistent picture of 
 the extremes of divergence between the two groups of subjects. 
 
 The results of the group comparisons lead to the conclusion that 
 those who are interested in dealing with human beings (the sales- 
 men) are differentiated from the students of mechanics by greater 
 social ability, credulity or suggestibility, adaptability, excitability, 
 self-confidence, talkativeness, present-mindedness, and good-na- 
 ture. The students of mechanics show a greater tendency to be self- 
 conscious, careful of details, inhibited and cautious, reticent, ab- 
 sent-minded, and glum. They are more capable.of making fine co- 
 ordinations, and excel in slow, painstaking, and accurate effort. 
 There is no evidence for two corresponding sharply defined types 
 of personality, but evidence points to the utility of recognizing “ex- 
 
vi INTRODUCTION 
 
 tremes,” those few at the social extreme possessing all the traits of 
 the completely socially adapted person, and those at the mechan- 
 ical extreme possessing all of the traits of the person who is com- 
 pletely devoted to machines or inanimate objects. Most individuals 
 will deviate from an absolutely neutral state toward one or the 
 other extreme, depending upon which of the two opposed classes 
 of traits is dominant in the make-up of their personality. A use of 
 this distinction is provided in the present study by the presentation 
 of a composite scoring method which discriminates effectively be-- 
 tween the socially adapted and the mechanically adapted. | 
 
 At a later date the writer expects to show the relation of these 
 results to the concepts of the introvert and the extrovert, which 
 have been taken over by many psychologists. There is a remarkable 
 correspondence between the traits described in the theoretical liter- 
 ature as native to the introvert and the extrovert and those which 
 differentiate the mechanically inclined from the socially inclined, 
 the mechanically inclined corresponding to the introvert and the 
 socially inclined to the extrovert. The greater objectivity of a dis- 
 tinction based on ability to motivate people and ability to handle 
 machines warrants the use of this criterion for distinguishing the 
 extrovert from the introvert, rather than the criterion of facing | 
 reality or withdrawing fromit. 
 
TABLE OF CONTENTS 
 
 CHAPTER |, The Problem and Modes of Attack 
 The Problem 
 Procedure 
 Subjects 
 Forms Used 
 Cuapter II. Group Comparisons Based on Test Scores 
 The Method 
 Test VI. General Intelligence 
 Test IX. Will-Temperament 
 Directed Association Test 
 Series 12. Kent-Rosanoff Free Association Test 
 Summary 
 CuaptTer III. Group Comparisons Based on Answers to 
 Questionnaires 
 The Method 
 Description of the Questionnaires 
 Results 
 Summary 
 CHAPTER IV. Group Comparisons Based on Ratings 
 The Method 
 Results 
 Summary 
 Cuapter V. Intercorrelations of Tests and Ratings 
 The Theory of General Ability 
 Method Employed in this Study 
 Results 
 Summary 
 CHAPTER VI, Application of Results in a Total Score 
 The Method: 
 Results 
 Summary 
 CuapTer VII. Summary and Conclusions 
 BIBLIOGRAPHY 
 
 OW # 
 

 
GHA DER | 
 
 THE PROBLEM AND MODES OF ATTACK 
 
 THE PROBLEM 
 
 The great complexity of behavior of the adult individual seems 
 to offer an almost insurmountable barrier to systematization, yet 
 some sort of an approach must be made toward a classification of 
 people if psychology is to advance toward its goal of the prediction 
 and control of human behavior. The approach to this goal through 
 the study of instincts has failed, for the moment at least, to provide 
 materials for the study of adults; if the reader is in doubt, he need 
 only review the teeming literature on the subject and note the con- 
 tradictory statements and ex-cathedra pronouncements with which 
 it abounds. Genetic studies of instincts reveal little for the under- 
 standing of adults; the instincts become so disguised and are ex- 
 pressed in such a multitude of ways as to render any instinctive 
 explanations of conduct mere broad generalizations. Factors other 
 than instinct, such as immediate heredity and early environment, 
 are also important in determining the psychological make-up of 
 the adult. If-we are to interest ourselves in adult psychology, we 
 must take the adult as he is, the resultant of a number of forces 
 acting upon him. We must proceed to classify adults as to both 
 qualitative and quantitative differences, just as the zoologist classi- ° 
 fies his specimens into phyla, classes, orders, and so forth. Adults 
 have been classified on scales ranging from genius to idiocy, and 
 from sanity to insanity, presumably quantitative and qualitative 
 differentiations. The classifications of adults into types of person- 
 ality’ has met with indifferent success. 
 
 A certain common-sense attitude toward types of personality 
 cannot be disregarded. Whether as psychologists or laymen, we 
 1 By personality I mean the sum-total of the reaction tendencies of an individual 
 
 toward his fellow men, as opposed to his reactions toward inanimate objects. 
 Character is personality judged by the moral standards of a particular group. 
 
2 MAX FREYD 
 
 speak of typical ministers, typical pugilists, typical ward bosses, 
 and so forth. Such individuals constitute types in the sense that 
 they had common interests which led them into the same line of 
 work, where their training put the finishing touches of uniformity 
 on them. While there may not be fundamental psychological dif- 
 ferences between people belonging to diverse types, the differences 
 are important, popularly recognized, and relatively permanent. 
 
 In approaching the problem of human types from the psycho- 
 logical standpoint, we are immediately confronted with a variety 
 of difficult questions. Is a type distinguished by one central trait, 
 or by acluster of traits? In how many traits must individuals agree 
 in order to belong to the same type? How close a similarity in any 
 given measured trait shall be accepted as a criterion of agreement? 
 How can the traits be best measured ? How important relatively are 
 heredity and environment as factors which place men in types? 
 What is the final proof of the existence of types? The difficulties 
 pointed out show the need for experimental evidence. | 
 
 The common psychological point of view is against the existence 
 of types. If we set up as our criterion of the existence of types that 
 the individuals belonging to them shall be similar in even the re- 
 motest trait, we may as well let the question of types rest as far» 
 as any practical applications of our findings are concerned, for the 
 number of persons falling into types if this criterion were main- 
 tained would be infinitesimal. If a limited number of traits were 
 accepted as the distinguishing marks of a type, the most favorable 
 place to find types would be among those engaged in the same kind 
 of work. If it can be shown that individuals engaged in the same 
 occupation are not only in agreement in their abilities, interests, 
 and personal history, but also possess the same traits of personali- 
 ty, the results will not only be of theoretical interest, but will be of 
 immediate application in vocational guidance. 
 
 For the purpose of investigating the existence of types among 
 vocationally selected people, the writer has made a study of dif- 
 ferences between salesmen on the one hand and students of en- 
 gineering and mechanics on the other. These two opposed groups 
 are characterized most prominently by a social inclination and an 
 ability to motivate others in the case of the salesmen, and a primary 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 3 
 
 interest in mechanics on the part of the students of engineering 
 and mechanics. 
 
 PROCEDURE 
 
 A very good statistical consideration of psychological types is to 
 be found in the third volume of Thorndike’s Educational Psy- 
 chology (20). He outlines two ways of discovering types: 
 
 “The first is by direct measurement of individuals in toto, and 
 of their differences 
 
 “The second way . . . is by discovering what each amount 
 possessed of any trait implies concerning the individual’s condition 
 in other traits.”” (pp. 347-348) 
 
 In a detailed consideration of these methods and the results ob- 
 tained with their use, Thorndike comes to the conclusion that the 
 data at hand do not warrant the statement that types exist; that 
 the most probable condition of an individual is one of mediocrity 
 in most traits, with slight variations toward extremes in the re- 
 mainder of the traits. His main argument against the existence 
 of types is that distributions of various traits or combinations of 
 traits are unimodal, whereas if types existed each type would be 
 represented by a mode in the distribution curve. The unimodal dis- 
 tribution curves argue for a single type—mediocrity. 
 
 Regarding the first method, Thorndike states that no individual 
 has ever been, measured in toto. Even in cases of the most thorough 
 measurements no great similarities between individuals have been 
 found. The difficulty of the first method leads us naturally to the 
 second. Why measure all the individual’s traits if one trait may 
 stand for the amount possessed of any of a number of traits which 
 intercorrelate highly? Thorndike’s conclusion on this point, based 
 on such data as he could obtain, is as follows (p. 362): 
 
 “Tt is very, very hard to find any case of a negative correlation 
 between desirable mental functions. Divergence toward what we 
 vaguely call better adaptation to the world in any respect seems to 
 be positively related to better adaptation in all or nearly all re- 
 spects. And this seems especially true of the relations between 
 original capacities.” 
 
 A great deal hinges on the use of the word “desirable”; and in 
 
6 MAX FREYD 
 
 SUBJECTS 
 
 The men who served as subjects for this study fall into two 
 general classes : those who are mechanically inclined and those who 
 are socially inclined. Each of the classes may be subdivided into 
 primary and secondary groups. A wide series of test measure- 
 ments, answers to questionnaires, and ratings were obtained on the 
 two primary groups. Where results were available on the second- 
 ary groups they were used as checks on the results obtained with | 
 the primary groups. 
 
 In order to obviate any possible sex differences, only men were 
 used in this study, with the exception of a very small percentage of 
 one of the secondary groups. 
 
 Primary Mechanical Group. This group consisted of thirty 
 seniors in the College of Industries at Carnegie Institute of Tech- 
 nology, tested in the fall of 1921. These men were in the fourth 
 year of a curriculum designed to prepare them for executive posi- 
 tions in industry. Their courses were practical, and to a large ex- 
 tent consisted of laboratory work in the occupations in which they 
 aimed to become executives. The departments of instruction in- 
 cluded the following: Building Erection and Estimating; Elec-— 
 trical Equipment and Construction; Cost and Factory Account- 
 ing; Machinery Production; Power Plant Operation and Manage- 
 ment; Printing Industry; and Heating, Ventilating, and Sanitary 
 Equipment. While these men are prepared to become industrial 
 executives, the problem of motivating others forms a very small 
 proportion of their duties. It is safe to say that their success is 
 largely dependent on their technical information and ability, and 
 that their interests and abilities are predominantly technical and 
 mechanical, rather than social. Ability to motivate others does not 
 enter into their tasks in the sense that it enters into the work of the 
 salesman. Their prestige as executives will rest mainly on their 
 technical background. To motivate others they need only give or- 
 ders. The salesman, on the other hand, requires a complex social 
 ability in order that he may motivate others, and can in no sense 
 rely on authority in accomplishing that motivation. He must rely 
 on his own persuasive ability and personality, and must be actuated 
 by an interest in motivating people. In their daily behavior the in- 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 7 
 
 dustries seniors displayed markedly the fact that their abilities 
 were overwhelmingly mechanical rather than social, in comparison 
 with the groups of salesmen who were studied. 
 
 Primary Social Group. The social groups consisted in all cases 
 of salesmen. The primary group was composed of thirty men 
 chosen from the School of Life Insurance Salesmanship, Fall 
 Class, 1921-22, at Carnegie Institute of Technology. This school 
 consists of a course of eleven weeks’ duration, which is repeated 
 three times a year. It draws representative men with sales interests 
 from the various corners of the country. The thirty men chosen 
 out of the class of approximately sixty were those who stood high- 
 est in the tests developed by Ream (17) for the prediction of life 
 insurance salesmanship. The probability is high that these men 
 will sell more life insurance than the remaining thirty men in the 
 class. To some extent the tests used in the selective process were 
 such as to cause wider divergences between this group and the 
 primary mechanical group than would be expected if no selective 
 process were used. In most cases, however, the results of secondary 
 groups were available as checks on our differences. The sales 
 primary group seems well fitted to represent individuals whose 
 interests and abilities are mainly social rather than mechanical in 
 nature. The commodity which they sell is intangible and depends 
 for its appeal largely on the personality and persistence of the sales- 
 man. These salesmen have to make as many as forty calls in order 
 to make a sale. In their social contacts the members of this group 
 displayed a more pleasing personality than the industries seniors. 
 There were, of course, a few exceptions, but the statement holds 
 for the majority of the group. 
 
 Secondary Mechanical Groups. The following groups of me- 
 chanically inclined men were used in obtaining additional evidence 
 on the results obtained with the primary group: 
 
 Forty-three seniors in Mechanical Engineering, Class of 1922, 
 Case School of Applied Science, tested in December, 1921. 
 
 Seventeen seniors in the College of Industries, Class of 1921, 
 Carnegie Institute of Technology, tested in May, 1921. 
 
 Eighteen seniors in Mechanical Engineering, Class of 1921, 
 Carnegie Institute of Technology, tested in May, 1921. 
 
8 MAX FREYD 
 
 Nineteen seniors in Commercial Engineering, Class of 1921, 
 Carnegie Institute of Technology, tested in May, 1921. 
 
 Secondary Social Groups. The following sales groups provided 
 results which were used to supplement the results obtained with 
 the primary sales group: 
 
 Thirty men, constituting the remainder of the School of Life 
 Insurance Salesmanship, Fall Class, 1921-22, Carnegie Institute of 
 Technology, tested in October, 1921. 
 
 Seventy-six members of the School of Life Insurance Salesman- 
 ship, Spring Class, 1920-21, tested in April, 1921. ; 
 
 Fifty members of the School of Life Insurance Salesmanship, 
 Winter Class, 1921-22, tested in January, 1922. 
 
 One hundred and thirty-five life insurance salesmen with the 
 Edward A. Woods Company, of Pittsburgh, tested in July, 1921. 
 
 Eighty-one salesmen with the Burroughs Adding Machine Com- 
 pany. | 
 Twenty-one salesmen with the Cleveland Discount Company, 
 mainly bond salesmen, tested in November, 1921. 
 
 Twenty-four salesmen with the Harrison Real Estate Company, 
 of Buffalo, New York, mainly bond and real estate salesmen, tested 
 in November, 1921. 
 
 Forty-nine salesmen with the Procter and Gamble Company, 
 tested in January, 1922. | 
 
 An additional group on whom limited results were available con- 
 sisted of thirty-three seniors in the Westinghouse Technical Night | 
 School, Class of 1921, East Pittsburgh, Pennsylvania. These men 
 correspond roughly to the Industries groups, but the situation is 
 complicated by the fact that among them were numbers of sales- 
 men employed by the Westinghouse Company who were receiving 
 training in technical work in order to enable them to sell more ef- 
 fectively. Very few of the men indicated on their test blanks the 
 positions they occupied, and we are consequently at a loss as to the 
 inclinations of the majority of the men. 
 
 Forms UseEp 
 
 For purposes of comparing the groups several tests and ques- 
 tionnaires and a rating scale were used. A number of factors op- 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 9 
 
 erated to limit the scope of the program of forms, among which 
 were: the limit of two hours placed on the program; the fact that 
 only group tests could be used; the range of test forms used in the 
 Bureau of Personnel Research; and the availability of results with 
 the use of the same forms on the secondary groups. 
 
 Tests. While the principal stress is laid on differences in person- 
 ality, an intelligence test was included as a check on the factor of 
 general intelligence in the group differences found. The test used 
 for this purpose was Bureau of Personnel Research Test VI, a 
 spiral-omnibus test with items chosen from the Army Alpha Ex- 
 amination (9g). The test was given with a fifteen minute time limit. 
 
 A second test was Bureau Test IX, a group modification of the 
 Downey Will-Temperament Test (4, 16). This test series consists 
 of eleven parts, yielding with the use of ratios, twelve scores. This 
 test series was originated by Downey for individual testing of a 
 -number of volitional and temperamental traits, using modifications 
 of-handwriting for the purpose. In the present study use was made 
 of the fourth edition of a group modification of the test developed 
 by the staff of the Bureau of Personnel Research. 
 
 A third test was the Kent-Rosanoff free association test ar- 
 ranged for group testing. In this form, instead of giving his re- 
 sponse orally, the subject responds by writing the first word called 
 to mind by the stimulus word. The stimulus words are printed in 
 a list; one signal serves to start the individual on the list and one 
 to stop him. The chief interest in the test lay in the coefficient of 
 commonality of response, and in the classification of responses 
 made by Wells (22, 24). 
 
 A fourth test, which I have chosen to call a Directed Association 
 Test, was intended to measure ability at directed thinking, the sup- 
 position being that this ability could be measured by the rapidity 
 and brevity with which an individual completed a chain associa- 
 tion between two dissimilar words. 
 
 Questionnaires. A number of exhaustive questionnaires cov- 
 ering items of personal history, interests, attitudes, etc., were used 
 in order to supplement the measures yielded by the tests. Material 
 for. these questionnaires was drawn largely from previous lists pre- 
 pared in the Bureau of Personnel Research, and from the psychi- 
 
10 MAX FREYD 
 
 atric and psychoanalytic literature. The fields covered by the sev- 
 eral questionnaires are roughly the following: 
 
 Questionnaire 1. Early Personal History. 
 
 Questionnaire 2. Vocational Interests. 
 
 Questionnaire 3. Avocational Interests. 
 
 Questionnaire 4. Likes and Dislikes. 
 
 Questionnaire 5. Beliefs. 
 
 Rating Scale. A graphic rating scale (6) was constructed in 
 order to measure a number of traits some of which presumably 
 were brought out by the tests and questionnaires. With the use of 
 the same scale, self-ratings and ratings by friends and acquaint- 
 ances were obtained, and the groups compared in each of the traits 
 on the scale. The measurements yielded by this scale were also used 
 as a partial check on the validity of the tests, and as materials 
 for the intercorrelation of traits in discovering constellations of 
 traits. 
 
 The entire series of measures was obtained on the two primary 
 groups and on the seniors in Mechanical Engineering at Case 
 School of Applied Science. Measures of the remaining groups 
 were incomplete, lacking in every case the ratings, the Directed As- 
 sociation Test, and Questionnaires I, 3, and 5. In some groups 
 additional measures were missing. 
 
 More detailed discussion of the forms used will be found in the 
 following three chapters. 
 
CHAPTER II 
 
 GROUP COMPARISONS BASED ON TEST SCORES 
 
 THe METHOD 
 
 The first method of studying personality types, that of making 
 group comparisons in distributions of abilities, offers no great 
 theoretical or statistical difficulties. This procedure was followed 
 in the present instance in order to form an estimate of the differ- 
 ences between the two primary groups in the selected measures of 
 capacities. Where test results were available on primary and sec- 
 ondary groups, these were combined. The central tendencies of the 
 salesmen and of the mechanics students and the standard error of 
 the difference in central tendencies were then computed, enabling 
 us to place a statistical estimate on the reality of the difference in 
 ' central tendencies of the two groups. Where test results were avail- 
 able only on the primary groups, the less rigid statistical device 
 was employed of computing the range of scores of the middle 50 
 per cent of each group in each test (interquartile range), and plac- 
 ing a rough observational estimate on the reliability of the differ- 
 ences. Statistically, if the median of one group is coincident with 
 the first or third quartile of the other group, or, if 25 per cent of 
 one group lie beyond the median of the other group, the group dif- 
 ference is quite significant; with thirty cases in each group the 
 probability is approximately one in fifty that with further sampling 
 the medians will coincide, or that instead of 25 per cent, 50 per cent 
 or more of the one group will lie beyond the median of the other 
 group. With one hundred cases the probability is reduced prac- 
 tically to zero. Where 25 per cent or less of one group lie beyond 
 the median of the other, we shall speak of the difference in that 
 ability as very significant. A rough estimate must be placed on dif- 
 ferences which fall short of this criterion. 
 
 We cannot, in the absence of correlation coefficients with reliable 
 
12 MAX FREYD 
 
 ratings, state dogmatically what function is measured by each test, 
 but we can draw a common-sense analogy between the functions 
 involved in the performance of the test and certain psychological 
 concepts. Outlines of experiments for laboratory courses in psy- 
 chology (the cradle of psychological tests) do not hesitate to name 
 the functions for which each experiment is intended, such as atten- 
 tion, volition, suggestion, association, and the like. In the following 
 discussion of results, I have given the interpretation placed on the 
 tests by the originators or those who have subsequently made use © 
 of them. In Chapter 5 mention is made of correlation coefficients 
 obtained between tests and ratings, but in as much as the ratings 
 were secured under conditions more suitable for rough group com- 
 parisons than for correlations with tests, they should not be given 
 too much weight in this connection. The rating scale, moreover, 
 does not exhaust the number of possible abilities: there is always 
 the possibility that the test would correlate most highly with a trait 
 not on the scale. These coefficients will be mentioned in connection 
 with the interpretations of the tests. 
 
 The comparisons were first made between the two primary 
 groups: seniors in the College of Industries at Carnegie Institute 
 of Technology, and thirty salesmen selected from the School of 
 Life Insurance Salesmanship. Where the differences seemed sig- 
 nificant and the same measures were at hand on secondary groups 
 listed in the preceding chapter, the additional results are given.* 
 
 Test VI. GENERAL INTELLIGENCE 
 
 Group comparisons in this test are interesting mainly in so far 
 as they corroborate the aim to eliminate the factor of general in- 
 telligence in the group comparisons. The test consists of 184 items 
 selected from six of the tests in the Army Alpha Examination, 
 namely, arithmetical problems, synonym-antonym, disarranged 
 sentences, number series completion, analogies, and information. 
 The directions for all the different forms of tasks are assembled 
 at the beginning of the test. In the body of the test the items are 
 so arranged as to become progressively more difficult. Each type 
 
 1 The test figures for the secondary groups in Test IX were supplied by Mr» 
 L. D. Anderson from unpublished data. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 13 
 
 of item occurs at given intervals. The term “spiral-omnibus” has 
 been applied to this test to distinguish it from the earlier “‘cycle- 
 omnibus” tests in which the various types of task recurred without 
 any progressive increase in difficulty. Its superiority over the army 
 form lies in the greater ease of administering this form and the 
 briefer time required for testing. 
 
 In administering the test, four minutes are allowed for reading 
 the directions. At a signal the subjects commence the test and work 
 uninterruptedly for fifteen minutes. Originally twenty-five minutes 
 were allowed, but correlations with scores in the fifteen minute 
 period were so high (++.96 and +.97) as to warrant using the 
 shorter time limit (9). The different types of items are scored in 
 the same manner as in the Army Alpha Examination, but a sepa- 
 rate score for each type of item is not obtained. 
 
 The purpose of this test is to measure an individual’s mental 
 capacity or general intelligence. “By intelligence is meant ability to 
 . think quickly, accurately, and independently; to comprehend new 
 problems; to meet new and difficult situations; to understand and 
 to be able to carry in mind complicated directions, etc.” (14, p. 
 425) 
 
 Below are given the interquartile ranges of the two primary 
 groups and the Case School Seniors in Test VI. 
 
 OF a, Q, 
 
 ’ Succ. |. 1. salesmen 107.5 117.6 130.0 
 Industries seniors 07.5 110.3 130.3 
 Mech. Eng. seniors 105.0 115.0 130.0 
 
 The differences are not so great as to place the groups on dif- 
 ferent mental levels. Test VI correlates uniformly low with the 
 other tests used in this study. We may conclude, therefore, that 
 we have test results on factors other than conceptual intelligence. 
 
 Test [X. WILL-TEMPERAMENT 
 
 Although the Downey Will-Temperament Test is of compara- 
 tively recent advent in the field of psychological tests, the litera- 
 ture upon it is already quite extensive. A thoroughgoing discussion 
 of the test is, therefore, not in place in this study. The present dis- 
 cussion will be limited to a brief exposition of the form of the 
 
14 MAX FREYD 
 
 Will-Temperament Test used in this study, and to a reconsidera- 
 tion of the traits which it aims to measure. 
 
 The Will-Temperament Test, as originated by Downey (4), is 
 a series of twelve individual tests intended to measure volitional 
 characteristics, so arranged that when the deciles are plotted the 
 volitional pattern or profile of the subject becomes apparent, and 
 his classification with regard to three general volitional types may 
 be determined; the “hair-trigger” type; the wilful, aggressive 
 type; and the sluggish, painstaking type. The extent of each of 
 these three tendencies in his volitional make-up is supposed to be 
 measured by three or four tests. Handwriting is the medium by 
 which Downey arrives at this analysis, but in so doing 2G makes 
 no graphological assumptions. 
 
 “It may be taken for granted that will-functions must be tested 
 largely through some form of motor reaction . . . The motor 
 activity required for the present purpose should be a common one 
 and one which leaves behind it a permanent record. The one I 
 have chosen for exploitation is handwriting; I have introduced 
 modifications of it such as speeded, retarded, blocked, disguised, 
 and automatic writing . . . In addition to writing activity I 
 have included tests involving choice in which the intellectual factor 
 is subordinate to the impulsive one. . . . The choice of par- 
 ticular tests has been motivated largely by previous experimenta- 
 tion on handwriting and, to a lesser degree, on muscle-reading. 
 ; It should be emphatically stated that the present exploita- 
 tion of handwriting makes little use of graphic individuality and 
 no use whatever of graphological assumptions. It utilizes the writ- 
 ing movement merely as a convenient method for study of compli- 
 cations of motor impulses.” (4, pp. 8-9) 
 
 Downey’s original test series was intended for individual test- 
 ing, but she as well as other psychologists have arranged group 
 modifications of the series. The pioneer in this field in the Bureau 
 of Personnel Research is H. A. Richmond. The group form of the 
 test in use by the Bureau has passed into its fifth edition and has 
 been given to over two thousand persons. 
 
 The fourth edition of the Bureau modification of the test series 
 was used in this study. It consists of eleven parts, which with the 
 use of ratios yield twelve separate scores. The main difference be- 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 15 
 
 tween this series and the series for individual testing lies in the 
 substitution of a time limit for a work limit. Although the parts 
 are administered in their numerical order, the scores are obtained 
 and arranged according to Downey’s three-fold classification, 
 which has been revised in the light of the intercorrelations of the 
 parts. The order in which the scores are presented is as follows: 
 
 Parts I and Io. 
 Part 4, ratio. 
 Part 3. 
 
 Farts.0: 
 
 Part 6, disguises. 
 
 Hatiwant 
 Part 11. 
 Part 9, ratio. 
 
 Parts 2: 
 Hare 
 Part a7 
 Part 6, letters. 
 
 The first five scores are designed to measure speed of action and 
 decision, the next three aggressiveness and impulsiveness, and the 
 last four capacity for slow, painstaking effort and for making cer- 
 tain motor coordinations. The scores will be discussed in the above 
 order, rather than in their numerical order in the test booklet. In 
 designating the functions measured by these parts, I have relied on 
 the interpretations of Ream (16) and Downey (4).? Some addi- 
 tional scores were obtained and are presented at the conclusion of 
 this discussion. 
 
 Parts 1 and 10. Normal Speed of Writing. This is measured by 
 averaging the number of letters written in Parts 1 and 1o. In each 
 of these parts the subject is required to write the words “United 
 States” repeatedly in his usual style and at his usual speed. Thirty 
 seconds are allowed for each part. This score correlates —.37 with 
 ratings on cautiousness. (This figure and the coefficients which 
 
 will be given in connection with some of the remaining parts and 
 2 More complete descriptions of the parts are to be found in two publications 
 by Ream (16, 17). Copies of the test booklet may be obtained from the Bureau 
 of Personnel Research, Carnegie Institute of Technology. . 
 
16 MAX FREYD 
 
 tests, were abstracted from the unpublished eighth table mentioned 
 in Chapter 5.) 
 
 The results reveal small group differences. The following table 
 shows the results for the two primary groups. The figures for the 
 secondary groups are practically identical with these. 
 
 Q, Q, Q, 
 Succ. 1. i. salesmen 54 65 69 
 Industries seniors 61 66 70 
 
 Part 4, ratio. Ratio of Normal to Speeded Writing. In Part 4 
 the subject is required to write the words “United States” re- 
 peatedly as quickly as he can. Sixty seconds are allowed, and the 
 score is in terms of the number of letters written. Part 4, ratio, is 
 obtained by dividing the number of letters written in Part 10 (nor- 
 mal speed) by the number of letters written in Part 4 (greatest 
 speed.) The ratio is supposed to be an index of the subject’s ten- 
 dency to work at his greatest possible speed in his ordinary tasks, 
 in other words, his freedom from load (4, p. 20). 
 
 Group differences are practically non-existent in this part. The 
 following results are practically duplicated by the secondary 
 groups. 
 
 2h Q, Q, 
 Succ. 1. i. salesmen 35 39 43 
 Industries seniors 36 a7 43 
 
 Part 3. Speed in Checking Better Traits. In Part 3 thirty-six 
 pairs of opposite traits are presented, as careful-careless, accurate- 
 inaccurate, effeminate-manly, etc. The subject is required to go 
 down the list and check the one trait in each pair which he believes 
 is the better in most circumstances. Forty-five seconds are allowed, 
 and the score is in terms of the number of traits checked. 
 
 Although the type of decision involved is not entirely objective, 
 
 the personal element has been eliminated as far as possible. By an. 
 
 objective decision is meant one that is free from any reference to 
 the reagent himself, not necessarily one that is free from internal 
 facilitations. That the test calls for no great individual judgment 
 is shown by the fact that the men in both primary groups agree 
 consistently in their checks. No group differences appeared in this 
 test, taken as a measure of social standards. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 17 
 
 Results are given below and in the table at the end of this chap- 
 
 fcr. 
 Q, Op: Q, 
 
 Mechanical groups 25.0 30.0 34.0 
 Industries seniors 26.0 29.0 33.0 
 Misc. salesmen 18.3 23.0 29.6 
 Life ins. salesmen 19.4 25.4 B12 
 Succ. 1. i. salesmen 21.0 27.0 34.0 
 
 The second and fifth groups in the above table are the primary 
 groups. The “mechanical groups” include the primary and sec- 
 ondary groups; that is, the figures for the second group above are 
 also contained in those for the first group, but form a small pro- 
 portion of that group. Similarly, the thirty successful life insur- 
 ance salesmen are included with the “life insurance salesmen,”’ 
 forming a small proportion of that group. The Westinghouse 
 Technical Night School seniors are omitted in giving these test re- 
 sults. 
 
 There is a pronounced tendency for the mechanically inclined 
 people to show greater speed in this test than the salesmen. All 
 men in the primary and secondary groups considered, the chances 
 are I in infinity that a real difference does not exist between the 
 mechanically inclined and the salesmen. 
 
 Part 9. Speed in Checking Personal Traits. Here the same list 
 as in Part 3 is presented but the decision is made subjective by re- 
 quiring the reagent to indicate which of each pair he himself pos- 
 sesses. Sixty seconds are allowed, and the score is the number of 
 traits checked. 
 
 In this test we find that the differences between our primary 
 groups are reversed when the secondary groups are taken into 
 consideration. The primary sales group checks more traits than 
 the primary mechanical group, but with all cases considered the 
 difference is reversed and we find the mechanical groups differing 
 significantly from the sales groups in that they check more traits. 
 
 Results are given below and at the end of this chapter. 
 
 Q, Q, Q, 
 
 Mechanical groups 20.7 25.5 31.5 
 Industries seniors 21.0 24.0 30.0 
 Misc. salesmen - 16.6 21.5 26.5 
 Life ins. salesmen 16.5 21.0 26.5 
 
 Succ. 1. i. salesmen 22.0 26.0 30.0 
 
18 MAX FREYD 
 
 Part 6, disguises. Ability to Disguise Handwriting. The direc- 
 tions for Part 6 are as follows: 
 
 Write the words “United States” in the space below, trying to 
 disguise your handwriting in as many ways as possible and as 
 much as you can. Try out any disguise you can think of but do not 
 print. Take as much time as you need and copy the words as many 
 times as necessary. 
 
 Keep trying until you feel that you have made a copy that even 
 a handwriting expert could not identify as yours. 
 
 Three minutes are allowed for this part. The score for Part 6, 
 disguises, is the number of different successful disguises. In scor- 
 ing the part, a stencil is used on which are examples of successful 
 and unsuccessful disguises. One great difficulty with this part is 
 the lack of an objective method of scoring. Downey calls the 
 equivalent test in her series a test of flexibility, a term intended to 
 cover the ability to modify one’s routine activities, to dramatize, to 
 exercise imagination (4, p. 22). 
 
 Results are given below and at the end of this chapter. 
 
 Q, Q, Q; 
 Mechanical groups 27 2.9 4.0 
 Industries seniors ims 2.2 2.8 
 Misc. salesmen 2.3 3.3 4.6 
 Life ins. salesmen 2.4 3.5 4.5 
 Succ. 1. i. salesmen 2.4 3.5 4.3 
 
 The table above is presented graphically in Fig. 1. There is a 
 
 Industries Seniors meee accor 
 Misc. Salesmen RUSS ES 
 Life Ins. Salesmn SE TE TEE EES 
 30 Life Ins. Salesman Guim 
 t] t 4 e ’ ‘ ’ 
 Score 1.5 2.0 2.5 3.0 3.5 4.0 4.5 
 HG. I 
 
 Interquartile Ranges of Scores in Part 6, disguises, Test IX. 
 Vertical lines indicate medians 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 19 
 
 significant tendency for salesmen to excel in this sort of perform- 
 ance. All cases considered, the probability is 1 in 2301 that a real 
 difference does not exist. 
 
 Part 5. Enlargement of Writing under Distraction. The page 
 for Part 5 has a series of three numbered dots, one below the 
 other. The subject is required to place his pencil on the first dot 
 and write the words “United States” once while looking at the 
 ceiling. He then repeats this with the second dot. These two per- 
 formances do not enter into any scores. The directions with regard 
 to the third dot are in brief as follows: 
 
 This time it will be much more difficult. Listen carefully. You 
 are to do two things: write “United States’ repeatedly and at the 
 same time count the number of taps I make on the table. Write one 
 phrase just below the other. Watch me while writing and count- 
 ing, and when the tapping stops quit writing, and record the num- 
 ber of taps. Ready. Place your pencil on dot number 3. 
 
 The number of taps made is immaterial, and no record is made 
 of the accuracy of the guesses. The taps are made about one per 
 second, an occasional tap being skipped to avoid regularity. At the 
 end of twenty-five seconds the tapping is stopped and the subject 
 is told to quit writing. The performance with regard to dot 3 is 
 then repeated farther down the page. 
 
 The score for Part 5 is the sum of two ratios. The first ratio is 
 obtained as follows: The longest “United States” in each of the 
 above two trials is measured in millimeters and the two lengths are 
 averaged. This is then divided by the length in millimeters of a 
 “United States” of average length in Part 1. The purpose of this 
 ratio is to determine the degree to which the subject has enlarged 
 his handwriting under distraction. The second ratio is obtained by 
 averaging the number of letters written in each of the above two 
 trials and dividing this average by the average number of letters 
 written in Parts 1 and 10. This is to determine the degree to which 
 the subject has speeded up under distraction. The sum of the two 
 ratios is the score for Part 5 (Enlargement under distraction). 
 
 This part is supposed to distinguish between impulsiveness or 
 
20 MAX FREYD 
 
 explosiveness of will as opposed to inhibition, or between the 
 choleric and the phlegmatic temperament. Downey says of her 
 parallel test: 
 
 “The utilization of change in size of writing under distraction 
 of attention as an index of motor impulsion was suggested by pre- 
 vious work both on automatic writing and on muscle-reading. The 
 tendency to magnify or to decrease handwriting size not only un- 
 der conditions of distraction, but even with such a simple change 
 in conditions as results from withdrawal of visual control in writ- 
 ing with the eyes closed is an individual characteristic of consider- 
 able significance. Extensive experiment with subjects had shown 
 me that it was quite possible to anticipate the value of an individual 
 as a guide in muscle-reading by his tendency to magnification of 
 graphic movement when inhibitory checks were removed. One’s 
 value as a guide in muscle-reading is in turn significant of the de- 
 gree to which one possesses the explosive in contrast to the in- 
 hibited temperament” (4, pp. 23-24). 
 
 Results are given below and at the end of this chapter. In the 
 table below the life insurance salesmen include only those attend- 
 ing the School of Life Insurance Salesmanship. The remaining 
 insurance salesmen were not given this test. 
 
 Q, Q, U, 
 Mechanical groups 1.78 2.02 2.23 
 Industries seniors 1.62 1.80 2.05 
 Misc. salesmen 1.82 2.04 2.24 
 Life ins. salesmen (schools only) 1.72 1.07 2.25 
 Suce. 1. i. salesmen 1.85 2.11 2.20 
 
 A very significant tendency for the primary sales group to excel. 
 in this test (show greater impulsion) is not corroborated by the 
 results of the secondary groups. 
 
 Part 11. Assurance in Visual Memory. A large chart with the 
 following characters on it is exposed for one and one-half minutes, 
 with instructions to study it carefully and try to remember it. 
 
 j IX N 
 F 5 any 
 VI d 8 
 
 At the end of this period the chart is withdrawn and the subjects 
 are given time to answer the questions in Part 11. The latter is in 
 form of a true and false statement test on the characters in ‘the 
 chart. The following is an illustration. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 21 
 
 1. There were eight capital letters in the chart. TRUE FALSE 
 The subjects are required to underline either TRUE or FALSE. 
 If they are absolutely sure of the answer they are asked to doubly 
 underline the word. There are fifteen such statements. The fol- 
 lowing paragraph appears after the last statement : 
 
 If you finish before time is called, you may check your accuracy 
 in the last three statements. The word FALSE should be under- 
 lined after statements 13, 14, and 15. If you have not done this 
 you are at liberty to change your answers. 
 
 As a matter of fact the word FALSE should be underlined 
 only after statement 13. The score for Part II is in terms of the 
 number of words doubly underlined after the statements and the 
 effectiveness of the false suggestion made in the paragraph quoted 
 above. If FALSE is not underlined after both statements 14 and 
 15, the score is 16, otherwise it is 0. Scores may therefore range 
 from o to 31. 
 
 This is intended to be a measure of assurance or resistance to 
 suggestion (16, p. 11). The higher the score the more assured the 
 subject is. There is a slight tendency for the person who scores 
 high in this test to be rated as unsociable and close-hearted. 
 
 Results are given below and at the end of this chapter. The table 
 below is presented graphically in Fig. 2. 
 
 Q, Q, Q, 
 
 Mechanical groups 22.5 30.8 31.5 
 Industries seniors 16.0 27.8 31.0 
 Life ins. salesmen (schools only) 12.2 15.6 28.5 
 Succ. 1. i. salesmen 11.0 15.3 22.0 
 
 There is a very significant tendency for the mechanical groups 
 to score higher in this test than the salesmen. All cases considered, 
 the probability is one in infinity that a real difference between the 
 two groups does not exist. 
 
 Part 9, ratio. Ratio of Objective to Subjective Speed of De- 
 cision. The score is the ratio of the number of traits checked in 
 Part 3 to the number of traits checked in Part 9. This score 
 has been treated as a measure of a self-consciousness by Ream 
 (16, p. 11). The score in Part 9 alone would seem at first 
 sight to be a measure of the trait, but this score must be in- 
 
22 MAX FREYD 
 
 Industries Seniors ee 
 Life Ins. Schools SE 
 30 Life Ins. Salesmen SACRE 
 ’ t ‘ i] i ( 
 Score 10 15 20 25 30 35 
 Hie, 2 
 
 Interquartile Ranges of Scores in Part 11, Test IX 
 
 terpreted in the light of an objective decision. In this way we 
 can determine the inhibiting effect of introducing the personal fac- 
 tor into a decision. If the ratio is low, the personal factor has not 
 changed the situation at all and the subject may be described as 
 free from self-consciousness. If the ratio is high, the person has 
 hesitated about personal judgments because, hypothetically, a com- 
 plex regarding himself has caused him to become embarrassed. The 
 same mechanism works to throw into relief complex-indicators in 
 the free association experiment. Our correlations show that the 
 person who gets a high score in the test (and is theoretically self- 
 conscious) is rated as lacking in self-esteem and adaptability. 
 
 Results are presented below and at the end of the chapter, and 
 in Fig. 3. 
 
 Be, Q, 42; 
 Mechanical groups 98 1.21 1.48 
 Industries seniors 96 1.19 1.36 
 Misc. salesmen 80 1.05 1.38 
 Life ins. salesmen 88 1.15 1.50 
 Succ. 1. i. salesmen 93 1.07 1.16 
 
 There is a very significant difference between the primary 
 groups, with the mechanics students obtaining the higher ratios 
 (more self-conscious). This difference is reduced very much with 
 the inclusion of the secondary groups. . 
 
 Part 2. Inability to Write Slowly. In Part 2 there are three 
 groups of words, each word being followed by a blank line. In the 
 first and second groups are five words of fifteen letters each, while 
 the third group has eight words of ten letters each. The subject is 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 23 
 
 Mechanical Groups nan nn I 
 Industries Seniors a ad 
 Miscel. Salesmen aL 
 Life Ins. Salesmen SR a 
 
 50 .Life Ins. Sdlesmen emanemees:/eerors 
 
 Scare 60 .90 1.00 1.10 1.20 1.30 1.40 1.50 
 Hie. 3 
 Interquartile Ranges of Scores in Part 9, ratio, Test IX 
 
 required to write each word on the blank line, as slowly as he 
 possibly can. Forty-five seconds are allowed for each of the first 
 two parts, and sixty seconds for the third part. The directions for 
 the three parts are similar, except that the exhortation to write 
 ' slowly becomes greater. The first two groups are ignored in scor- 
 ing, as they exist merely for the purpose of impressing the sub- 
 ject with the unusual type of performance that is demanded. The 
 score is the number of letters written in the last section. Downey 
 calls the equivalent test in her series a measure of ““power of motor 
 control and inhibition” (4, p. 22). 
 
 Results are presented below, at the end of the chapter, and in 
 Fig. 4. 
 
 Q, Q, Q, 
 
 Mechanical groups 2.0 6.0 10.5 
 Industries seniors 1.0 2.0 5.2 
 Misc. salesmen 5.0 11.5 25.5 
 Life ins. salesmen 3.5 7.3 ives 
 Succ. 1. i. salesmen 4.0 8.4 12.0 
 
 A very significant tendency exists for the mechanical groups 
 to write more slowly in this test. 
 
 Part 8. Speed of Writing in Small Space. Part 8 consists of six 
 short sentences (18 to 36 letters), each of which is followed by a 
 line with vertical bars at its ends. In each case the line is shorter 
 than the printed sentence. The directions are as follows: 
 
 Copy each of the sentences below as rapidly as possible on the 
 
24 MAX FREYD 
 
 Mechanical Groups a mae 
 Industries Seniors ieee 
 Miso. Salesmen Se ee ene nn 
 Life Ins. Salesmen ‘ACR NE NS aN TE 
 30 Life Ins. Salesmen wrorme}meaney 
 . e t 
 Soor e 64S aan "46 yy aot ae meee 
 Fic. 4 
 
 Interquartile Ranges of Scores in Part 2, Test IX 
 
 line after each sentence. Be careful not to let the writing extend 
 beyond the end of the line. 
 
 Forty-five seconds are allowed. 
 The score is the total number of letters written on the lines. Let- 
 
 ters which extend beyond the bars at the ends of the lines are not 
 counted. Downey says regarding the equivalent test in her series: 
 
 “This test impresses a double mental set, namely, adjusting one’s 
 movements to a limited space and doing it as rapidly as possible. 
 . . . Just what such ability to coordinate impulses indicates is 
 something of a question. My study of the individuals who make 
 particularly good records or who fail overwhelmingly suggests 
 that it serves as an index to what is called “keeping one’s head’; 
 ‘presence of mind’; perhaps orderly handling of a complex situa- 
 tion without previous practice” (4, pp. 18-19). 
 
 To the writer it seems that the test measures merely a skilled co- 
 ordination such as is required in certain of the finer mechanical 
 operations. This view is strengthened by the fact that the two 
 highest correlations of this test with ratings are +.37 with accu- 
 racy is work and +.40 with quickness in work. 
 
 Results are given below, at the end of the chapter, and in Fig. 5. 
 
 Q, Q, OF 
 
 Mechanical groups 83 93 105 
 Industries seniors 82.5 92.5 07 
 Misc. salesmen 54 68 83 
 Life ins. salesmen 63 75 90 
 
 Succ. 1. i. salesmen 75° 85 95 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 25 
 
 The mechanical groups do much better in this sort of task than 
 the salesmen. The differences between the primary groups are 
 small, but with the secondary groups considered, the differences 
 are very significant. 
 
 Mechanical Groups sean tienen 
 
 Industries Seniors 
 
 Misc. Salesman mL OC aes 
 Life Ins. Salesmen SEES 
 30 Life Ins. Salesmen Eo 
 ‘ 1 ' ' ! ' a 
 Score 50 60 70 80 90 100 110° 
 Fic. 5 
 
 Interquartile Ranges of Scores in Part 8, Test IX 
 
 Part 7. Carelessness in Copying.’ This part consists of three 
 handwritten sentences (fourteen letters each), written by differ- 
 ent persons. The instructions are to copy the sentences as exactly 
 as possible. Space for this purpose is provided between the sen- 
 tences. One minute and a half are allowed for this part. The score 
 is in terms of the number of divergences from the originals in such 
 details as dotting i’s, preserving the correct slant, closing the loops 
 of a’s, etc. The subject is scored only on the first two sentences, 
 and in each of these fifteen mistakes are possible, in other words, 
 there are fifteen details in each of these two sentences which are 
 observed for purposes of detecting mistakes. A stencil is used in 
 scoring which is laid over the test and has arrows which point to 
 these details, enabling the scorer to observe rapidly in each sen- 
 tence how many of these fifteen standard details have been over- 
 looked in copying the sentence. As in Part 6, disguises, the scoring 
 method is weak in that it is quite subjective. One error in the ad- 
 ministration of this test is the fact that speed is not taken into ac- 
 8 Although it does not concern this study directly, it is worth while noting that 
 when standings are plotted, as on the first page of the test booklet, high decile 
 
 standings accompany high numerical scores, except in Part 9, ratio, Part 2, and 
 Part 7, where the reverse is true. 
 
26 MAX FREYD 
 
 count. It was presumed that everyone would finish the first two 
 sentences in the time allotted, but this did not prove to be the case. 
 In the results given below, the difference between the primary 
 groups is reversed with the secondary groups. The difference be- 
 tween the primary groups was due to the fact that the primary 
 sales group wrote fewer words than the industries group. The 
 median number of words written by the former group was 16, the 
 latter group 25. If the test is rescored on the basis of the ratio of 
 the number of errors made to the number of errors possible in the 
 amount written (according to the scoring method), we find that 
 the two primary groups differ in the same direction as the sec- 
 ondary groups. These results will be given later. 
 
 The results obtained with the present scoring method are given 
 below, at the end of the chapter, and in Fig. 6. 
 
 Q o Q, 
 Mechanical groups 3.0 5.0 9.0 
 Industries seniors 28 2.9 4.5 
 Misc. salesmen 6.0 9.0 13.0 
 Life ins. salesmen 4.6 9.0 14.0 
 Succ. 1. i. salesmen 1.3 2.8 4.7 
 
 The differences between the mechanical groups and the salesmen 
 are very significant, and indicate that the salesmen show less care 
 for detail in such work than the mechanical groups. 
 
 Part 6, letters. Number of Disguises Attempted. The score is 
 the number of letters written in Part 6. The equivalent test in the 
 original Downey series is called a test of tenacity or perseverance. 
 
 Mechénical Groups rs 
 
 Industries Seniors aero ed 
 
 Misc. Salesmen SESS ERPS 
 Life Ins. Salesmen ae eS 
 
 50 Life Ins. Salesmen mmm 
 
 4 6 
 Fic. 6 
 Interquartile Ranges of Scores in Part 7, Test IX 
 
 to- 
 @- 
 K 
 ° 
 ~ 
 wo 
 ad 
 > 
 
 Score 0 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 27 
 
 The test involves persistence in attempting an indefinite task. Re- 
 sults are given below and at the end of the chapter. 
 Oe rem anerar t | 7°F 
 
 Mechanical groups 76 QI III.5 
 Industries seniors 65 83 105 
 Misc. salesmen 76 100 130 
 Life ins. salesmen 68 90 120 
 Succ. 1. i. salesmen 78 102.5 128 
 
 There are no important group differences in this test. There is 
 a slight tendency for salesmen to make a greater number of at- 
 tempts at disguising their handwriting. 
 
 In addition to the scores regularly obtained from the test series, 
 the following measures were obtained. 
 
 Agreements Between Checks in Parts 3 and 9. This score was 
 obtained at the suggestion of L. D. Anderson, as a possible meas- 
 ure of conceit. The score is the number of agreements between 
 the traits checked in the first column of Part 3 and the first column 
 of Part 9. In Part 3 we have the opinion of the reagent on the 
 most desirable traits to possess; in Part 9 he indicates which of 
 these desirable traits he himself possesses. The number of agree- 
 ments shows how desirable he considers his own traits. 
 
 Results obtained with the primary groups are given below. 
 
 Q, Q, Q, 
 
 Industries seniors He 15.0 15.8 
 Succ. 1, i. salesmen 12.5 13.8 16.0 
 
 The mechanical group shows a slightly greater number of agree- 
 ments than the sales group. 
 
 Traits Checked in Part 3. Group comparisons were made in the 
 frequencies with which the different traits in Part 3 were checked. 
 No important differences were found, the great majority checking 
 the traits in the socially acceptable way. This is added proof of the 
 objectivity of the decision made in the test. 
 
 Traits Checked in Part 9. Similarly, group comparisons were 
 made in the frequencies with which each trait in Part 9 was checked. 
 While not to be accepted as a self-rating, it is often the case that 
 under stress of time and under test conditions a person will give 
 a truer estimate of himself than he would if he had time to reflect 
 and rationalize. No startling differences were discovered, the great- 
 est ones being the following: 8 of the primary sales group checked 
 
28 MAX FREYD 
 
 “slow,” whereas 17 of the primary mechanical group checked that 
 trait; 18 salesmen and 12 industries students checked “quick’’; 9 
 salesmen and I industries student checked “cowardly” ; 17 salesmen 
 and 27 industries students checked “brave.” 
 
 Part 4. Speeded Writing. Scores were in terms of the number of 
 letters written in Part 4. This is a measure of the subject’s greatest 
 speed of writing. It is more or less a measure of coordination of 
 impulses. The following results were obtained with the primary 
 groups. 
 
 Qe Qa hens 
 Industries seniors 149 169 187 
 Succ. 1. i. salesmen I51I 163 175 
 The industries students show a slight superiority in speeded writ- 
 
 ing. 
 
 Part 7. Revised Scoring. This part was rescored for reasons 
 previously given. The revised scoring method is to obtain the ratio 
 of the number of errors made to the number of possible errors by 
 our scoring method, in the amount written. It leaves out of consid- 
 eration the third sentence, but very few work on it. Results are 
 as follows for the primary groups. 
 
 Q, Q, Q, 
 Industries seniors SG .24 rae 
 Succ. 1. i. salesmen 14 25 30 
 
 This reverses the differences found when Part 7 is scored in the 
 customary manner. The difference is slight, with the salesmen ex- 
 hibiting a greater tendency toward making errors. 
 
 DIRECTED ASSOCIATION TEST 
 
 This test was printed on two sheets, the first containing the di- 
 rections and the second containing the body of the test. The first 
 page is reproduced herewith. 
 
 DIRECTED ASSOCIATION TEST 
 EXAMPLE: 
 Column I 1st Solution ‘and Solution 
 
 NAIL NAIL NAIL 
 hammer rust 
 handle 
 wood 
 tree 
 roots 
 damp 
 
 WATER WATER WATER 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 29 
 
 EXPLANATION: 
 
 Note that in Column I there is no direct relationship between “NAIL” and 
 “WATER.” You are to write words in the blank space in the column so that 
 a continuous association is formed from “NAIL” to “WATER.” The example 
 has been worked for you in two ways. Look at the first solution. Here “NAIL” 
 suggests hammer,” “hammer” suggests “handle,” and so forth, “damp” finally 
 suggesting “WATER.” Thus a continuous association is formed from “NAIL” 
 to “WATER,” each word being suggested by the preceding one. The same 
 thing is done in the second solution, but with only one word. 
 
 On the following page you will find ten columns such as Column I. In each 
 column you are to form a continuous association from the first to the last 
 word. 
 
 No stated number of words is required in each list, but make your lists as 
 brief as possible. Work rapidly. 
 
 Do not turn page until told to do so. 
 
 The first and last words in the ten columns were as follows: 
 sun-faith, day-blotter, potato-cement, sky-fever, inch-black, jour- 
 ney-guilty, rifle-square, fun-question, sound-habit, copper-ignor- 
 ant. 
 
 -In giving this test the directions and examples on the first page 
 were read aloud to the subjects and any questions answered. Then 
 three minutes of continuous work were allowed on the test. The 
 scores used in comparing the groups were: (1) the number of 
 words per column, and (2) the number of columns completed. The 
 character of the words written and the closeness of their associa- 
 tion were ignored in scoring the test, as being impossible of simple 
 and objective measurement. Group comparisons were not made by 
 columns. 
 
 This test was devised and used in the present study under the 
 assumption that it measures ability in directed thinking. The per- 
 son before whom the test is placed is required to direct his thought 
 processes toward a definite end, and in order to receive a high score 
 must keep his attention continually fixed on that end. It is easily 
 imaginable that a person given to phantasy building, whose think- 
 ing, in other words, is not directed toward serviceable ends, will 
 lose his moorings at the first word and float away on a tide of 
 words with no port in view. This type would be represented by 
 those who on the average supply long lists of words and who also 
 take plenty of time in completing the test. At the other extreme we 
 should find the individual who is always present-minded and who 
 
30 MAX FREYD 
 
 directs his thinking toward definite objective ends. Such a person 
 will get a high score in the test, 7.¢., he will write few words and 
 finish most of the lists. The two types are represented in the two - 
 solutions on the first page of the test. 
 
 The number of words per column correlates +.28 with ratings 
 on present-mindedness. The number of columns completed shows 
 the extraordinary correlations of -++-.54 with ratings on self-asser- 
 tion and +.64 with ratings on bearing. 
 
 The following results were obtained with the primary groups 
 and the Case School seniors in Mechanical Engineering. 
 
 Numbers of words per column: 
 
 Q, Q, Q, 
 Industries seniors 1.60 2.25 3.05 
 Mech. Eng. seniors 2.02 2.43 2.75 
 Succ. 1. i. salesmen 1.85 ee) 2.95 
 
 Number of columns completed : 
 
 Q, Q, Q; 
 Industries seniors 37 5.4 7.0 
 Mech. Eng. seniors 6.0 74 8.0 
 Succ. 1. i. salesmen 5.0 6.7 8.8 
 
 In each case the engineering students obtain the highest median 
 score and the industries students the lowest. Since the salesmen 
 occupy an intermediate position it is difficult to interpret the re- 
 sults. If we ignore the engineering students, we note that the sales- 
 men excel in speed of association, but that the industries students 
 excel in directness and brevity of association. 
 
 SERIES 12. KENT-ROSANOFF FREE ASSOCIATION TEST 
 
 In its original form, as developed by Kent and Rosanoff (11) 
 this test consisted of one hundred words which were presented 
 orally to the subject one at a time. The subject responded in each 
 instance with the first word which came to his mind, the response 
 being recorded by the examiner. The test was in no sense a group 
 test. The speed of response was at first recorded, but was later 
 ignored owing to the magnitude of their undertaking. 
 
 The originators of the test aimed to make a study of “certain 
 disorders of the flow of utterance which appear to be dependent 
 upon a derangement of the psychical processes commonly termed 
 association of ideas” by “an application of the experimental meth- 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 31 
 
 od known as the association test” (11, p. 5). In order to have on 
 record the responses of normal people so that these could be com- 
 pared with the responses of the insane, the association list which 
 they arranged was given to one thousand normal subjects. From 
 an analysis of this material the authors are led to observe that “the 
 one tendency which appears to be almost universal among normal 
 persons is the tendency to give in response to any stimulus word 
 one or another of a small group of common reactions” (I1,p.14). 
 The tendency among the insane is to give unique responses, but it 
 is conceivable that an original person of great ability may give 
 uncommon responses. 
 
 The possibilities of this test in the study of vocational differ- 
 ences has been scarcely recognized. Since Kent and Rosanoff stan- 
 dardized the test in 1910, little use has been made of it on normal 
 adults. Wells and others have given us classifications of the types 
 of responses, and Murphy has used it to some extent for the com- 
 parison of people who are inclined toward literature with those 
 who are inclined toward science. Most of those who have used 
 the test were drawn to it by an interest in abnormal psychology. 
 Its application to normal adults as a measure of various volitional 
 and associational tendencies, rather than as a means of detecting 
 insanity or complexes, has been passed over. 
 
 The Bureau of Personnel Research has adapted this test for its 
 use by printing the list and administering it as a group test, with- 
 out any very apparent change in the function measured. No 
 changes were made in the lists of words. The words are printed 
 consecutively with blank spaces opposite them in which the re- 
 agents write their responses. The directions and examples are read 
 to the reagents by the examiner and any questions regarding them 
 are answered. A starting signal is given, at which the subjects com- 
 mence to_record their responses to the words in sequence, and at 
 the expiration of four minutes a stopping signal is given. This 
 form is far easier to administer than the individual test, and rein- 
 troduces the factor of speed. The error of timing is reduced, since 
 it can enter only into the first and last responses given by the sub- 
 ject. We have only a measure of average time per word, instead 
 of a separate measure for each word. Speed of writing is a factor 
 
32 MAX FREYD 
 
 which enters into the score on the number of responses given. The 
 
 correlation of Speeded Writing (Part 4 of Test IX) with the 
 
 number of responses in the Kent-Rosanoff Test is +-.37. 
 Following are the directions for the test: 
 
 Below is a list of 100 words. Write after each word the first word it sug- 
 gests (the first word that comes to your mind). 
 Do not write sentences. 
 Do not stop to think. 
 Work as quickly as you can. 
 EXAMPLES: 
 
 grass....green 
 school.. .teacher 
 
 For the purpose of this study, this test has been scored in thir- 
 teen different ways, each of which is described below together with 
 its results. ; 
 
 Number of Responses. This score is simply the number of word 
 responses given by the subject. General intelligence is probably the 
 dominant factor in this performance. In this connection Wells 
 says: . 
 
 “Probably no one could seriously assert that increased emo- 
 tional reaction is the sole cause of lengthened association time, 
 though it need not be questioned that it is an important contrib- 
 uting cause. The principal other factors are the failure to find a 
 response that is intellectually satisfying, and a failure of predomi- . 
 nance of any one among the associations presented” (23, p. 20). 
 
 We might very well assert that these factors are inversely cor- 
 related with intelligence. Speed of writing and general intelligence 
 are the two tested functions which correlate most highly with the 
 number of responses. The correlation coefficients are respectively 
 +.37 and +.31. 
 
 The interquartile ranges for the two primary groups are given 
 below. The figures are practically identical for the two groups. 
 
 Q, Q, Q, 
 Industries seniors EO." 02 70 
 Succ. 1. i. salesmen 55 62 70 
 
 Number of Spaces Omitted. The score was the number of spaces 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 33 
 
 omitted up to the last word response given. Blocking of response is 
 an extreme result of the factors operating to increase reaction 
 time. Results for the primary groups are given below. 
 Q, Q QQ, 
 Industries seniors 4 9 Bi 
 Succ. 1. i. salesmen 3 8 1.4 
 
 There is a very slight tendency for the industries students to 
 omit more responses. 
 
 Coefficient of Commonality. The commonality score on a re- 
 sponse is obtained by referring to the Kent-Rosanoff tables of re- 
 sponses given by one thousand normal adults (11, p. 77). In these 
 tables each response word is preceded by a number which repre- 
 sents the number of individuals (out of the thousand) who gave 
 that response to the particular stimulus word. For example, after 
 the stimulus word “dark” we find appearing first the word 
 “afraid,” with the figure 6 before it, indicating that six of the 
 one thousand subjects gave the word “afraid” in response to the 
 word “dark.” In scoring any person’s test paper, we give him a 
 score of 6 if he writes the word “afraid” after the stimulus word 
 “dark.” The score used in this study is the sum of the commonality 
 scores of the first twenty-five responses given by the subject. 
 
 This score is a criterion of the extent to which a person’s 
 thoughts run in normal objective channels (24, p. 74). Responses 
 with low commonality value are unique responses which usually 
 refer to the stibject’s individual experience. We cannot, however, 
 go so far as to say that a person who scores low has psychopathic 
 tendencies, since unusual responses may be characteristic of a 
 very original person. The genius and the imbecile may find a com- 
 mon ground in their commonality score. 
 
 Results obtained with the primary groups are given below. 
 
 Om Q, 0; 
 
 Industries seniors 1600 2100 3400 
 Succ. 1. i. salesmen 2300 3500 4400 
 
 The sales group shows a very significant superiority in this 
 function. 
 
 Responses with Commonality Value of Zero. This score is the 
 number of responses entering into the preceding score whose com- 
 
34 MAX FREYD 
 
 monality value is zero, t.e., the number of words which were not 
 found in the Kent-Rosanoff tables. 
 
 Results obtained are as follows: 
 Q, Q, Q, 
 Industries seniors 1.6 21 4.5 
 Succ. 1. i. salesmen ny, 1.5 2.6 
 
 As a corollary of the results of the preceding scoring method 
 we find the industries students obtaining higher scores in the num- 
 ber of zero value responses. 
 
 Responses with Commonality Value of 1-5 Inclusive. These re- 
 sponses are slightly less unique than the preceding ones. Scores are 
 based on the first twenty-five responses given by the subjects. 
 
 0. Q, OQ. 
 Industries seniors 2.8 4.0 6.0 
 Succ. 1. i. salesmen 1.7 2.9 4.8 
 
 Here again we find industries students obtaining higher scores. 
 
 Five additional scores were obtained, using the classification 
 given by Wells (22, 24). Effort toward the classification of types 
 of responses arose among psychiatrists. Some of the better known 
 classifications are those of Wells, Murphy, Krapelin,Aschaffen- 
 burg, Munsterberg, Claparéde, etc. Most of the classifications are 
 detailed and more logical than psychological, and they offer great 
 obstacles to practical use. Wells’ classification seems to possess 
 more points of merit than the others, and was therefore chosen for 
 use in this study. His classification is given on the following page. 
 This extract is copied verbatim from “Association Tests’ (24, 
 
 P. 77). 
 
 1, The egocentric reactions may be typified by— 
 
 a. Predicate reactions. Cloud-ominous, flower-pretty, crooked-line, red-rose, 
 scratch-cat, lion-roar, money-wish, invent-machine, weasel-stealth, beauty-rose, 
 safe-quite, almost-grown, sing-well, never-decide, nicely-very (including the 
 responses yes and no). 
 
 b. Responses in the form of proper names. Citizen-New York, boy-Johnny, 
 mountain-Kearsarge. 
 
 c. Reactions interpreting the stimulus word as a proper name. Eagle-news- 
 paper, park-square. 
 
 d. Reaction involving the response of a pronoun. Hand-you, health-me. 
 
 e. Interjections, failures of response or repetitions of the stimulus word. 
 
 2. The supraordinate category is confined strictly to the individual-genus 
 order, defined in such examples as, priest-man, potato-vegetable, lily-flower, 
 cow-animal. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED — 35 
 
 3. The contrast group is composed, of course, of reactions in which the re- 
 sponse means the opposite of the stimulus and is made up of such associations 
 as, good-bad, trouble-pleasure, scatter-gather, fertile-sterile, and the like. 
 
 4. The miscellaneous category is composed essentially of the remaining 
 reactions of the “inner” type. It includes about 45 percent of all associations. 
 
 5. The speech-habit group is composed of associations by familiar phrase 
 (stand-pat) ‘word compounding (play-ground) simple sound associations 
 (tease-sneeze) and syntactic changes (high-height). 
 
 The purpose of a classification of this sort is to gain an insight 
 into some of the mental processes. 
 
 The only modification made in the use of this classification was 
 to substitute the word “subjective” for “egocentric” in naming 
 the first category, in order to avoid the connotations of the latter 
 word. The responses falling into this group are not egocentric in 
 the sense that they disclose a self-centered personality, but in the 
 sense that they are not objective or universal. 
 
 The scores for each of the five categories consist of the number 
 of responses belonging to that category out of the first twenty- 
 five words in the list; not necessarily the first twenty-five responses 
 made by the subject, since omissions of response are classified as 
 subjective responses. 
 
 Some difficulty was encountered in using this classification since 
 a great many words could fit with equal justification into several 
 of the classes. 
 
 Subjective Responses. The following results were obtained with 
 
 the primary groups. 
 Oo; (Oia Oy, 
 Industries seniors 5.5 11.0 16.3 
 Succ. |. i. salesmen 3.0 8.0 12.0 
 
 In view of the fact that subjective responses have low common- 
 ality value, we should expect these results. 
 
 Supraordinate Responses. Very few responses of this type oc- 
 curred. Twenty-one of the industries students and twenty-six of 
 the salesmen made no responses of this type. 
 
 Contrast Responses. Since this has proved to be the most im- 
 portant single score yielded by the test, it deserves special attention. 
 It correlates -++.92 and -++.98 with the coefficient of commonality, 
 and since it is a much easier score to obtain than the commonality 
 score (a matter of several seconds compared with over a minute) 
 
36 MAX FREYD 
 
 it may be very easily substituted for that score. It also correlates 
 —.9o and —.86 with the number of subjective items, another im- 
 portant score which is difficult to obtain. It yields fairly high cor- 
 relations with most of the other scores obtained from this test, 
 and consequently is the most significant and representative score 
 
 yielded by the test. 
 All of the groups were scored on the number of contrast re- 
 
 sponses. Results are given below and in Fig. 7. 
 Industries Seniors STAB ee 
 
 West. N.S. Seniors ——— 
 
 Harr. RE. Salesmen Sra RS LD 
 
 Cleve. Disc. Salesmen 
 
 Life Ins. Schools AIT I 
 
 30 Life Ins. Salesmen Sr NST RES 
 Score ee EY ee SN) ice 
 Fic. 7 
 
 Interquartile Ranges of Number of Contrast Responses, 
 Series 12, Kent-Rosanoff Test 
 
 Q, Q. Q, 
 Mechanical groups 1.6 4.7 9.8 
 Industries seniors 1.0 2.7 7.0 
 West. N.S. seniors 1.4 28 4.0 
 Cl. Disc. Co. and Harr R.E. Co. salesmen 6 2.5 7.5 
 Life ins. salesmen (schools only) 1.3 7.6 II.0 
 Succ. 1. i. salesmen 39 7.3 83:3 
 
 The large difference between the primary groups is cut down 
 considerably by the secondary groups, yet we find the life insur- 
 ance salesmen showing a slight superiority as a whole over the 
 mechanics students. (See table at end of chapter. ) 
 
 Miscellaneous Responses. The following results were obtained 
 
 on the primary groups. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 37 
 
 Q, Q, Q; 
 Industries seniors 7.0 10.0 12.6 
 Succ. 1. i. salesmen 8.0 9.9 10.9 
 
 No group differences appear. 
 Speech-Habit Responses. Results obtained with the two primary 
 groups are given below. 
 
 Q, Q. Q, 
 Industries seniors 4 8 1.6 
 Succ. 1. i. salesmen 6 1.4 2.0 
 
 There is a significant tendency for salesmen to make more re- 
 sponses of this nature than industries students. 
 Responses were also classified into three grammatical cate- 
 gories. 
 Noun Responses. Results on the primary groups follow. 
 Q, Q, Q, 
 Industries seniors 12 15 19 
 Succ. 1. i. salesmen 12.5 14 18.5 
 
 No group differences appear. 
 * Adjective and Adverb Responses : . 
 ; Q, Q, a 
 
 Industries seniors 5.0 6.8 9.8 
 Succ. 1. i, salesmen 6.0 9.1 10.5 
 
 The salesmen show a significant tendency to give more re- 
 sponses of this class than the industries students. The difference 
 is not as great as in the number of contrast responses. 
 
 Verb and Participle Responses: 
 
 Q, Q, Q, 
 
 Industries seniors 23 Bia 4.3 
 Succ. 1. i. salesmen 1.0 28 3.9 
 
 There is some tendency for industries students to make more 
 responses of this sort than the salesmen. 
 
 SUMMARY 
 
 The results which may be classified as highly significant from 
 the statistical standpoint are the following: 
 
 Mechanically inclined people excel in 
 Speed of checking better traits (Part 3, Test IX) 
 Speed of checking personal traits (Part 9, Test IX) 
 Assurance in visual memory (Part 11, Test IX) 
 Speed of writing in small space (Part 8, Test IX) 
 Number of subjective responses in the Kent-Rosanoff test. 
 
38 MAX FREYD 
 
 Salesmen excel in 
 Ability to disguise handwriting (Part 6, disguises, Test IX) 
 Inability to write slowly (Part 2, Test IX) 
 Carelessness in copying (Part 7, Test 1X) 
 The coefficient of commonality in the Kent-Rosanoff test. 
 The following are less significant results: 
 
 Mechanically inclined people excel in 
 Enlargement of writing under distraction (Part 5, Test IX). 
 Ratio of objective to subjective speed of decision (Part 9, 
 
 ratio, Test IX) . 
 
 Agreements between checks in Parts 3 and 9. 
 Speeded writing (Part 4, Test IX) 
 
 Salesmen excel in 
 Number of disguises attempted (Part 6, letters, Test IX) 
 Speech-habit responses in the Kent-Rosanoff test. 
 Contrast responses in the Kent-Rosanoff test. 
 
 SUPPLEMENTARY NOTE 
 
 The following table gives a more accurate estimate of the 
 tested differences than is afforded by comparisons of interquartile 
 ranges. The figures show how significant are the differences in 
 means between the Mechanical Group and the Miscellaneous Sales- 
 men and the Life Insurance Salesmen combined, in those tests in 
 which results on these large groups are available. In Part 3 of Test 
 IX, for instance, the group referred to as the Mechanical Group 
 is compared with the Miscellaneous Salesmen and the Life Insur- 
 ance Salesmen combined. Differences between the primary groups 
 have not been treated in this manner since the procedure has no 
 value with small groups. The statistical treatment follows that of 
 Fernald, Hayes, and Dawley (5, p. 109 et seq.) 
 
 TABLE I 
 AVERAGES CHANCES THAT REAL 
 TEST SALES MECH. DIFFERENCE DIFFERENCE DOES 
 ‘GROUPS GROUP NOT EXIST ARE I IN 
 Test IX 
 Batt 24.34 28.44 4.10 6.30 Infinity 
 Part oy > 21.63 25.66 4.03 4.74 555,550.0 
 Part 6, disg. 3.00 2.60 .40 2:33 2,301.0 
 Part 5 197.98 201.64 3.66 1.16 8.1 
 Part I1 17.10 25.50 8.40 8.40 Infinity 
 Part 9, ratio 122.00 125.76 3.76 885 5.3 
 Part 2 14.18 9.00 5.18 3.08 31,546.0 
 Part 8 73.02 04.7 Die 9.22 Infinity 
 atten 9.68 6.75 2.03 6.42 Infinity 
 Part 6, letters 103.58 101.76 1.82 .408 3.2 
 
 Kent-Ros. Test 
 Cont. Resp. 5.34 5.08 .26 52 Ge) 
 
CHAPTER III 
 
 GROUP COMPARISONS BASED ON ANSWERS 
 TO QUESTIONNAIRES 
 
 THE METHOD 
 
 Questionnaires have been used for gathering data in diverse 
 experiments by many psychologists, including Galton, Hall, Leuba, 
 and others. The questionnaire method is objective and has the ad- 
 vantages that no timing is necessary and groups do not have to be 
 assembled for purposes of administering the papers. For the same 
 reasons, however, a certain caution must be observed. If the pur- 
 pose of the questionnaire is obvious there is nothing to prevent 
 the subject from filling it out in the way which he thinks will best 
 serve his interests. In order to make the answers more valuable, 
 the purpose of the questionnaire must not be too apparent. If under 
 these circumstances the subject guesses at the purpose and fills out 
 the questionnaire falsely, he runs no greater chance of hitting upon 
 the “right” answers than he would in a test. 
 
 In this study a series of five detailed questionnaires was submit- 
 ted to the two primary groups and to the Case School seniors. 
 Group comparisons were made in answers given. By this means it 
 was intended to study differences in personality as expressed in in- 
 terests and attitudes, and to determine the importance of personal 
 history as a cause for any differences in personality disclosed. 
 
 No statistical evaluation of the differences was made, owing to 
 the immense amount of computation involved. In the presenta- 
 tion of the results, the original figures are given in most instances. 
 
 DESCRIPTION OF QUESTIONNAIRES 
 
 The subject matter of the five questionnaires is roughly the fol- 
 lowing: Questionnaire 1, early personal history; Questionnaire 2, 
 vocational interests; Questionnaire 3, avocational interests; Ques- 
 tionnaire 4, likes and dislikes; Questionnaire 5, beliefs. The 
 
40 MAX. FREYD 
 
 groundwork of the lists was a series of questions submitted to 
 C. S. Yoakum by students enrolled in a seminar in personnel re- 
 search in 1919, at Carnegie Institute of Technology. These ques- ~ 
 tions were intended to reveal a cross-section of an individual’s 
 personal history and interests at several important stages of his 
 development. It is due to the incentive furnished by Yoakum that 
 the series of questions was compiled and that modified forms were 
 used in this study. In addition to this source, items were drawn 
 from the published material of the following men: Allport, Baade, 
 Lipman, and Stern, Cimbal, Healy, Hoch and Amsden, Meyer, 
 Moore, Pressey, Sommer, Stern, Wells, and Woodworth. A num- 
 ber of forms in use at various institutions were also consulted. 
 Items for the questionnaire on beliefs were contributed by M. A. 
 Bills and L. D. Anderson, of the staff of the Bureau of Personnel 
 Research. Some use was made of the lists given by Kemble, 
 Nathan and Mencken, and Sumner. 
 
 Questionnaire 1. Early Personal History. The personality of an 
 individual is the resultant partly of heredity and partly of en- 
 vironment. The latter commences its modifying influence directly 
 after birth, and during childhood so molds the individual that the 
 impressions of this period have more influence in shaping his per- 
 sonality than those of any later period. By finding out what these 
 impressions were, we can better understand the personality of the 
 adult. In our case we can determine to what extent the individuals 
 in the same vocational group were subjected to the same influence ~ 
 in childhood. 
 
 The most thorough attempts to determine the effect of environ- 
 ment in shaping the adult personality have come from psychi- 
 atrists. Before attempting to cure his patient the psychiatrist 
 makes it a point to have before him all the material he can assemble 
 on the past history of the patient. He can then trace the causes of 
 the malady and apply such curative methods as-the case demands. 
 
 That personal history is an important factor in vocational choice 
 and success has been proved by Andrews (1), Goldsmith (8), and 
 Ream (17). These workers have found that certain facts yielded 
 by questions ordinarily included in application blanks have such 
 great vocational differentiating value that workers may with fair 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 41 
 
 success be chosen merely on the basis of their answers to these 
 questions. Their results have been carried over into this study on 
 a larger scale. 
 
 Questionnaire I consists of seventy questions covering paren- 
 tage, and early home and social environment, etc., up to the ap- 
 proximate age of twelve. 
 
 Questionnaire 2, Vocational Interests. Questionnaire 3, Avo- 
 cational Interests. Questionnaire 4, Likes and Dislikes. 
 
 Questionnaire 2 covers items of vocational interest. The third 
 page is a modification and enlargement of a form used by Moore 
 (13) for determining occupational interests. The 72 occupations 
 are listed in alphabetical order, but they are readily classified into 
 general groups, such as scientific occupations, literary occupations, 
 selling, indoor work, outdoor work, executive work, etc. 
 
 Questionnaire 3 covers such items as hobbies, literary interests, 
 diversions, athletic interests, social tendencies, religious tenden- 
 cies, etc. 
 
 Questionnaire 4 contains 129 items toward which the subject 
 expresses his attitude, and which fall into three general groups: 
 first, a series of physical attributes of people, up to item 26; sec- 
 ond, a series of mental attributes of people, items 27 to 76; and 
 third, a series of miscellaneous items, 77 to the end. In addition to 
 the detailed tabulation of the answers given by groups of indi- 
 viduals, this questionnaire offers information on personality in the 
 total number of each kind of symbol checked in response to the 
 first 76 items. Correlations with ratings (Chapter V) show that 
 the person who tends to check a great many “?” after the first 76 
 items tends also to be rated as flexible and adaptable (r == +.48). 
 The person who checks a great many L’s and L!’s after these 
 items tends to be rated as submissive (r == +.43), self-conscious 
 (r = +.43), and unadaptable (r == +.31). Group comparisons 
 on the basis of the number of such symbols checked reveal no dif- 
 ferences. | 
 
 An individual may go into an occupation not only because of 
 certain factors in his personal history or because of an interest in 
 that occupation, but also for the reason that he has a number of 
 concomitant interests which are gratified by the type of work he 
 
42 MAX FREYD 
 
 has elected to enter upon. In addition to sales ability or abilities, 
 which may be difficult to fathom, successful salesmanship is due 
 to communities of interests all of which find an outlet in sales 
 tasks; such interests, for example, as in talking to and meeting 
 people, in variety in daily work, in walking around rather than 
 sitting at a bench all day, and so on. It has been shown that records 
 of such interests are valuable supplements to tests for the selection 
 of salesmen. 
 
 It does not matter much for present purposes how we define 
 interests. In behavioristic terms, interest in an object may be de- 
 fined as a tendency to make positive reactions toward that object. 
 Introspection would add the affective element, or “likes.’’ We are 
 not concerned here with interests that are mere reflexes, such as 
 the interest in a bright light in the periphery of vision. Rather we 
 are concerned with those which are developed in a social environ- 
 ment and are based on either heredity or social habit. 
 
 Very little scientific study of interests has been made up to the 
 present. The outstanding researches are those which Thorndike 
 (19) has carried out on interests in school studies. The methods 
 used were such as to make the results of limited significance. They 
 point to the conclusion that courses tend to maintain their relative 
 degree of interest throughout school or collegiate life; that a high 
 correlation exists between a person’s interest in a school subject 
 and his ability in that subject as judged by himself; and that a 
 significant, but somewhat lower, correlation exists between inter- 
 est in a subject and ability in it as measured by school grades. The 
 conclusions give weight to any arguments for the use of interests 
 as auxiliaries to measures of abilities. 
 
 The analysis of interests has played a prominent part in two 
 major studies produced by members of the staff of the Bureau of 
 Personnel Research. Moore (13) used an interest questionnaire in 
 differentiating among a group of graduate engineers those who 
 inclined toward design engineering and those who inclined toward 
 sales engineering. The section of his interest questionnaire which 
 proved to be the most reliable for the differentiation of the two 
 groups was a record of occupational preferences. A scoring 
 method for this blank was determined empirically; that is, if a 
 
THE SOCIALLY AND, THE MECHANICALLY INCLINED 43 
 
 larger proportion of the sales engineers than the design engineers 
 liked a certain occupation, this reaction to the occupation was 
 assigned a score favorable to sales engineering. The total score 
 for any person indicated whether his inclinations were in the di- 
 rection of design engineering or sales engineering. No attempt 
 was made to determine the relative significance of the items en- 
 tering into the total score, all such items being weighted one. By 
 means of the total score on choice of occupations, 82 per cent of 
 the men were placed properly, in other words, with the most favor- 
 able critical score, there was an overlapping of only 18 per cent. 
 
 Ream (17) used interest blanks very much like Questionnaires 
 2 and 4 of the present series, in his study of methods of selecting 
 successful life insurance salesmen. In his study a refinement of 
 statistical procedure was introduced, namely, where the propor- 
 tions of successful and unsuccessful salesmen who liked or dis- 
 liked an item differed considerably, the ratio of the difference in 
 proportion to its standard error was calculated. All items in which 
 this ratio was one or more entered into the total score on the in- 
 terest blanks. All items were weighted one. Distributions of total 
 scores on the interest blanks showed that by their means approxi- 
 mately 80 per cent of the men were placed correctly. 
 
 Any attitude which has a strong emotional accompaniment may 
 be called a “complex.’’ Thus Hart (Psychology of Insanity) 
 speaks of the photography complex. Pressey has tried to measure 
 the complexes of an individual by the questionnaire method (15). 
 His contribution presents two aspects of interest: (1) the blank 
 is made extremely concise and compact by printing lists of words 
 with directions for expressing one’s attitude toward them at the 
 top of the page, each word thus representing what would ordinari- 
 ly be a complete question; (2) a method is given for obtaining 
 a total score. Pressey selected for his total score ten items crossed 
 out in greater proportion by successes in school and ten others 
 crossed out in greater proportion by failures. By subtracting each 
 of the latter from the former (checked by any student) he ob- 
 tained a total score which differentiated successes from failures. 
 The overlapping of the distributions for the two groups, using the 
 most favorable critical score, was only Io per cent (11/106). 
 
44 MAX FREYD 
 
 Questionnaire 4 serves a somewhat similar purpose. Complexes 
 are here disclosed by checking L! or D! These checks are due to | 
 emotionally toned incidents in the previous experience of the re-» 
 agent by which the object of the complex acquired a gil or 
 unfavorable halo. 
 
 Questionnaire 5. Beliefs. The final questionnaire consists of a 
 number of questions on which people tend to disagree. The 
 answers are therefore measures of attitudes, perhaps emotionally 
 toned, and in a limited sense serve to gauge the subject’s informa- 
 tion. 
 
 That attitudes expressed in answers to such questions are of 
 vocational importance, is shown by one study at least, that of 
 Leuba (12, p. 255) on the belief in God and immortality. He com- 
 pared leaders and lesser men in the physical and biological sciences 
 in the proportion of their numbers believing in God and immor- 
 tality. The proportions are as follows: 
 
 PER CENT BELIEVING IN 
 GOD IM MORT. 
 
 Physical sciences Leaders 34.8 40.0 
 Lesser men 49.7 ra 
 Biol. sciences Leaders 16.9 25.4 
 Lesser men 39.1 45.1 
 
 These figures have no direct bearing on the present study except 
 as they show the possibility of obtaining important. group differ- 
 ences in reactions to moot questions. 
 
 The headings of the various questionnaires and specimen ques- 
 tions are given on the following pages. 
 
 QUESTIONNAIRE 1 
 Earty PersonaL History 
 
 This information is wanted for statistical purposes only. It will be kept 
 strictly confidential. Your answers will be seen by no one except the person 
 who does the tabulating. 
 
 Consult no one in filling out this questionnaire, except to get necessary in- 
 formation to which you do not have access. 
 
 If a question is followed by the letters Y N, draw a circle around the Y 
 if your answer is Yes, and draw a circle around the N if your answer is No. 
 
 Leave no questions unanswered. 
 
 1. What is your height?........ feet Sais. inches. Weight........ pounds. 
 4. In what state or, country: weresyou) borti? ..c.251.. as seca ued cee 
 
22. 
 
 ae 
 
 24. 
 
 SO HRN AM 
 
 THE SOCIALLY AND THE MECHANICALLY INCLINED 45 
 
 In what. State or country wasryour father: Dom? .civdiesdoscscsecassare 
 What was your father’s'nationality, (ancestry,) 2? cc. csccccc ccs c cress occ 
 In what state or country was your mother born? .......ccecseccsscccees 
 What was your mother’s nationality (ancestry)? .........seeeeeeecerees 
 Were your parents divorced after your birth? Y N. 
 
 How old were you at the time of your father’s death, if he is not living? 
 
 Which of your parents did you like the better? Father ( ), Mother (_ ). 
 Which of your parents exercised the most influence on your life? 
 Father ( ), Mother ( ). 
 
 Do you feel that if you had your childhood to live over again, that you 
 would desire a change in your home and other environmental conditions? 
 2 N. 
 
 Would you choose more sociable parents? ( ), less sociable ones ( ), 
 or the same (_ )? 
 
 What was the general tenor of your family life? Formal ( ), Informal 
 ( ), Refined ( ), Unrefined ( ), Happy ( ), Gloomy (_ ), Quarrel- 
 some ( ), Peaceable (_ ). 
 
 Who was the dominant one in the family? ccc. ccc enemies ap Sawectslecs 
 As a child how were you treated in general in comparison with your broth- 
 ers and sisters? Better ( ), Worse ( ), Average ( ). 
 
 (There are 70 questions altogether.) 
 
 QUESTIONNAIRE 2 
 
 VocATIONAL INTERESTS 
 
 (Directions are similar to those for Questionnaire 1.) 
 
 Did you ever plan to take a civil service examination? Y N. Did you 
 ever take a civil service examination? Y N. Did you pass? Y. WN. 
 Is your present vocational choice due to the advice of a friend (_ ), rela- 
 tive ( ), vocational adviser ( ), teacher ( ), school friend (_ ), ac- 
 quaintance ( ), someone in the same vocation (_ ), to reading ( ), or 
 to a consideration of your own abilities ( )? You may check more than 
 one. 
 
 In what occupation and position do you see yourself ten years from now? 
 
 SiG) ©6010) ¥18).0)01 6 6 6% 2. ave 0.0» 60's) 6 6 © O10) 6, 6 60 0 0 6 6 © 9 © 00.000 0,0 00 8 68 2 6 9 810 8 0.019 & 
 
 What income do you expect to command at 50? $............ 
 Did you ever quit a job because 
 
 The work was too hard for you physically?...... aN 
 ApMerewdaSetoO MCh T101SC ley suis. «oi «ain elise silence SY aoN| 
 PEM OEM AS- 100 CILty ! twion> « 0.0.0 s ames maeee bass! 
 The work was too monotonous? .............6. VoaNG 
 You were not promoted often enough? .......... Y N 
 You disagreed with the boss? ............seeeee: Yon, 
 You felt that the job offered no future?.......... avn LN) 
 
 (There are 13 questions altogether.) 
 
46 MAX. FREYD © 
 
 CHOICE OF OCCUPATIONS 
 
 (Third page of Questionnaire 2.) 
 
 After each of the occupations listed there are three symbols. Cross out one 
 of the symbols after each occupation, as follows: 
 
 ‘Cross out L if you would like doing that kind of work. 
 
 Cross out D if you would dislike doing that kind of work. 
 
 Cross out ? if you have decided feelings toward that kind of work, or know 
 nothing about it. 
 
 Disregard any salary or social differences or any possible family objections. 
 Consider only your interest and satisfaction in doing each of the kinds of work 
 listed. You are not asked whether you would take up the occupation per- 
 manently; you are merely asked if you would enjoy that kind of work. Assume 
 that you have the ability necessary for each of the occupations. 
 
 Be sure to cross out one symbol after each of the occupations. Do not skip 
 any. 
 
 Actor Leigh) Mechanical engineer L ? D 
 Architect bia a 8 Member of Congress L ? D 
 Artist OWT a) Musician bee AB, 
 Astronomer ee, Newspaper reporter L ? D 
 Auctioneer pee ANB) Novelist Li dae el 
 
 (There are 72 occupations in the list.) 
 
 QUESTIONNAIRE 3 
 
 AVOCATIONAL INTERESTS 
 
 (Directions are similar to those for Questionnaire 1.) 
 
 3. Have you ever constructed a piece of furniture or household appliance 
 other than in a school shop? Y N. 
 
 4. Do you think that you could find out what was wrong with a watch that 
 did not run? Y N. An electric motor? Y N. An autor Y N. A 
 gasoline motor? Y N. 
 
 Did you ever build a coop or a house for an animal? Y N. 
 
 Nw 
 © 
 < 
 o 
 ee 
 ° 
 cS 
 o 
 < 
 o 
 st 
 Po 
 @ 
 oO 
 co 
 us) 
 o 
 co 
 rab) 
 as 
 , 
 a 
 w 
 ~~ 
 ri 
 a 
 = 
 ) 
 cr 
 ° 
 3 
 o 
 wn 
 ~ 
 
 7. Have you ever made a collection of stamps? Y N. Coins? Y N. 
 Postal cards?” Y- = Ni \Whatelsetosicaaee cos cs > lee «na Gerard ae nee 
 
 8. Did you ever own a camera? Y N. Which type of picture do you pre- 
 fer to take, scenes (_ ) or snapshots of people (_ )? Check only one. 
 
 44. Estimate how many smokers, lodge meetings, card parties or other social 
 affairs of your own sex you have attended during the past year.......... 
 Estimate how many mixed social affairs you have attended during the 
 Past Veataadces te cee Include dances, parties, socials, etc. Do not include 
 ordinary “dates.” 
 
 45.’ How many plays have you acted in?..:4.6-..0..acs cea cue een 
 
 46. Can you successfully tell a story to a group of people? Y N. 
 
 47. Name the social clubs, fraternities, and business organizations to which 
 Vou Delong. son's one's s apps o's ooo ce WN ao vie seme Ore ce ene eae 
 Draw a line under each one which you attend regularly. : 
 
 (There are 76 questions in this list.) 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 47 
 
 QUESTIONNAIRE 4 
 
 LIKES AND DISLIKES 
 
 Cross out one of the symbols after each of the items below, as follows: 
 
 Cross out L! if you like the item very much. 
 
 Cross out L_ if you like the item. 
 
 Cross out ? if you have no decided feelings toward the item. 
 
 Cross out D_ if you dislike the item. 
 
 Cross out D! if you dislike the item very much. 
 
 You may be in doubt in some cases. Use your own judgment, always check- 
 ing the symbols in the way which expresses your most frequent attitude. 
 
 Be sure to cross out one symbol after each of the items. Do not skip any. 
 
 Blondes Os Sere ek Be 
 Brunettes Dial DAD 
 Very old people Felsh eet aly 4 
 Children Bi aes De 
 Cripples Lp Lake Ee 
 People who disagree with you LAalee Ly) | 
 People more intelligent than you L! L ? D D! 
 People less intelligent than you L! L ? D D! 
 People who borrow things Lie LoaterDa Dt 
 Quick-tempered people j TN ea ar aie AD 
 ‘Conventions Pigiiveiae DID} 
 Sporting pages Bevis wri at 
 Love stories Dieser Dip 
 Detective stories ee Lee Loe LA 
 “Life” SE hh OO 8 
 
 (There are 129 such items.) 
 
 QUESTIONNAIRE 5 
 
 BELIEFS 
 
 Draw a circle around Y if your answer is Yes. 
 Draw a circle around N_ if your answer is No. 
 
 This information is wanted for statistical purposes only. 
 Your answers will be kept strictly confidential. 
 Do you believe 
 1. That Japan and the United States will be at war with each other within 
 the next forty years? Y N. 
 That divorce laws should be made stricter? Y N. 
 That prize fights should be prohibited? Y N. 
 That capital punishment should be abolished? Y N. 
 That unstinted service is always rewarded? Y N. 
 (There are 82 questions in this list.) 
 
 Qk N 
 
48 MAX FREYD 
 
 RESULTS 
 
 The numbers at the heads of the paragraphs below refer to the 
 questions in the original lists. Only significant positive results are 
 presented here.* The groups compared are the two primary groups 
 and the Case School seniors. 
 
 Questionnaire 1. Early Personal History 
 
 5. Eighty per cent of the fathers of the salesmen were born in 
 this country; 60 per cent of the fathers of the industries students ; 
 and 53 per cent of the fathers of the engineering students. 
 
 6. The per cents of fathers of American or British nationality 
 among the three groups to whom the questionnaire was given are: 
 primary sales group, 73 per cent; primary mechanical group, 56 
 per cent; and Case School seniors, 46 per cent. American nationali- 
 ty: 23, 13, and II per cent, respectively. 
 
 7. The per cents of mothers born in this country are: salesmen, 
 83 per cent; industries seniors, 43 per cent; and engineering stu- 
 dents, 72 per cent. 
 
 8. The per cents of mothers of American or British nationality 
 are as follows: salesmen, 67 per cent; industries students, 43 per 
 cent; and engineering students, 39 per cent. American nationality, : 
 26, 6, and 7 per cent, respectively. 
 
 12. Education of fathers: 
 PER CENTS FINISHING 
 GRADES HIGH SCHOOL COLLEGE 
 
 Succ. 1. i. salesmen 82 46 19 
 Industries seniors 73 31 4 
 Mech. Eng. seniors 77 42 16 
 
 13. Education of mothers: 
 PER CENTS FINISHING 
 GRADES HIGH SCHOOL COLLEGE 
 
 ‘Succ. 1. i. salesmen 96 54 14 
 Industries seniors 78 44. ) 
 Mech. Eng. seniors 86 43 Zz 
 
 1It is impossible to present the complete questionnaires and results here. 
 Those interested in examining the questionnaires may obtain copies from the 
 Bureau of Personnel Research at Carnegie Institute of Technology. Those 
 questions not reported on in this section revealed no group differences or 
 merely insignificant ones, or showed differences but were omitted because the 
 data were irrelevant to the present study. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED — 49 
 
 15. Parent “taken after”: 
 PER CENT TAKING AFTER 
 FATHER MOTHER BOTH NEITHER 
 
 Succ. 1. i. salesmen 57 30 10 3 
 Industries seniors 47 50 (a) 3 
 Mech, Eng. seniors 51 42 a, ) 
 
 16. Side of ancestry resembled: 
 PER CENT RESEMBLING 
 FATHERS MOTHERS 
 
 Succ. |. i. salesmen 57 43 
 Industries seniors 47 53 
 Mech. Eng. seniors 49 46 
 
 17. There is a slightly greater tendency among the salesmen to 
 be the oldest or youngest of a large family of boys. They belong 
 to larger families than the members of the other two groups. In 
 this respect the engineering students occupy an intermediate po- 
 sition. The number of only children is proportionately the same 
 for the three groups. 
 
 -19. The occupations of fathers were classified in six rough 
 groups. The number of times each type is mentioned as a father’s 
 
 occupation is shown below. 
 SALESMEN IND. MECH.ENG, 
 
 Farming 6 2 4 
 Commerce 12 4 20 
 Industry 5 12 10 
 Professional 10 I 9 
 Sales” «. 2 ) 5 
 Clerical (o) 6 I 
 
 20. Parent liked better: 
 PER CENT PREFERRING 
 FATHER MOTHER BOTH 
 
 Succ. 1. i. salesmen 7 40 53 
 Industries seniors O 63 37 
 Mech. Eng. seniors 15 60 30 
 
 21. Parent exercising the most influence on their lives: 
 PER CENT NAMING 
 : FATHER MOTHER EQUAL 
 Succ. 1. i. salesmen 27 57 17 
 
 Industries seniors te) 93 7 
 Mech. Eng. seniors 34 59 7 
 
 22. Percentages of men who would want a change in their home 
 
50 MAX FREYD 
 
 and other environmental conditions if they had their lives to live 
 
 over again: 
 PER CENT 
 Succ. 1. i. salesmen 23 
 Industries seniors 33 
 Mech. Eng. seniors 29 
 
 24. Practically all of the men characterize their family life as 
 “informal, refined, happy, and peaceable.’’ Five engineernig stu- 
 dents, one industries student, and no salesmen check “unrefined.”’ 
 Five engineering students, three industries students, and no sales- 
 men check ‘“‘quarrelsome.”’ 7 
 
 25. Dominant member of family: 
 
 PER CENT NAMING 
 FATHER MOTHER NEITHER 
 
 Succ. 1. i. salesmen 57 17 26 
 Industries seniors 50 30 20 
 Mech. Eng. seniors 62 BT 3 
 
 26. Treatment as a child compared with brothers and sisters: 
 PER CENT TREATED 
 BETTER WORSE AVERAGE 
 
 Succ. 1. i. salesmen T3it hin 0 87 
 Industries seniors 7 7 86 
 Mech. Eng. seniors 13 fa) 87 
 
 27. Financial circumstances of family: 
 
 PER CENT INDICATING 
 POOR MODERATE WELL-TO-DO 
 
 Succ. 1. i. salesmen 10 60 30 
 Industries seniors 17 80 3 
 Mech. Eng. seniors 9 12 79 
 28. Social inclinations of family: 
 PER CENT 
 MORE OFTEN INVITER MORE OFTEN INVITED 
 Succ. 1. i. salesmen 43 57 
 Industries seniors 65 35 
 Mech. Eng. seniors 64 36 
 
 30. Percentages whose family acquaintanceship was limited to 
 people of the same religious belief : 
 
 PER CENT 
 Succ. 1. i. salesmen 23 
 Industries seniors 30 
 
 Mech. Eng. seniors 30 
 
THE SOCIALLY AND. THE MECHANICALLY INCLINED 51 
 
 31. Father’s religious belief : 
 
 PER CENT 
 CATHOLIC PROTESTANT 
 Succ. 1. i. salesmen 3 83 
 Industries seniors 26 53 | 
 Mech. Eng. seniors 7 81° 
 32. Mother’s religious belief : 
 PER CENT 
 CATHOLIC PROTESTANT 
 Succ. 1. i. salesmen 3 90 
 Industries seniors 23 60 
 Mech. Eng. seniors 7 83 
 41. Per cents who had a desire to run away from home: 
 PER CENT 
 Succ. 1. i. salesmen 23 
 Industries seniors 10 
 Mech. Eng. seniors 19 
 Per cents running away from home: 
 PER CENT 
 Succ. 1. i. salesmen 7 
 Industries seniors 3 
 Mech. Eng. seniors 2 
 
 43. A tendency is shown for salesmen to be brought up in a bet- 
 
 ter residential district than the industries students. The latter are 
 more often residents of small towns. 
 
 44. Per cents desiring a wealthier environment if they had their 
 lives to relive: . 
 
 PER CENT 
 Succ. 1. i. salesmen 17 
 Industries seniors 30 
 Mech. Eng. seniors 28 
 
 45. Per cents desiring more refined playmates: 
 
 PER CENT 
 Succ. 1. i. salesmen 27 
 Industries seniors 33 
 Mech. Eng. seniors 26 
 
 46. Per cents desiring more cheerful conditions: 
 
 PER CENT 
 Succ. 1. i. salesmen 17 
 Industries seniors oF 
 
 Mech. Eng. seniors 35 
 
52 MAX FREYD 
 
 53. Fisthights engaged in before the age of twelve: 
 PER CENT WINNING 
 MEDIAN NO. MORE THAN HALF HALF LESS THAN HALF 
 
 Succ. 1. i. salesmen 3.7 23 60 17 
 Industries seniors 13.0 64 18 18 
 ‘Mech. Eng. seniors 5.0 4I 37 22 
 
 66. Per cent of income earned while attending high school: 
 MEDIAN PER CENT 
 
 Succ. |. i. salesmen yi 
 Industries seniors II 
 Mech. Eng. seniors 9 
 
 69. Per cent of expenses earned at college: 
 MEDIAN PER CENT 
 
 Suce. 1. i. salesmen 9 
 Industries seniors 21 
 Mech. Eng. seniors 14 
 
 Questionnaire 2. Vocational Interests. 
 
 6. Reasons for present vocational choice: 
 FREQUENCY AMONG 
 SALESMEN IND. MECH. ENG. 
 
 Advice of a friend II 3 7 
 Advice of a relative 10 3 14 
 Advice of a school friend . O 6 
 Advice of someone in same vocation 12 3 7 
 Consideration of own abilities 22 20 35 
 
 8. Income expected at the age of 50: 
 
 MEDIAN . 
 Succ. |. i. salesmen $19,300 
 Industries seniors 16,500 
 Mech. Eng. seniors 10,500 
 
 g. Reasons for ever quitting jobs: 
 FREQUENCY AMONG 
 SALESMEN IND. MECH. ENG. 
 
 Not promoted often enough 6 3 2 
 _ Disagreed with the boss 5 9 6 
 11. Per cents bothered if they were watched at work: 
 : PER CENT 
 Suce. 1. i. salesmen 5 
 Industries seniors 31 
 Mech. Eng. seniors 30 
 
 The occupational choices which showed the greatest group dif- 
 ferences are given in Table II. No statistical procedure was used 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 53 
 
 in selecting these items beyond omitting all which did not show 
 differences of ten per cent or more. If the item was skipped or if 
 both L and D were checked, credit was given for checking the 
 question mark, 
 
 TABLE II 
 
 Group Comparisons in Choice of Occupations. Percentages of each group 
 checking L (Like) after the occupations. The groups are the two primary 
 groups and the Case School seniors in mechanical engineering. 
 
 PER CENTS OF EACH GROUP CHECKING 
 L AFTER THE OCCUPATION 
 
 SA. IND. M.E. 
 Actor 46 20 23 
 Architect 43 50 55 
 Astronomer 13 36 23 
 Automobile salesmen 63 53 37 
 Auto racer 26 40 34 
 Auto repairman 10 46 49 
 Aviator 56 60 58 
 Bank teller 20 33 vidi 
 Baseball player 46 66 28 
 Carpenter 20 40 a7 
 Cook 3 13 9 
 Dentist 10 16 13 
 Draftsman 10 43 39 
 Editor 43 30 32 
 Explorer 36 60 72 
 Factory manager 40 90 90 
 Farmer 23 26 15 
 Foreign correspondent 50 30 18 
 Hotelkeeper or manager 26 36 15 
 Labor arbitrator 20 33 18 
 Lawyer 60 36 18 
 Locomotive engineer 20 43 34 
 Machinist 10 50 46 
 Magazine writer 53 2 28 
 Mechanical engineer 30 43 90 
 Musician 50 70 690 
 Newspaper reporter 40 20 22 
 Novelist 33 13 20 
 Office clerk - g 10 2 
 Office manager 50 66 51 
 Philanthropist 56 43 30 
 Politician 33 23 9 
 Preacher 2 3 9 
 Promoter 50 70 49 
 
 Real estate salesman 43 26 13 
 
54 MAX FREYD 
 
 Tape []—Continued 
 PER CENTS OF EACH GROUP CHECKING 
 L AFTER THE OCCUPATION 
 
 Ship officer 10 60 44 
 Shop foreman 3 50 51 
 Social worker 33 30 15 
 Specialty salesman 56 20 30 
 Steeple-jack 6 10 18 
 Stock broker 53 63 37 
 Surgeon 53 30 37 
 Toolmaker 3 50 32 
 Traveling salesman 36 40 32 
 Watchmaker tan 23 18 
 
 Questionnaire 3. Avocational Interests 
 
 1. The following table presents the most important differences 
 
 in attitudes toward school subjects. 
 
 PER CENTS CHECKING L (LIKE) 
 AFTER THE SUBJECT 
 
 SA. IND. M.E. 
 Shop work 20 83 95 
 Economics 80 50 86 
 Sociology 67 26 60 
 Mechanical drawing 33 53 88 
 Manual training 30 63 93 
 Ancient languages 13 17, 4 
 Civics mg, 63 86 
 Public speaking 53 40 53 
 
 3. Per cents who constructed a piece of furniture or a household 
 appliance other than in a school shop: 
 
 PER CENT 
 Succ. 1. i. salesmen 43 
 Industries seniors 75 
 Mech. Eng. seniors. 79 
 
 7. Percents making collections of 
 STAMPS COINS POSTAL CARDS 
 
 Succ. 1. i. salesmen 59 50 48 
 Industries seniors 44 44 41 
 Mech. Eng. seniors 55 ‘AI 46 
 8. Per cents who owned cameras: 
 PER CENT 
 Succ. 1. i. salesmen 90 
 Industries seniors 75 
 
 Mech. Eng. seniors 72 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 55 
 
 Per cents preferring to take pictures of people rather than 
 scenes : 
 
 PER CENT 
 Succ. 1. i. salesmen ; 28 
 Industries seniors 23 
 Mech. Eng. seniors 30 
 
 18. Magazines read regularly: 
 , NUMBER WHO READ THE MAGAZINE 
 REGULARLY AMONG 
 
 SA. IND. M.E. 
 Literary Digest 12 8 5 
 American Magazine 20 7 ug 
 Saturday Evening Post 9 nig 19 
 System yi oO 0 
 Annalist 0) 5 (0) 
 Outlook O 10 oO 
 
 19. Favorite magazine: 
 
 ; SA IND. M.E 
 Literary Digest 5 4 I 
 American Magazine 15 8 10 
 Saturday Evening Post 6 4 10 
 
 20. Per cents who have had the impulse to express their opin- 
 ions on current topics in letters to newspapers : 
 
 PER CENT 
 Succ. 1. i. salesmen 62 
 Industries seniors 50 
 Mech. Eng. seniors 5I 
 
 22. Per cents who have written poems : 
 
 PER CENT 
 Succ. 1. i. salesmen 21 
 Industries seniors 33 
 Mech. Eng. seniors 44 
 
 The same proportions between the groups held for those who 
 had poems appearing in print. 
 28. Median age at which the groups learned to dance: 
 
 MEDIAN AGE (YEARS) 
 
 Succ. 1. i. salesmen 15 
 Industries seniors 17.5 
 Mech. Eng. seniors 18 
 
 30. The salesmen have engaged in each of the following activi- 
 ties to approximately twice the extent of the industries or mechan- 
 
56 MAX FREYD 
 
 ical engineering students: dramatics, musical organizations, de- 
 bating, school politics, and public speaking. 
 37. Per cents who smoke: 
 
 PER CENT 
 Succ. 1. i. salesmen 80 
 Industries seniors 50 
 Mech. Eng. seniors 37 
 
 40. Per cents of men who spent a vacation traveling around for 
 the mere pleasure of traveling: 
 
 PER CENT 
 Succ. 1. i. salesmen 52 
 Industries seniors 30 
 Mech. Eng. seniors 41 
 
 42. Median number of states in which a month or more was 
 spent: 
 
 MEDIAN 
 Succ. 1. i. salesmen 5.25 
 Industries seniors 3.00. 
 Mech. Eng. seniors 2.00 
 
 43. Median number of cities and towns lived in, for six months 
 or more: 
 
 Succ. 1. i. salesmen 4.3 
 Industries seniors 3.0 
 Mech. Eng. seniors 2.3 
 49. Numbers having (only) a few close friends: . 
 FREQUENCY 
 Succ. 1. i. salesmen oO 
 Industries seniors 5 
 Mech. Eng. seniors 2 
 
 Numbers having both close friends and a great many ordinary 
 friends: 
 
 FREQUENCY 
 Succ. 1. i. salesmen 29 
 Industries seniors 22 
 Mech. Eng. seniors 36 
 
 51. Median number of persons with whom a social correspon- 
 dence is maintained: 
 MEDIAN NUMBER 
 
 Succ. 1. i. salesmen Y 4 
 Industries seniors 5 
 Mech. Eng. seniors 3.5 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 57 
 
 52. Median number of social letters written per week: 
 
 MEDIAN NUMBER 
 
 Succ. 1. i. salesmen a0 
 Industries seniors 3.30 
 Mech. Eng. seniors 1.90 
 53. Religious beliefs: 
 PER CENT 
 CATH. PROT. JEWISH NO PREF, NON-BEL. 
 Succ. 1. i. salesmen 3 90 (a) 3 3 
 Industries seniors 23 67 3 7 fe) 
 Mech. Eng. seniors 3 85 5 3 5 
 
 55. Per cents who have tried to convert anyone to their re- 
 ligious point of view: 
 
 PER CENT 
 Succ. 1. i. salesmen 18 
 Industries seniors 37 
 Mech. Eng. seniors 19 
 
 60. Per cents believing in: 
 
 SA IND. M.E 
 The evil eye 6 fe) 2 
 Yogi philosophy 13 fe) oO 
 Fletcherism a7 10 37 
 Vivisection 47 23 4 
 Memory systems 77 43 74 
 Pelmanism 27 13 16 
 Mental telepathy 60 30 63 
 New Thought 20 3 16 
 Osteopathy 57 37 WE: 
 Phrenology 40 10 49 
 Astrology 27 43 14 
 Character analysis (Blackford) 63 50 79 
 Vegetarianism 33 47 37 
 
 68. Per cents believing: 
 PER CENT 
 SA, IND. M.E. 
 
 That they have had less than their 
 share of luck 13 27 a 
 
 71. Per cents who have fainted away: 
 
 PER CENT 
 Succ. 1. i. salesmen 37 
 Industries seniors 10 
 
 Mech. Eng. seniors 26 
 
58 MAX FREYD 
 
 74. Personages considered their ideal by the three groups: 
 
 FREQUENCY 
 SA, IND. M.E. 
 
 Lincoln II 13 18 
 Roosevelt 8 I 3 
 Washington 4 3 3 
 Robert E. Lee 2 Ce) fe) 
 Napoleon I 4 2 
 Christ I I 3 
 Carnegie I ee ts) 
 
 The following personages are mentioned once by the industries 
 students: Vanderlip, Merriwell, Brutus, and Moses. The follow- 
 ing are mentioned once by the engineering students: Caesar, Han- 
 nibal, J. P. Jones, Grant, King Richard of England, Edison, Alex- 
 ander, and Aeneas. 
 
 Brill (2) calls the answer to this question the person’s empathic 
 index. The answer indicates the person’s trend of adjustment to 
 the world. A certain personage is admired because of a wish to 
 emulate his mode of reaction to the world. He expresses in his life 
 the conscious and unconscious strivings of those who consider him 
 their ideal. 
 
 Questionnaire 4. Likes and Dislikes 
 
 The most important group differences found are the following: 
 the salesmen show a greater proportionate liking for fat men, con- 
 ventions, “Life,” and interviews; the industries students show a 
 greater proportionate liking for very polite people and ‘““The New 
 Republic.” The per cents of the salesmen, industries students, and 
 mechanical engineering students, respectively, checking L! or L 
 after each of these items are: fat men: 55, 17, and 37; very polite 
 people: 47, 90, and 40; conventions: 73, 53, and 49; “Life”: 97, 
 77, and 93; “The New Republic” : 33, 77, and 14; interviews: 77, 
 33, and 51. 
 
 Questionnaire 5. Beliefs 
 Table III shows the most important group differences found in 
 attitudes toward the questions. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 59 
 
 Taste III 
 
 Group Comparisons in Reactions to Questions in Questionnaire 5. Percentages 
 checking Y (yes) among the two primary groups and the Case School 
 Seniors in Mechanical Engineering. 
 
 PER CENT CHECK- 
 ING Y (YES) 
 
 DO YOU BELIEVE SA. IND. M.E. 
 3. That prize fights should be prohibited? ..............000. 333 20un IG 
 5. That unstinted service is always rewarded? ..........000+ 70 40 49 
 Pueiaiat au men are Created, CUAll .) o «dscns dusesscwsaccsens [01s tae sk 
 
 8. That the United States should build a stronger navy than 
 
 AAU MOCCL TAUION Cites ck aera ees eh ae hee he hears a oe hae 4340, 1c co 
 
 10. That preparedness for war is the best guarantee of peace? 73 33 30 
 15. That modern writers are not as good as the classical writers? 50 30 32 
 foe Liateevery, Malninas- Nis: Price £2 co a tnccw hs wns eislne™ din ga sejeins 20s SO me 26 
 17. That Mars is inhabited by living beings similar to man?... 43 26 62 
 20. That few men of good taste or intelligence are found among 
 
 thetri chris ae eels ec nl Moree ee erate eternal ci we OP c0 4 
 24. That “spare the rod and spoil the child” is the best maxim 
 Py USE TAIY Paiste CUUOTEN face yeu at es vaso chases sees bo ee Te Bae: 
 
 32. That Americans are stronger physically than Europeans?.. 93 63 65 
 33. That the more children a man has, the greater his value to 
 
 FUCRCOMIUIUY 287 were sis ak Mara in Dele bis cere ew he ie 20 AS ees 
 42. That the boy who does poorly in school usually turns out a 
 STICCESS UE Tan ahetetee ceerahirern eet eerhoele ss einciere a nai e's 6 26 9 
 
 44. That the accumulation of money makes a man heartless?.. 20 46 20 
 45. That the government should grant pensions to the aged?.. 46 67 49 
 
 58. That a man’s clothes are an index of his character? ...... 70 40 60 
 ee bmatewarels CVer TUStMADICL. vccas 56s he cei esin's soe wee ruie 83. 50 SI 
 64. That the union principle of the closed shop is better than 
 the open.shop for American industry?.............. 16 36 39 
 65. That Abraham Lincoln is the foremost man in American 
 DSTEV Sates avalos oe nie ivi’ faa et amare 0's ao ciel mee 20 50 76 
 SUMMARY 
 
 The outstanding features of the data presented in the preceding 
 pages may be summarized as follows: 
 
 OQuestionnatre I. 
 
 (1) A greater proportion of the sales group than the other two 
 groups are of native or British Protestant stock. Their parents 
 were better educated and wealthier than those of the industries 
 students. They were thus afforded the advantage of a better social 
 adaptation in this country, which may have been a factor in their 
 choice of salesmanship as a vocation. 
 
¢ 
 
 60 MAX FREYD 
 
 (2) The industries students were apparently subjected to great- 
 er maternal influence in childhood than the salesmen. We find that 
 more industries students than salesmen “‘took after” their mothers; 
 more resembled the mother’s side of their ancestry; more of them 
 liked their mothers in preference to their fathers; more of them 
 state that their mothers exercised the most influence on their lives; 
 and more of them state that their mothers were dominant in the 
 family. A possible explanation is that the predominant influence of 
 the mother may have caused the industries students to become less 
 successfully adapted toward other individuals. They may have 
 clung to their mothers’ apron-strings for too long a period in child- 
 hood. On the other hand, the attachment to the mother may have 
 been symptomatic of more fundamental factors influencing per- 
 sonality. We should expect to find among the industries students 
 compensatory reactions for the weaknesses expressed in their atti- 
 tudes toward their mothers, in that they may have taken an ag- 
 gressive attitude toward their social environment instead of a tact- 
 ful one. 
 
 - (3) The industries students were evidently not wholly satisfied 
 with their early environment. A greater proportion of them than 
 of the salesmen would desire a change in their home and other 
 early environmental conditions if they were to live their lives over 
 again. More of them would prefer a wealthier environment, more 
 refined companions, and more cheerful conditions. They were 
 treated slightly worse than their brothers and sisters. Their early 
 lack of tactful social adjustment is demonstrated by the fact that — 
 they engaged in more-fistfights as children than the salesmen. 
 
 Questionnaire 2. 
 
 (4) In their choice of occupations the salesmen tend to choose 
 the characteristically social occupations and the industries students 
 the mechanical ones. Occupations preferred in greater proportion 
 by the salesmen are: actor, automobile salesman, editor, foreign 
 correspondent, lawyer, magazine writer, newspaper reporter, nov- 
 elist, philanthropist, politician, preacher, real estate salesman, 
 specialty salesman, and surgeon. The industries students are more 
 inclined toward the following occupations: astronomer, auto fe- 
 
THE SOCIALLY AND THE*-MECHANICALLY INCLINED 61 
 
 pairman, bank teller, baseball player, carpenter, draftsman, ex- 
 plorer, factory manager, hotelkeeper or manager, labor arbitrator, 
 locomotive engineer, machinist, mechanical engineer, musician, 
 office manager, promoter, ship officer, shop foreman, stock broker, 
 toolmaker, and watchmaker. It is an interesting fact that salesmen 
 prefer literary occupations, while the industries students prefer 
 financial occupations. 
 
 Questionnaire 3. 
 
 (5) In their attitudes toward school subjects the industries 
 students show the expected liking for shop work, mechanical draw- 
 ing, and manual training. The salesmen prefer economics, sociolo- 
 gy, civics, and public speaking. 
 
 (6) The salesmen show their social proclivities in that they 
 learned to dance earlier in life and maintain a greater social cor- 
 respondence than the industries students. They probably make 
 themselves less objectionable by not trying to convert people to 
 their religious point of view, although in this respect the indus- 
 tries students show a more marked sales tendency than the sales- 
 men themselves. 
 
 (7) Salesmen are more credulous in that they tend toward a 
 greater belief in pseudosciences. 
 
 (8) Among the salesmen the ideal personage is more often 
 Roosevelt and less often Lincoln. Roosevelt expresses a more 
 normal vigorous empathic index. Lincoln expresses an ethical and 
 conscientious attitude toward life. 
 
 Questionnaire 4. 
 
 (9) Salesmen show a greater preference for fat men, conven- 
 tions, interviews, and “Life.” Industries students show a greater 
 preference for very polite people and “The New Republic.” The 
 connotations of a preference for good-natured fat men and for 
 “Life” are obvious. Salesmen probably show less liking for very 
 polite people because such people are less free in their social ex- 
 pressions and are more difficult prospects to handle. 
 
 Questionnaire 5. 
 
 (10) In their beliefs salesmen show greater credulity and con- 
 
62 MAX FREYD 
 
 formity with popular opinion than the industries students. The lat- 
 ter display to a great extent the laboring man’s attitude. 
 
 In most respects the mechanical engineering students occupy an 
 intermediate position, although their leanings are in the direction 
 of the industries students rather than the salesmen. We may con- 
 clude that engineering draws men with slightly greater social pro- 
 clivities than a course such as the one which the industries students 
 follow. 
 
CHAPTER TV 
 
 GROUP COMPARISONS BASED ON RATINGS 
 
 Group comparisons based on ratings were made in order to deal 
 with a number of traits, (some of which presumably were also 
 measured by the tests and questionnaires), which seemed to be re- 
 lated to the personalities studied. Ratings were obtained on only 
 three groups: the two primary groups and the Case School seniors 
 in mechanical engineering. The data include self-ratings by the 
 members of these groups, plus ratings on them by acquaintances. 
 
 THE METHOD 
 
 All the ratings reported here were made on the same scale, a 
 graphic rating scale which afforded ratings on twenty traits. The 
 graphic rating scale is a form in which the rating is indicated by 
 a check along a straight line, under which are printed descriptive 
 phrases indicative of varying degrees of the trait, from one ex- 
 treme to the other. The advantages of this form of scale are set 
 forth in another place by the writer (6). The directions for the 
 present scale and an illustrative item are given below. 
 
 CCA ENT GerA sl ON Ge © Boe cy tetck ree serra cere aie Seine caro ne enero eure err 
 INSTRUCTIONS FOR USING THE RATING SCALE 
 
 . Let these ratings represent your own judgments. Please do not consult anyone in making 
 
 them. 
 
 . In rating this person on a particular trait, disregard every other trait but that one. Many 
 ratings are rendered valueless because the rater allows himself to be influenced by a general 
 
 favorable or unfavorable impression which he has formed of the person. 
 
 ; When you have satisfied yourself on the standing of this person in the trait on which you 
 are rating him, place a check at the appropriate point on the horizontal line. You do not 
 have to place your check directly above a descriptive phrase. You may place your check at 
 
 any point on the line. 
 Is he absent-minded or wide awake? 
 
 Jery absent-minded. Often becomes Usually present- Always wide awake and 
 ontinually absorbed abstracted and out of minded alive to present 
 n thought touch with his situation 
 
 surroundings 
 
 In the reproduction the length of the line on which the rating is 
 made is reduced. In the original it is 65% inches long. 
 
64 MAX FREYD 
 
 The “good” ends of the scales were alternated in order to coun- 
 teract a motor tendency to check at one edge of the blank. The 
 time required for making the twenty ratings was approximately 
 ten minutes. 
 
 To score the ratings a stencil was used, one of whose edges was 
 marked off for a distance equal to the length of the graphic rating 
 line. This space was divided off into twenty spaces of equal length, 
 consecutively numbered from left to right. The stencil was placed 
 beneath the line on which the rating was made, so that this line 
 coincided with the marked-off space on the stencil. The score was 
 the number of the space over which the check was made. If the 
 check was made over a division line between two spaces, the score 
 to the right was assigned to the rating. An X-shaped check was 
 scored at the intersection of the two lines, and a V-shaped check 
 at the point of the V. If more than one check was made on the 
 same scale for the same person, indicating doubt on the part of the 
 rater, the average of the ratings was taken. Since the same stencil 
 was used for all the scales, all scores read from left to right. No 
 total score was obtained. 
 
 The following additional directions were used in obtaining self- 
 ratings from the industries students. Similar directions were used 
 in obtaining self-ratings from the other two groups. 
 
 “Tt is requested of each senior in Industries that he fill out the 
 enclosed blank and mail it to this office. Our purpose in asking for 
 this information is to provide a check on the tests which were re- 
 cently given, by comparing ratings on character traits with the © 
 test scores. 
 
 “The rating scale which is enclosed was not intended for self- 
 rating, but we wish you to use it in that way. Let the ratings which 
 you make represent your judgments on your own traits. That is, 
 you are to judge yourself, taking the point of view of an impartial 
 observer. If you care to, you may add marginal comments to make 
 your ratings clearer or to supplement them. 
 
 “We wish to make it absolutely clear to you that these ratings 
 will be kept confidential. None of the records will go out of this 
 office. They will not be made known to the employment office or to 
 any of your instructors. Do not sign your name to the blanks.” 
 
 The scales were numbered so that they could be identified 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED _ 65 
 
 when returned. The industries students returned their ratings by 
 mail, but the other two groups rated themselves in the classroom, 
 with instructions similar to the above. In order to get ratings by 
 acquaintances on these men, as checks on their self-ratings, each 
 member of the three groups gave the names of five men to whom 
 the scale could be sent. To each of these references were sent a 
 copy of the scale and a letter requesting ratings on the student. 
 Out of approximately 400 rating scales sent out, 80 per cent were 
 returned completely filled out. Except in rare instances, no two 
 men were rated by the same judge; that is, with rare exceptions, 
 no two men gave the same reference. All self-ratings and ratings 
 by acquaintances were made anonymously. 
 
 Distribution tables of self-ratings and ratings of others were 
 drawn up and the interquartile ranges were calculated. The ratings 
 of others include all ratings, and since some men were rated by 
 five references and others by merely one, the ratings of others do 
 not completely represent the groups. All of the engineering stu- 
 dents were rated by three or more people, but about one-fifth of the 
 other two groups were not rated at all by others. 
 
 RESULTS 
 
 The interquartile ranges of the self-ratings and the ratings of 
 others for the three groups are shown graphically in Fig. 8. In 
 order to make the results more apparent and to bring out features 
 which would not otherwise be noted, the rating scale instead of the 
 scale of scores is placed below the lines. 
 
 There are several outstanding facts about these ratings. 
 
 (1) The general tendency is for raters to make checks directly 
 above descriptive phrases. This is to some extent an error in the 
 use of the scale which should be overcome with practice, and to 
 some extent a weakness in the scale itself. 
 
 (2) Several of the scales, especially those for Traits 1 and 2, 
 are shown by the interquartile ranges to be defective. The central 
 phrase in every scale should be a neutral one, or should express 
 the most common condition of the trait found in the population 
 at large. The defects of these two scales are made evident by the 
 displacement of the distributions, due to the fact that the central 
 
66 MAX FREYD 
 
 phrases are more nearly allied in meaning to the unfavorable than 
 the favorable ones. 
 
 (3) Several other scales,—those for Traits 16 and 17, for in- 
 stance,—are defective because, as indicated by the narrow range 
 of ratings, the intermediate phrases are more nearly like the ex- 
 tremes than they are like the central phrase. Raters therefore 
 tended to avoid them. 
 
 (4) The general trend is for acquaintances to rate the men 
 higher or more favorably than they would rate themselves. This is 
 true for Traits 1, 2, 3, 4,6, 7,9, 11, 12,14, 19, and 20. Whether this 
 is due to self-depreciation on the part of the subjects of the study 
 or to flattery on the part of their acquaintances, cannot be deter- 
 mined. As we are not in possession of the facts, the ratings must be 
 accepted at face value. 
 
 (5) The differences between the groups in self-ratings and in 
 the ratings of others are in most cases slight, but they gain sig- 
 nificance from the fact that in almost every case where a difference 
 appears in self-ratings it is corroborated by the ratings of others, 
 and from the fact that the differences are consistent. The direc- 
 tions of the differences between the successful life insurance sales- 
 men and the industries seniors are shown below, based on self- 
 ratings. 
 
 1. Salesmen are more wide-awake 
 2. Salesmen are more good-natured 
 3. Salesmen are neater 
 4. Salesmen are more excitable 
 7. Salesmen are less self-conscious 
 g. Salesmen have a more impressive bearing 
 *10, Salesmen are more self-confident 
 II. Salesmen are more even-tempered 
 13. Salesmen are more adaptable 
 *14. Salesmen make friends more quickly © 
 *15. Salesmen are more open-hearted 
 18. Salesmen are more talkative 
 *20. Salesmen work faster 
 
 The differences starred are those which may be called very sig- 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 67 
 
 nificant, according to the criterion for the significance of differ- 
 ences explained in Chapter 2. 
 
 The above differences are reversed by the ratings of others in 
 Traits 9, 11, and 20. Every one of the differences but these three is 
 corroborated by the ratings of others on the men. In addition, the 
 differences are borne out by the self-ratings of the seniors in 
 mechanical engineering. These self-ratings differ from the self- 
 
 ceorsume nist ames occa et 
 Sales 
 ——— 
 eonenee cree 
 Industries 
 a 
 cenerfecammaiemc ener 
 Mech. Eng- 
 ef 
 Very absent-minded. Often becomes Usually present- Always wide awake and 
 Continually absorbed abstracted and out of minded alive to present 
 in thought touch with his situation 
 surroundings 2 
 Trait 2 
 me Rep 
 Sales 
 eee 
 Ind. 
 —__}-—____ 
 Me E. 
 —j—_ 
 Very good-natured. Agreeable Rather glum and Grouchy and V i 
 nites : ll-natured 
 Has winning unresponsive unpl bf ne 
 manner : npleasant and uncivil 
 Trait 2 
 eon tess ee 
 Sales 
 —_—}—. 
 Sores Seen 
 Ind. 
 a oo 
 ooceerencomenares 
 Me E. } 
 Extremely neat Appropriately and Inconspicuous Somewhat careless Very slovenly 
 and clean. Almost ~ neatly dressed in dress in his dress and unkempt 
 a dude. 
 Trarus 
 
 Fic. 8a. Interquartile Ranges of Ratings. 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 the ratings of others. 
 
68 MAX FREYD 
 
 ratings of the salesmen in the same direction as do those of the in- 
 dustries seniors in every one of the traits in the above list. In ap- 
 proximately half the traits they differ more from the salesmen than 
 the industries seniors do, but in every case this difference is in the 
 same direction as the industries seniors. 
 
 Eee reo Saco 
 Sales 
 nana 
 Industries BA 
 comers 
 Meche Eng. 
 Very excitable and Easily stirred Usually cool and Always cool-headed 
 high-strung self-contained and collected 
 Trait 4 
 See aT 
 Sales i 
 Industries 
 ee 
 Mech. Eng. ‘ 
 seer i in Asserts himself Very aggressive 
 Very submissive Saree bh vhe frequently and insistent 
 Trait 5 
 Ee eT 
 Sales 
 ——_—_—_——" 
 a ed 
 Industries 
 ooo, 
 oemoramren eum 
 Meche Eng 
 —————_—_—_— 
 d Careful Usually accurate Careless Tnexact and 
 Med dacireer and reliable negligent 
 Trait 6 
 
 Fic. 8b. Interquartile Ranges of Ratings. 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 the ratings of others. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 69 
 Oe 
 Sales 
 a 
 Industries 
 _—_—_—— HY 
 EE 
 Meche Eng. 
 a een 
 Always at ease Seldom flustered by Self-conscious Frequently Painfully self-con- 
 
 actions or remarks on occasions embarrassed scious and ill at 
 
 with reference ease. Very sensitive 
 
 to himself 
 
 Tra ite. 
 se celuesieseraiaiemmmmansianadaen aaememmenimaammmial 
 Sales 
 era a TECTED 
 Industries 
 ep ET 
 
 Mech. Eng. 
 
 Always acts on 
 the spur of the 
 montent 
 
 Sales 
 Industries 
 
 Mech. Eng» 
 
 Looked down on 
 
 Fic. 8c. 
 
 Impulsive. Always Shows moderate Cautious, deliberate Extremely wary and 
 
 makes prompt deliberation and considerate hesitant. Acts only 
 
 decision after careful con- 
 sideration 
 
 Trait 8 
 
 eorsmenpecen 
 __—t+ 
 Se ST ETE RR 
 See 
 SSR eT a MEETS 
 jnecnnmencare esas 
 
 Excites admiration. 
 Very impressive 
 
 Noticeable for good 
 
 Unimpressive physique d 
 physique and bearing 
 
 and bearing 
 
 Trait 9 
 
 Interquartile Ranges of Ratings. 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 
 the ratings of others. 
 
70 
 
 Sales 
 
 Industries 
 
 Mech. Enge 
 
 Judges himself 
 capable of 
 
 Exaggerates 
 his abilities 
 
 MAX FREYD 
 
 ee 
 eee et es 
 
 Knows just what 
 he is capable of 
 
 Underestimates his 
 own abilities 
 
 anything 
 
 Sales 
 
 Industries 
 
 Meche Eng. 
 
 Often has extreme 
 ‘ups or downs in 
 
 mood. Shows ela- 
 tion or depression 
 
 Sales 
 
 Industries 
 
 Mech. Eng- 
 
 Never speaks 
 depreciatingly 
 of others 
 
 Fic. 8d. 
 
 Trait 10 
 
 a 
 ee 
 Usually even- Even-tempered. 
 
 Has ups or downs Shows change of 
 
 at times without feelings when con- tempered 
 apparent reason ditions warrant 
 Trait 1l 
 rasa npr ar er 
 ———_— 
 orrreesar RTE 
 st 
 Seen gee 
 +———_—_— 
 Rarely criticises  Coniments on out- Criticises 
 others standing weaknesses others 
 
 or faults of others 
 
 Trait 12 
 
 Interquartile Ranges of Ratings. 
 
 Considers himself 
 incapable of much 
 success 
 
 Maintains the same 
 mood in spite of 
 cause for changing 
 
 Extremely critical 
 of others 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 
 the ratings of others. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 
 
 71 
 Bree ele erase eT 
 Sales 
 ee 
 Industries 
 —_—t+ 
 AED RARER SNEED A OTE 
 Mech. Eng. 
 re re 
 Hidebound. Slow to take up Progressive Quick to pick Is always adapting 
 Runs in a rut new ideas tendencies up new ways himself and taking 
 and habits up new ideas 
 Trait 13 
 amo reap isfor ns 
 Sales 
 es enna 
 hen eeeeeredem nee 
 Industries 
 ———_} —__. 
 Meche Eng. 
 
 ——E——— 
 
 Makes friends quickly 
 
 Has quite a number 
 and easily. Very 
 
 More interested in ideas 
 of friends 
 
 Lives almost 
 or things thanin persons by Ricans entirely 
 popular 
 Trait 14 
 GER ERNIE LET PRR OE SOE 
 Sales 
 ee 
 PEE LTE LE 
 Industries 
 Ht 
 ee RR 
 Mech. Eng. 
 
 th 
 
 Often confesses his 
 thoughts and feelings 
 to friends and 
 acquaintances 
 
 At times unburdens 
 spontaneously to 
 
 Will occasionally Never unburdens. 
 friends 
 
 unburden when Rarely talks about 
 questioned himself 
 
 Trait 15 
 
 Fic. 8e. Interquartile Ranges of Ratings. 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 the ratings of others. 
 
72 MAX FREYD 
 
 emersseat 
 Sales 
 + 
 cE AD 
 Industries 
 Sere 
 cepa 
 Meche Eng; 
 oe 
 Avoids companion- Associates with Shows a normal, Gives much time Associates more 
 ship of women women wholesome interest to the ladies with women than 
 infrequently in feminine society with men. A 
 Jady’s man 
 Trait 16 
 er dE 
 Sales 
 eenl ores 
 Industries 
 mela 
 Mech. Eng- 
 ee 
 Daredevil. Shows Will take every Gets “cold feet” Risks nothing. 
 great “nerve” reasonable chance Takes no-chances 
 Trait 17 
 eer 
 Sales 
 eaten ran rr EER 
 du 8 
 Industrie 
 = 
 Mech. Eng. 
 > 
 Talks seldom. Does not uphold his Moderately More than upholds Great talker. 
 When questioned end of the con- talkative his end of the Always going 
 answers briefly versation conversation 
 Trait 18 
 
 Fic. 8f. Interquartile Ranges of Ratings. 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 the ratings of others. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 73 
 
 excercise 
 Sales p 
 SSR LE 
 Industries 
 Ee eee 
 Meche Eng- 
 ———___ + ____ 
 ia- Shows poor Shows some.appre- Shows good Has excellent taste 
 oe lcs hai taste r ciation of artistic artistic judg- in art, music and 
 Taste runs to the value ment literature 
 cheap and ugly 
 Trait 19 
 erence 
 Sales 
 ee ee Sere meer pene 
 Industries 
 emer sien ferns 
 e Eng. 
 Mech g 
 Ext ly ‘quick. Fast worker Works just fast Slow worker. 
 Raphi worker enough to get by Barely moves 
 Trait 20 
 
 Fic. 8g. Interquartile Ranges of Ratings. 
 
 Vertical lines indicate medians. Heavy lines indicate self-ratings and light lines 
 the ratings of others. 
 
 SUMMARY 
 
 Although the rated differences are not so great as to withstand 
 detailed statistical examination, they gain significance from the 
 fact that they are psychologically consistent, that self-ratings and 
 the ratings of others corroborate each other in most of the traits, 
 and that the differences between the primary groups in self-ratings 
 are borne out by the self-ratings of engineering seniors. 
 
 The most significant differences between the primary groups 
 are in Traits 4, 10, 14, and 15. They lead to the conclusion that 
 salesmen are more excitable, more self-confident, and more open- 
 hearted than industries seniors, and make friends more quickly. 
 
 There is a less significant tendency for salesmen to be more wide- 
 awake, good-natured, adaptable, and talkative, neater, and less 
 self-conscious than industries seniors. 
 
GHAP DER W 
 
 INTERCORRELATIONS OF TESTS AND RATINGS 
 
 Though not an application of Spearman’s method of proving 
 the existence of a general common factor, the method to be de- 
 scribed is related to it in such an intimate way that it will be neces- 
 sary to give first a brief summary of the development of Spear- 
 man’s two factor theory (3, 10, 18). 
 
 THE THEORY OF GENERAL ABILITY 
 
 Spearman first put forth his theory of general ability in 1904 
 (18). He found that after intercorrelating a number of measure- 
 ments of mental abilities, the correlation coefficients could easily 
 be arranged in a hierarchical order, 7.e., so that on the whole each 
 coefficient was greater than any to the right of it in the same row 
 or below it in the same column. 
 
 “This fact of ‘hierarchical’ order which he had thus discovered 
 was taken by Professor Spearman to indicate the presence of some 
 common fundamental function which saturates in different degrees 
 the different activities, and is the sole cause of correlation between 
 them except in the case of very similar activities. 
 
 “Tt can easily be shown that if all the correlations are due solely 
 to one common or general factor, then the correlation coefficients - 
 will be in perfect hierarchical order. 
 
 “Since clearly perfect hierarchical order cannot be expected in 
 any experimental research, it becomes important to. know what 
 deviation from prefection can be allowed without giving up the 
 idea of a general factor: or on the other hand, what approach to 
 perfection can be attained without the presence of a general fac- 
 tor.” (3, pp. 165-166) 
 
 Spearman devised a formula for correcting for observational 
 errors which, applied to his correlation coefficients, enabled him 
 to arrange them in perfect hierarchical order. The suppositions 
 underlying this formula have been called in serious question by 
 
 Thomson (3). 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 75 
 
 Researches between 1904 and 1912 differed in the degree of 
 hierarchical order obtained. In 1912 Hart and Spearman pro- 
 posed a criterion for the existence of a hierarchical order, whose 
 underlying idea was that the correlation of adjacent columns of 
 correlation coefficients, if perfect hierarchical order existed, would 
 equal unity, and that its value is therefore a measure of the com- 
 pleteness of the hierarchical order. 
 
 We need not concern ourselves in detail with the criticisms lev- 
 eled against the theory by Thomson, aside from noting that he 
 produced hierarchical order through a random overlapping of 
 group factors in the absence of a general factor, and that he at- 
 tacks the validity of Spearman’s correctional formulae. He re- 
 places Spearman’s two-factor theory (a factor specific to the 
 ability, and general common factor: intelligence) by a sampling 
 theory of ability, preferring “to think of a number of factors at 
 play in the carrying out of any activity such as a mental test, these 
 factors being a sample of all those which the individual has at his 
 command.” (3, p. 188) 
 
 MeETHOD EMPLOYED IN THIS STUDY 
 
 A possibility of the table of intercorrelations which seems to 
 have been overlooked by these writers is to study through them the 
 differences existing between groups as regards the interrelation- 
 ship of their traits. It cannot be said that differences greater than 
 those to be expected by chance between two tables of intercorrela- 
 tions of the same measurements obtained from two different 
 groups of subjects would account for two general factors, but it 
 seems possible that these differences would at least show what 
 specific factors were common to the members of a group. Sup- 
 pose that in Group A we find the three highest intercorrelating 
 measurements to be 1, 2, and 3; and in Group B the three measure- 
 ments showing the highest intercorrelations to be 8, 9, and 10; 
 all these intercorrelations to be over +-.90. Even though these dif- 
 ferences were too great to be ascribed to errors of observation, 
 they could not be attributed to the degree to which the variables 
 were saturated with a general common factor. They must be due 
 to the constitution of the two groups, and these varying factors in 
 
76 MAX FREYD 
 
 the two groups can only be described in terms of the intercorre- 
 lating traits themselves. Thus, for unknown reasons, Traits I, 2, 
 and 3 always accompany each other in equal degrees in the mem- 
 bers of Group A, forming a constellation of traits for that group; 
 similarly, Traits 8, 9, and 10 form a constellation for Group B. 
 
 A second possibility is to make tables of the intercorrelations 
 of traits obtained from merging the two groups, and then having 
 found what traits form constellations, all men considered, to de- 
 termine to what degree the groups differ in the extent of the con- 
 stellation which they possess. If, for instance, we should find 
 that Traits 5, 6, and 7 formed a constellation with the data from 
 all the groups combined, it would be interesting to note if the 
 groups differed in their possession of these traits, the latter accom- 
 panying each other in a large amount in one group and in a small 
 amount in the other group. This could be determined by comparing 
 distribution tables of the amounts of each trait possessed by the 
 two groups. 
 
 I have used both of these procedures with the groups on which 
 the figures were available: the two primary groups and the Case 
 School seniors in mechanical engineering. The correlations were 
 all obtained by the Spearman fourfold table method, which 
 although merely an approximation and subject to considerable 
 error, is perhaps as reliable as any, considering the number of 
 persons in the groups. The following tables of intercorrelations of 
 measurements were obtained : 
 
 1. Intercorrelations of self-ratings for the sales primary group. 
 
 2. Intercorrelations of self-ratings for the mechanical primary 
 group. Trait 17 was omitted, as the form of distribution of this 
 trait did not permit of the calculation of correlations. 
 
 3. Intercorrelations of self-ratings for the Case School seniors 
 in mechanical engineering. Trait 17 was omitted. 
 
 4. Intercorrelations of average ratings for 84 cases. These 84 
 cases were members of the three preceding groups who had been 
 rated by two or more acquaintances. The average rating for any 
 man in any trait was a figure obtained by first averaging the opin- 
 ion of his acquaintances on this trait, and then averaging this com- 
 bined opinion with the man’s self-rating on that trait, giving his 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED rip 
 
 self-rating and the combined judgment of his acquaintances equal 
 weight. 
 
 5. Intercorrelations of all test scores for the sales primary 
 group. 
 
 6. Intercorrelations of all test scores for the mechanical pri- 
 mary group. 
 
 7. Intercorrelations of all test scores for the 84 men mentioned 
 in connection with the fourth table above. 
 
 8. Correlations of all test scores with the average ratings of 
 these 84 men. 
 
 These tables of intercorrelations are not reproduced here. 
 
 In calculating the correlation coefficients where ratings were 
 concerned, one end of the range of actual ratings was arbitrarily 
 accepted as the better end. In many of the scales either end repre- 
 sents an extreme which is undesirable, but checks were rarely 
 made over these extremes. Within the limits of distributions of 
 checks, the good end is easily distinguished. In correlating ratings, 
 the rating at the ‘good’ end of the scale was called high, and the 
 trait was called by the ‘good’ name. In doing this numerical scores 
 were disregarded as expressions of the amount of the trait, and 
 were used merely to obtain medians. The effect is as if the good 
 end of the scales were always to the right instead of being alter- 
 nated, and as if a rating to the right of the median were consid- 
 ered high and to the left low, and the name of the trait were always 
 the right extreme. 
 
 In obtaining intercorrelations where tests were concerned, a 
 high score represented a high standing. 
 
 It is unfortunate that writers on the general factor theory give 
 us no method for the closest approximation to hierarchical order 
 where that order is not immediately and obviously made. Brown 
 and Thomson mention in passing two methods, or rather two 
 variations of the same method (3, p. 177, footnote, and p. 185). 
 The method rests on the probability that the ability represented 
 in the upper row and left-hand column in a hierarchical order 
 approaching perfection will have a larger total correlation than 
 any other ability represented in the table, and that the total cor- 
 relations of the other abilities will be in the same order as their 
 
78 MAX FREYD 
 
 position in the table. A brief method of ranging the measured 
 abilities in hierarchical order is, therefore, to range them in the 
 order of the sums of their correlations with the other abilities 
 (their total correlation). 
 
 In doing this we are immediately confronted with the question: 
 shall we take the algebraical totals of the correlation coefficients 
 or shall we disregard the signs in computing the sums? In the 
 one example given by Brown and Thomson all the coefficients are 
 positive. If, however, the hierarchy were arranged with only high 
 positive coefficients in the upper left-hand corner, we should ex- 
 pect a number of high negative correlation coefficients in the lower 
 right-hand corner. The last situation is practically impossible, since 
 a group of traits cannot show very high negative intercorrelations. 
 It is possible for a group of traits to show a great many high posi- 
 tive intercorrelations and a few negative ones. Thus, Traits 1, 2, 
 and 3, may intercorrelate very highly (averaging ++.90, for ex- 
 ample), and Trait 4 may correlate negatively with each of these 
 traits. The constellation of traits would then consist of Traits 1, 
 2, 3, and 4, the first three accompanying each other to the same 
 degree and being accompanied by inverse amounts of Trait 4. 
 
 In obtaining total correlations, therefore, the signs of the cor- 
 relation coefficients have been ignored. 
 
 RESULTS 
 
 As the tables of intercorrelations will not be reproduced here, 
 the reader will have to keep in mind what they aim to present. 
 Each of the eight tables mentioned contains all the correlation 
 coefficients resulting when each of the variables (ratings or tests) 
 is correlated with each of the other variables with which the par- 
 ticular table is concerned. In addition, for each variable in each 
 table, the total correlation is given, a figure which represents the 
 sum of the correlation coefficients of that variable with each of 
 the other variables in the table, the signs of the coefficients being 
 neglected. 
 
 As a preliminary step, the total coefficients of the variables (self- 
 ratings) in the first table were correlated with those in the second 
 table and with those in the third table, and those in the second table 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 79. 
 
 with those in the third table. These correlation coefficients were 
 calculated by the rank order method. Thus in the first table, Trait 
 2 ranked first with a total correlation of 7.37, in the second table 
 Trait 2 ranked fourteenth with a total correlation of 4.77, and in 
 the third table Trait 2 ranked thirteenth with a total correlation 
 of 5.05. The correlation of the values of the total correlation coef- 
 ficients in the first and second tables is —.21, first and third —.13, 
 and second and third +.27. In other words, the figures for the 
 sales group correlate negatively with those for the two groups of 
 mechanics students, whereas the latter correlate positively with 
 each other. This indicates that the two groups of mechanics stu- 
 dents will resemble each other as regards their constellations of 
 traits more closely than either will resemble the sales group. 
 
 Out of each of the first three tables of intercorrelations the five 
 traits with the largest total correlations were chosen. The inter- 
 correlations of these traits are given below. 
 
 ‘Sales primary group (data from first table) : 
 
 TRAIT 2 II 19 15 6 
 2 +.64 -.51 —A5 73 
 II +.71 —.66 +.56 
 19 —.71 +.31 
 15 —.56 
 Mechanical primary group (data from second table) : 
 TRAIT. 5 8 18 I 4 9 
 8 —.51 +.51 —.59 .00 
 18 +.66 +16 +.977 
 I +.51 +.51 
 4 +.68 
 Case School seniors (data from third table) : 
 TRAIT 18 4 5 13 16 
 18 gl aero? tooy +.59 
 4 +.64 37 +.40 
 5 + .66 +.28 
 13 +.56 
 
 In each of the above cases the traits are arranged in the order 
 of their total correlations. In the first table, the five traits showing 
 the highest total correlations are in order, 2, I1, 19, 15, and 6. 
 
 In the sales group positive amounts of Traits 2, 11, 19, and 6, 
 
80 MAX FREYD 
 
 accompany negative amounts of Trait 15. The constellation for 
 this group is therefore the following: 
 2. Good-nature 
 11. Even temper 
 19. Artistic taste 
 6. Accuracy in work 
 15. Close-heartedness 
 It should not be assumed that these phrases describe the group. 
 They represent extremes of traits which accompany each other in 
 these individuals. It is just as logical to give the names of the op- 
 posite traits to the constellation. By reference to Fig. 8 we find that 
 the salesmen are more good-natured than the industries. students, 
 whereas the latter are more close-hearted. No important differ- . 
 ences appear with respect to artistic taste, accuracy in work and 
 even temper. The discriminating value of this constellation is 
 therefore very slight. oni 
 For the industries seniors Traits 18, 1, 4, and 9 inter-correlate 
 positively. Trait 8 shows no consistent relationship to the other 
 traits and is consequently omitted in naming the constellation: 
 18. Talkativeness 
 1. Present-mindedness 
 4. Cool-headedness 
 9. Good bearing 
 Referring to Fig. 8, salesmen excel in talkativeness and present- 
 mindedness, whereas the industries seniors excel in the other two 
 traits. | 
 The constellation for the seniors in mechanical engineering is: 
 18. Talkativeness 
 4. Cool-headedness 
 5. Self-assertion 
 13. Flexibility 
 16. Sociability with the other sex 
 No valuable conclusions as to group differences in constellations 
 of traits can be drawn from these results. 
 Turning to the intercorrelations of average ratings (fourth 
 table), we find the intercorrelations of the five traits with the 
 largest total correlations to be as follows: 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 81 
 
 TRAITS . 18 13 I 2 20 
 18 +.37 -+-.10 +.16 +.37 
 13 +.45 +.29 +.66 
 
 I +.43 +.51 
 2 +.16 
 
 The constellation, though not very coherent, consists of the fol- 
 lowing traits: 
 18. Talkativeness 
 13. Flexibility 
 1. Present-mindedness 
 2. Good-nature 
 20. Quickness in work 
 
 In every one of these traits the sales group excels the industries 
 students. Here we have the basis for a group differentiation as 
 regards a constellation of traits. These traits form a constellation, 
 perhaps in the general population, and also individually differen- 
 tiate the two groups. The terms in the above list are then the best 
 oties to apply to the general traits of the groups in differentiating 
 them. 
 
 Similar comparisons with regard to test scores are less reliable 
 since we have included several measures from the Kent-Rosanoff 
 Test which are very closely related. 
 
 The five test scores with the largest total correlations for the 
 sales group (fifth table) are all in the Kent-Rosanoff Test, being 
 Adjective and Adverb Responses, Noun Responses, Speech-Habit 
 Responses, Contrast Responses, and Responses with a Common- 
 ality Value of 1-5. The comparison of these results with those of 
 the sixth table produced entirely negative results. No differentiat- 
 ing constellations were discovered. 
 
 The intercorrelations of test scores for three groups combined 
 (seventh table) yield more favorable results. The intercorrelations 
 of the five tests with the largest total correlation are given below. 
 All are in the Will-Temperament Test. 
 
 TEST PT.9 PTS.IANDIO PT.9Q,F. POCO ee rr. Ast 
 Pt.9 +.03 —.61 + .31 +.19 
 Pts) wand-1o —.16 +.25 +.43 
 IDE Orat: —.48 —.37 
 
 Brie dL +.13 
 
82 MAX FREYD 
 
 The constellation consists of negative amounts of Part 9, ratio, 
 accompanying positive amounts of all the other tests. 
 
 An ideal arrangement would exist here if one group of subjects 
 excelled the other in all of these tests except Part 9, ratio, and if 
 every test in the five correlated highly with the constellation of 
 rated traits on the preceding page. We should then have found a 
 differentiating constellation of rated traits and adequate measures 
 cf them for use on other groups of subjects. 
 
 The sales primary group excels the industries group in Part 9, 
 Part 6, letters, and Part 4, ratio. The industries students excel in 
 Part 9, ratio, and very slightly in Parts 1 and 10. This seems to 
 offer a possibility that the ideal arrangement would be obtained. 
 The correlations of each of the five tests with each of the five 
 ratings are given below (from eighth table). 
 
 
 
 TEST TRAIT 18 13 I 2 20 
 Pt. 9 Paid 3) 4 =.257 9-349 Bog anodes 
 Pts. I and Io 00 +.22 +.16 +.16 00 
 PtOnie + .06 —.37 00 + .06 —.28 
 PteOnels +.13 +.25 +.03 +.13 +.03 
 PteAyeys —.16 .0O + .03 —.22 —.06 
 
 This random group of correlation coefficients closes this line of | 
 approach. 
 
 SUMMARY 
 
 The results of this attempt to find differentiating constellations 
 of traits are for the most part negative. This may be due partially 
 to the unreliability of the correlation coefficients, and partly to 
 the nature of the method for selecting the traits entering into the 
 constellation. 
 
 The intercorrelations of average ratings showed the most posi- 
 tive results. With the data from the three groups combined, the 
 following traits formed a constellation: 
 
 18. Talkativeness 
 
 13. Flexibility or adaptability 
 1. Present-mindedness 
 
 2. Good-nature 
 
 20. Quickness in work 
 
 In all of these traits the salesmen rate higher than the industries 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 83 
 
 seniors. The conclusion to be drawn from this is that the above 
 five traits form a constellation (as far as our data show) and that 
 the salesmen are distinguished from those who are mechanically 
 inclined in that the former possess these traits to a more marked 
 degree. Of all the rated differences between the two groups, these 
 five traits could most readily be represented by one distribution 
 curve on which salesmen and mechanically inclined people would 
 occupy distinct positions. 
 
CHAPTER VI 
 
 APPLICATION OF RESULTS IN A TOTAL SCORE 
 
 It is of interest to know how well the two general groups that 
 are dealt with in this study can be differentiated by a total score 
 which takes into consideration the significant individual items. If 
 the items can be combined in such a way as to produce a total score 
 which will clearly differentiate between men of extreme mechani- 
 cal tendencies and those of sales tendencies, we shall have a con- 
 venient mode of discovering the propensities of an individual, and 
 can use this information for vocational guidance along broad lines 
 rather than specific ones,—lines which will probably lead to greater 
 vocational happiness than any other method of vocational guidance 
 can afford. 
 
 We have found that no single test or other measurement com- 
 pletely differentiates the two primary groups. If the most sig- 
 nificant measurements are combined we shall have a total score 
 which, although it may not differentiate as well as the most sig- 
 nificant individual test, will yield the best combination of differen- 
 tiation and reliability. Although an individual may not conform 
 to his group in getting a high score in a certain test he may make 
 up for this by having all of the interests of his group, and this re- . 
 flected in his total score will place him properly. The extreme in- 
 dividuals will have a greater number of the interests and abilities 
 of their group than will any of the intermediate individuals. 
 
 Tue MetTHOD 
 The method of combining scores offers an interesting problem. 
 I have followed with some modifications, the procedure of Ream 
 (17). The ratings and all the questionnaires but 2 and 4 were left 
 out of consideration because of the difficulty of gathering the ma- 
 terial and scoring the blanks in any future applications of the 
 method. 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 85 
 
 The application of the method of combining scores to the ques- 
 tionnaires has already been explained (7). It need be necessary 
 only to review briefly what that procedure involves. 
 
 The problem is to discover the significant group differences in 
 answers to particular items in the questionnaires, and to bring 
 these differences together into a total scoring method by which 
 their cumulative differentiating power will be a maximum. The 
 steps in the solution of this problem are as follows: 
 
 1. For each of the two groups to be compared, make tables 
 showing the frequency with which each symbol in the question- 
 naires is checked. 
 
 2. Select for statistical treatment those symbols which seem 
 to show significant differences in the proportions of the two groups 
 checking them. 
 
 3. In the case selected, determine the difference in proportion. 
 
 4. Determine the standard error of the difference in proportion 
 by the formula given by Yule (formula 6, p. 269). The size of 
 this standard error is an index of the probability of the difference 
 in proportions arising through chance due to the fact that we have 
 merely sampled the total number of persons engaged in the two 
 types of occupation; in other words, it is an index of the univer- 
 sality of our findings. 
 
 5. List the cases in which the difference in proportions is at 
 least twice as great as its standard error. These items should all re- 
 ceive equal weight in the total score. Yule claims that if the differ- 
 ence in proportion does not exceed three times the error of the 
 difference, it may be obliterated by an error of simple sampling on 
 taking fresh samples in the same way from the same material. But 
 for practical purposes the probability of the difference being ob- 
 literated is very little less if the error is twice the difference in pro- 
 portion, and this enables us to use a great many more items. 
 
 6. Determine which items will enter into the total score in a pos- 
 itive way and which in a negative way. This will depend on which 
 group it is desired to favor in the total score. Thus, if we wish to 
 “favor” the sales group, we assign positive values to symbols 
 which are checked by a larger proportion of the salesmen than the 
 students of mechanics, and negative values to symbols where the 
 
86 MAX FREYD 
 
 opposite condition exists. The resultant total scores will reveal 
 “higher” scores on the part of the salesmen, but this does not in- 
 dicate a superiority, since the scores are merely relative, the main 
 object being to differentiate the groups. It does not matter which 
 group we select to assign the higher scores, but having selected one, 
 plus and minus signs should be consistent throughout. All symbols 
 should be given equal value, that is -+-1 or —1; since the criterion 
 for inclusion is high (see 5 above) no weighting of symbols is 
 necessary. 
 
 7. The total score for any person will be the algebraical sum of 
 the positive and negative values attached to the significant items 
 (symbols) which he checks. 
 
 The application of this method to test scores is fairly obvious. 
 In the case of the two primary groups, critical scores were found 
 by inspection which divided the distributions for the two groups 
 into sections in which there were significantly different propor- 
 tions of the members of the groups. For example, critical scores 
 were set at 2 and 11 for the number of contrast responses in the 
 Kent-Rosanoff Test. Seventeen per cent of the sales group and 
 53 per cent of the industries group obtained scores of 2 or less. 
 Forty per cent of the sales group and 7 per cent of the industries 
 group obtained scores of 11 or more. Both of these differences 
 are significant and consequently enter into the total scores. The sig- 
 nificances of the differences in proportion may be determined in the 
 same manner and by the same formula as differences in propor- 
 tions checking the symbols in the questionnaires. Thus, scores. of 
 II or more in the number of contrast reponses may be treated the 
 same as checking D after an item in a questionnaire; scores of 2 
 or less may be treated as checking L, and scores of 3 to Io inclu- 
 sive as checking ?. The significant differences are in the propor- 
 tions checking L and D (scoring above 11 or below 2). Setting 
 critical scores corresponds to the second step in the above outline. 
 
 A possible objection to this scoring method is that it leaves out 
 of consideration in most cases the majority of the groups;that is, 
 the majority of both groups may obtain scores of 3 to 10 inclusive 
 in the number of contrast responses. This objection is not valid for 
 the following reasons: 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 87 
 
 I have not set critical scores nor included in the total score items 
 of questionnaires, no matter how significant the differences in pro- 
 portion may have been, unless a fair proportion of the total num- 
 ber of individuals were taken into consideration. There would be 
 no point in setting a critical score if only two members of one 
 group and none of the other group were beyond it, or accepting 
 an item (in a questionnaire) which was checked by 2 per cent of 
 one group and 8 per cent of the other. It would require a great 
 number of such partial scores to obtain a total score which would 
 differentiate. f 
 
 Those whose scores do not fall into any of the compartments 
 formed by the critical scores and who consequently do not receive 
 either a plus or a minus score for that measure, are not totally 
 ignored in the composite score for that reason. Their score for that 
 test is zero. This makes the dividing line between the two groups 
 in the total score tend to be around zero, a convenient place for it. 
 It also makes scoring a simpler matter. The method could be made 
 more obvious by assigning to all those who score below 2 in the 
 number of contrast responses a score of 1 toward their total score, 
 assigning 2 to those who score from 3 to 10, and assigning 3 to 
 those who score 11 or above. 
 
 A third answer to this objection is that even if a certain propor- 
 tion of the groups were ignored in the total score because of their 
 scores in any individual measurements, we should still have a 
 logical basis for differentiation with our total score. We may 
 assume that the factors which operate to cause one to go into an 
 occupation are very diverse, and that any two or three are suf- 
 ficient to place him in that occupation. Then as long as an indi- 
 vidual falls into compartments showing significant differences in 
 two or more tests or answers to questionnaires, he shows evidence 
 of having been influenced by those two factors to place him in the 
 occupation in which he finds himself. Only the extreme individuals 
 are influenced by all of the factors. 
 
 RESULTS 
 
 The individual items from the two questionnaires and the tests 
 which entered into the final score for differentiating the industries 
 
88 MAX FREYD 
 
 seniors from the thirty successful life insurance salesmen, do not 
 in every case differentiate the secondary groups from each other, 
 but the results on the secondary groups were not tabulated until 
 the final scoring method was drawn up. 
 
 In determining differences in proportions of symbols checked 
 in Questionnaire 4, L! and L were combined and D and D! were 
 combined, and not tabulated separately. The reason for doing this 
 was that another form of the blanks in which only one kind of L 
 and D occurred was given to most of the secondary groups, and if 
 we wished to score them according to the final scoring method we 
 should not have been able to do so unless we made this change. 
 
 The items which entered into the total score for differentiating 
 the sales from the mechanical groups are given in Table 4 on the 
 following page. The scoring method was decided upon by com- 
 paring the two primary groups, and the results of this comparison 
 are in Table 4. 
 
 In obtaining the total scores the sales group was favored; that 
 is, where a greater proportion of salesmen than industries students 
 checked a given symbol, checking that symbol counted +-1 toward 
 the total score, and where the opposite relationship held, checking 
 the symbol counted —1 toward the total score. | 
 
 One member of the primary sales group failed to fill out several 
 of the forms, and his figures are omitted entirely in the presenta- 
 tion of results. Three members of this group also failed to com- 
 plete parts of Test IX. 
 
 Fig. 9 shows the results of the composite scoring method on the ~ 
 two primary groups. In this figure are shown the total composite 
 scores, composite scores on tests alone, and on Questionnaires 2 
 and 4, separately and combined. Each component of the total com- 
 posite score is very effective in differentiating the two groups." 
 The total score is, however, more effective than any of the compo- 
 
 1 The only check we have on the reliability of these scores is the following: 
 Eleven men in the Division of Cooperative Research at Carnegie Institute of 
 Technology filled out Questionnaire 4 twice at an interval of one month. Their 
 scores were obtained on those items which enter into the total score and occur 
 in that questionnaire. Of the eleven, five obtained the same total score on this 
 questionnaire, and four others gained or lost one point. ; 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 89 
 
 TABLE IV 
 
 Items which Entered into the Total Scoring Method for Differentiating Sales- 
 men from the Mechanically Inclined. These are the items toward which the 
 two primary groups (Sales. and Industries) differ in their attitudes. Check- 
 ing the symbol after the given item or making the critical score was ac- 
 corded the score in the last column. 
 
 TEST OR QUEST, ITEM CRIT. SCORE PER CENT MAKING WEIGHT 
 
 OR SYMBOL CRITICAL SCORES 
 OR CHECKING SYM. 
 Sales Industries 
 Test IX 
 Part 5 2.10 or more 52 Ty; +1 
 Part. 5 1.89 or less 26 60 —I 
 Part ti 16 or more 30 73 —I 
 Partert 14 or less 44 13 +1 
 Part 9, ratio 1.20 or more 17 43 —I 
 race, 2 6 or more 60 20 +1 
 Patter 3. +~=orless 27 70 —I 
 
 Kent-Rosanoff Test 
 
 Contrast Responses II ormore 40 07 +1 
 Contrast Responses 2 or less 17 53 — 
 Questionnaire 2 
 Actor. L 46 20 +1 
 Astronomer L 13 36 —I 
 Auto repairman it 10 46 —I 
 Draftsman L 10 43 —I 
 Factory manager 1 40 90 —I 
 Factory manager D 53 03 +1 
 Locomotive engineer L 20 43 —I 
 Machinist ib 10 50 —I 
 Machinist D 63 26 +1 
 Magazine writer ib 53 23 +1 
 Magazine writer D 20 50 —I 
 Preacher igs 23 03 +1 
 Ship officer LZ 10 60 —I 
 Ship officer D 56 20 +1 
 Shop foreman L 03 56 —I 
 Shop foreman D 73 16 +1 
 Specialty salesman Li 56 20 +1 
 Toolmaker L 03 50 —I 
 Toolmaker D 70 26 +1 
 Watchmaker 15, 03 23 —I 
 Questionnaire 4 
 Fat men L! or L 53 17 +1 
 Fat men DorD! 00 30 —I 
 Very polite people L!orL 47 90 —t1 
 Very polite people ? 33 00 +1 
 Conventions D or D! 07 27 —I 
 “Life” Lior L 97 ofe) +1 
 “Lite” re 00 23 —I 
 “New Republic” Bilgorals 33 ao, —I 
 “New Republic” ? 63 23 +1 
 Interviews Lior: "7 33 +41 
 
 Interviews Robot) | 22 67 —I 
 
90 MAX FREYD 
 
 ral ate Questionnaire 4 
 wl pl gt de Choice of Occup. 
 eg Eg BE eee Quest. 4 and 
 
 Choice of Ocoup. 
 
 feces SAB 
 
 Frequency 
 of 
 
 Total Score 
 re yp 
 . 70 7 ° $ Osa: 
 
 & 
 2 
 
 Fic. 9 
 
 Distribution of Total Scores and Scores on Separate Blanks for Thirty 
 Industries Seniors (unshaded area) and Twenty-nine Successful 
 Life Insurance Salesmen (shaded area). g 
 
 nents, the overlapping being only one. Another interesting point is 
 that the most favorable critical score for dividing the groups is 
 near the zero point in each case. 
 
 The total composite score was obtained for the secondary 
 groups. The interquartile ranges are presented in Fig. 10. The 
 mechanical groups include the Industries seniors, ’21 and ’22, Car- 
 negie Institute of Technology; Commercial Engineering and Me- 
 chanical Engineering seniors, ’21, Carnegie Institute of Tech- 
 nology; and Mechanical Engineering seniors, ’22, Case School of 
 
gI 
 
 THE SOCIALLY AND THE MECHANICALLY INCLINED 
 
 -@ 
 
 S910IG [e}JO], JO sasuey I[4enbsoj}uy] 
 
 OI ‘D1 
 Z 0 z t 9 Cres Ob" cla tiie: 
 ‘ . ® . ‘ 4 ‘ 8 
 ueuseTeg *sul SITT G2 
 (seseo 99z) ususeTes TIV 
 RERUN ERLE (sesvo 2g) stOTUeS °S*H °38OM 
 CTA ATES TD siczues seyiysnpul 
 A ST EN TED 
 
 (sesso PZT) sdncip TeopuByoo 
 
92 MAX FREYD 
 
 Applied Science. The sales groups include the entire membership 
 of the Life Insurance Schools of Spring, 1920-21, and Fall and 
 Winter, 1921-22; and the salesmen of the Harrison Real Estate 
 Corporation, the Burroughs Adding Machine Company, and the 
 Cleveland Discount Company. For further information about 
 these groups, see Chapter I. The Westinghouse seniors were a 
 composite group, some being salesmen and some mechanics. Most 
 of them neglected to state their occupation. Of the three making 
 the lowest scores, two were draftsmen (the only two in the group), 
 and of the three making the highest positive scores, one was a 
 salesman and the other a sales correspondent (the only ones in the 
 group). 
 
 The outstanding feature of this figure (Fig. 10) is the fact that 
 there is no overlapping of the interquartile ranges of the sales and 
 mechanical groups. The overlapping of the actual distributions is 
 less than one-fourth, which may include all those individuals who 
 are vocationally misplaced. 
 
 Another feature is the fact that the best critical score remains 
 around the zero point. 
 
 There is a tendency toward bimodality in the distribution curve 
 of total scores for all groups (Fig. 11c) although the tendency 
 is not marked. The modes correspond very roughly to the medians 
 of the two main groups. Fig. 11a and Fig.11b show the complete 
 distribution curves for the mechanical and sales groups. It is clear 
 from the curve for the mechanical group that if there were more 
 individuals in the group its mode would be around —3. If the 
 numbers in the two groups were equal the bimodality of the curve 
 for all cases (Fig. 11c) would be accentuated. 
 
 SUMMARY 
 
 Two important points are presented in this chapter. 
 
 (1) A method is explained for arriving at a scoring method 
 for differentiating groups. : 
 
 (2) A scoring method for differentiating salesmen from men 
 who are mechanically inclined, is presented. 
 
 The main purpose of the chapter is to show how the results ob- 
 tained in this study may be applied to practical problems in voca- 
 
93 
 
 THE SOCIALLY AND THE MECHANICALLY INCLINED 
 
 Las ot Ot 8 9 4 FA Q Z v 9 8 
 
 Olie.claeeyt 
 
 
 
 ot 
 
 
 
 *(saseo 
 Ivy) Us [[V 1OF saz09S 
 [eIOL JO aainy voNysiqg 02 
 JII “OI 
 og 
 OF 
 + = 
 Videe.dts Ola 8 9 v Z ) 2 id 9 8 ol 
 ° 
 oT 
 *(sased Sgz) 
 UIWISI[VS ][[Y OF sat09g 02 
 [®1O], JO VAIN UOTNQIySIG 
 ql “Oly 
 og 
 
 *(sosed PzZ1) 
 sdnoist) “Yoda IOF S9tOIS 
 [eJO], FO VAIND UOTINAIIYsIC 
 
 DII “Oly 
 
 
 
 61095 
 
 sesso Jo Iequni 
 
04 MAX FREYD 
 
 tional guidance. By the use of this scoring method, we can discover 
 the inclinations of any individual as regards mechanical or social 
 types of occupation, and advise him accordingly. 
 
 From the standpoint of the mere study of the personality of any 
 individual the results of this scoring method may be of interest. 
 The scoring method not only places the individual vocationally, 
 but places him with regard to constellations of personality traits 
 and interests. 
 
CHAr Ti RsVil 
 
 SUMMARY AND CONCLUSIONS 
 
 There is nothing in the results submitted to prove the existence 
 of types of people; rather can we conclude that extremes exist with 
 regard to any trait or combination of traits. If we assume the 
 normal probability distribution for any trait or combination of 
 traits, the persons lying beyond sigma, to be dogmatic, may be 
 said to belong to extremes. We find some evidence for a bimodal 
 distribution, which Thorndike sets up as a criterion of types, in 
 the distribution curves of total scores. In general, though, the evi- 
 dence is against the existence of types. The great majority of peo- 
 ple have no pronounced tendencies which warrant their placement 
 in extremes, but only show leanings toward one or another of ex- 
 tremes. Leanings toward extremes are of psychological value, and 
 in a sense this study has shown how they may be approached and 
 utilized. The importance has been shown of a distribution of ten- 
 dencies with mechanical inclinations at one extreme shading off 
 into social inclinations at the other extreme, and to some extent the 
 personality traits correlated with these tendencies have been deter- 
 mined. 
 
 The measures of personality used in this study are open to criti- 
 cism on several points. The test results may be interpreted as 
 measures of any one or more of a number of traits. The question- 
 naires may have been completed without regard for the truth. 
 The ratings may have been invalidated because of a bias in favor 
 of the persons rated, or because of errors of self-estimation. These 
 objections are not fatal, since the interpretations of the function 
 measured by any test will vary within rather narrow limits, and 
 the interpretation placed on the tests cannot therefore be greatly 
 in error; the purpose of the questionnaires was not so obvious as 
 to indicate the most profitable direction of falsification; and if the 
 raters were biased in favor of the subjects of the experiment, this 
 
06 MAX FREYD 
 
 bias would be a factor common to all groups of raters and if neces- 
 sary could be corrected for, but in any case this would not change 
 the standing of any group relative to the other groups. If the 
 ratings are not very accurate for any individual, they are at least 
 sufficiently accurate for group comparisons. The most effective 
 contradictions to the criticisms of the means by which the results 
 were obtained, come from the fact that the tendency is for the re- 
 sults obtained with the three sets of forms to corroborate each 
 other, and from the fact that the resulting description of the per- 
 sonality of the extremes of the two groups is psychologically 
 consistent. 
 
 The following table summarizes the basis for the conclusions 
 drawn as to differences in personality between the two groups. 
 
 TABLE V 
 
 TRAITS TESTS QUESTIONNAIRES RATINGS 
 Salesmen are more 
 Excitable Bratt. 4 
 Self-confident Trait 10 
 Open-hearted Trait 15 
 Quick to make friends 3:40 Trait 14 
 Present-minded Traits? 
 Good-natured 4:fat men, “Life” Trait 2 
 Adaptable Pt. 6, disguises Trait 13 
 Talkative Trait 18 
 Neat in dress ‘sh Draitens 
 Less self-conscious Pt. 9, ratio POT IS SY Trait 7 
 Credulous or suggestible Pt. 11 296, 632 06. 
 
 ; és 5:5, 7) 16 
 Objective or universal Kent.-Ros. com- 
 
 in their verbal ass’ns. monality 
 Careless of details ty 
 Deficient in fine motor Piaziand Piss 
 coordination 
 Less conceited Agreements between 
 checks in Pts. 3 
 and 9 
 
 Persevering Pt. 6, letters 
 
 How these differences in personality arose cannot be answered 
 in detail by examining the answers to the questionnaires. The pos- 
 sible explanation of the factors which operated to place the men 
 in the opposing extremes of occupation and personality are as fol- 
 lows: 
 
 (1) The mechanically inclined men were more often of non- 
 American or non-British extraction. Not being of the same stock 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED 97 
 
 as the majority of the inhabitants of the country, and not having 
 the advantages of training by parents of that stock, they were 
 handicapped in their adaptation to society and tended therefore 
 to enter occupations where mechanical rather than social ability 
 was of prime importance. 
 
 (2) The early family history of the mechanically inclined men 
 points to a less healthful and cheerful environment than that of 
 the salesmen. The indications are that the mechanics students were 
 dissatisfied with the financial status of their families, the charac- 
 teristics of their companions, and the generally cheerless tone of 
 their environment. They tend to claim they were treated worse than 
 their brothers and sisters, and engaged in more fist-fights than the 
 salesmen. 
 
 (3) Maternal influence was exerted predominantly on the me- 
 chanically inclined men and paternal influence on the salesmen. 
 How these influences shaped their personality is an obscure psycho- 
 logical problem. Without going unnecessarily deeply into the pos- 
 sible psychoanalytic explanation, the conclusion seems warranted 
 that the mechanics student was very early rebuffed by his human 
 environment, owing perhaps to a greater mechanical than social 
 intelligence, and subjected himself largely to the influence of the 
 person who naturally took the deepest interest in him,—his mother. 
 The mother’s influence may have caused him to become less vig- 
 orously adapted’ toward society, but a more plausible interpreta- 
 tion of the facts is that the same factors which caused in the me- 
 chanics student an attachment to his mother, gave rise to his per- 
 sonality traits and caused him to enter a mechanical occupation. 
 The influence of the mother may be accepted as an intensifying 
 factor in shaping personality, and as subordinate to the fact that a 
 greater proportion of the men were of foreign stock and in general 
 grew up ina less fortunate environment than the salesmen. There 
 is nothing to indicate that race is responsible for choice of occu- 
 pation or for personality differences such as we have been study- 
 ing. We should probably find socially inclined and mechanically 
 inclined individuals distributed in the same proportions in every 
 race and nationality, but we should also expect to find among a 
 submerged minority of different stock a greater proportion of men 
 
98 MAX FREYD 
 
 of mechanical inclinations, simply because they have not the train- 
 ing, appearance, and other qualities necessary to exert social in- 
 fluence. In America, especially, there is a marked tendency to look 
 with distrust on foreigners or those bearing foreign names. 
 
 (4) There remains, however, a large per cent of men of me- 
 chanical tendencies who are of native stock. The explanation given 
 in the preceding paragraph will not account for the differences be- 
 tween them and the salesmen. The most suitable explanation is to 
 describe them as lacking in social ability and possessing mechanical 
 ability, and the salesmen as possessing social ability and lacking 
 mechanical ability. Thorndike (21) postulates three kinds of intel- 
 ligence: mechanical, social, and conceptual, represented among 
 others by mechanics, salesmen, and engineers, respectively. The 
 mechanics students may dominate in mechanical intelligence and 
 be weak in social intelligence, and the salesmen display the opposite 
 tendency. It is the writer’s belief that conceptual intelligence is not 
 of a distinct sort, but displays itself in both mechanical intelligence 
 and social intelligence. Social intelligence may be conceived as the 
 ability to manipulate human beings so as to accomplish definite 
 ends, in contradistinction to mechanical intelligence, which is the 
 ability to manipulate machines. But conceptual intelligence is de-: 
 manded in both of these processes. Social intelligence and mechan- 
 ical intelligence are the only ones which may be distinguished be- 
 havioristically. Social intelligence is perhaps best illustrated by 
 pointing to the politician, the orator, or the salesman. The ma- 
 chinist or the draftsman illustrate the possession of mechanical 
 intelligence. The reactions of human beings and machines are of 
 not quite the same nature, and require a different technique. On 
 its passive side, social intelligence is shown in social suggestibility, 
 identification with the crowd in associations and beliefs, in suscep- 
 tibility to “being sold,” in apparent credulity. The credulous per- 
 son, or the person who appears to be credulous, is more apt to 
 make a favorable impression on others than the person who is 
 open in his disbeliefs. 
 
 (5) At least one of the traits displayed by the mechanics stu- 
 dents may be compensatory in nature. I refer here to conceit, ap- 
 parently measured by the means explained in Chapter II, and also 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED ~—99 
 
 shown in the statements of the industries students that they won 
 a majority of the fights they engaged in as children. A realization 
 on the part of many of the mechanics students that they lacked 
 certain of the elements of social ability may have wittingly or un- 
 wittingly caused them at least to assume the attitude of possessing 
 those elements. Conceit is a typical compensatory attitude. Incre- 
 dulity may in a sense be compensatory. It is an attitude assumed 
 by those who realize their weaknesses and are consequently on 
 guard that no one takes advantage of them. It is less easy to find 
 examples of compensation among the salesmen, since it is a lack 
 of social adaptation that is usually compensated for. Lack of 
 mechanical ability would be less often a cause of social mal-adapta- 
 tion than lack of sociability, and it is the realization that one is 
 socially poorly adapted that gives rise to compensation. Persever- 
 ance on the part of the salesmen may possibly be a compensatory 
 mechanism, and their lack of conceit may well be considered as 
 over-correction lest they be judged as conceited in the slightest 
 degree. 
 
BIBLIOGRAPHY 
 
 . Anprews, L. G. A Grading System for Picking Men. Sales Management, 
 
 January, 1922, P. 143. 
 
 . Britt, A. A. The Empathic Index and Personality. Med. Rec., 1920, 97, 
 
 131-134. 
 
 . Brown, W., and THompson, G. H. Essentials of Mental Measurement. 
 
 Cambridge, 1921. 
 
 . Downey, J. E. The Will Profile. 2nd Ed. University of Wyoming, Depart-. 
 
 ment of Psychology, Bulletin No. 3, 19109. 
 
 . Fernatp, M. R., Haves, M. H. S., and Dawtey, A. A Study of. Wore 
 
 Delinquents in New York State. New York, 1920. 
 
 . Freyp, Max. The Graphic Rating Scale. J. of Educational Psychology, 
 
 1923, 14, 83-102. 
 
 . Freyp, Max. A Method for the Study of Vocational Interests. J. of Ap- 
 
 plied Psychology, 1922, 6, 243-254. 
 
 . GotpsmitH, D. B. A Study of Five Hundred Personal History Blanks. 
 
 Guardian Life Insurance Co. of America. New York, 1922. 
 
 . Hansen, C. F., and Ream, M. J. The Predictive Value of Short Intel- 
 
 ligence Tests. J. of Applied Psychology, 1921, 5, 184-186. 
 
 . Hart, B., and SpearMAN, C. General Ability: Its Existence and Nature. 
 
 Brit. J. of Psychology, 1912, 5, 51-84. 
 
 . Kent, G. H., and Rosanorr, A. J. A Study of Association in Insanity. 
 
 Amer. J. of Insanity, July and October, 1910, Vol. 67. 
 
 . Leusa, J. H. The Belief in God and Immortality. Boston, 1916. 
 . Moorz, B. V. Some Principles and Practices of Personnel Selection with 
 
 Particular Reference to Graduate Engineers. Psychological Monographs, 
 1921, No. 138. 
 
 . Natronat ACADEMY OF ScIENCES. Memoirs, Vol. 15. Psychological Examin- 
 
 ing in the United States Army. R. M. Yerkes, Ed. Washington, 1921. 
 
 . Pressey, S. L. A Group Scale for Investigating the Emotions. J. of Abn. 
 
 Psychology and Soc. Psychology, 1921, 16, 55-64. 
 
 . Ream, M. J. Group Will-Temperament Tests. J. of Educational Psycholo- 
 
 gy; 1922, I3, 7-16. 
 
 . Ream, M. J. The Prediction of Successful Salesmanship. Thesis, Division 
 
 of Applied Psychology, Carnegie Institute of Technology, 1921. 
 
 . SPEARMAN, C. General Intelligence Objectively Determined and Measured. 
 
 Amer. J. of Psychology, 1904, 15, 201-293. 
 
 . THORNDIKE, E. L. The Correlation Between Interests-and Abilities in Col- 
 
 lege Courses. Psychological Rev., 1921, 28, 374-376. 
 
 . THORNDIKE, E. L. Educational Psychology. 3 Vols. New York, 1914. 
 
 Vol. 3. Mental Work and Fatigue and Individual Differences and Their 
 Causes. 
 
 
 
THE SOCIALLY AND THE MECHANICALLY INCLINED _to1 
 
 21. THoRNDIKE, FE. L. Measuring Human Intelligence. Harper’s Magazine, 
 1920, 140, 227-235. 
 
 22. Wetts, F. L. A Preliminary Note on the Categories of Association Reac- 
 tion. Psychological Review, 1911, 18, 229-223. 
 
 23. WELLS, F. LL. Some Properties of the Free Association Time. Psychological 
 Rev., 1911, 18, 1-23. 
 
 24. WoopwortH, R. S., and Wetts, F. L. Association Tests. Psychological 
 Monographs, 1911, No. 57. 
 
 25. Yue, G. U. Introduction to the Theory of Statistics. London, 1917. 
 

 

 
 
 
 
 7 ’ ' 
 i In 24 : 
 7 et g - i f Wi eA : 
 RE . Oth. i en ee ray, | ' may 
 Aa Cah Ny eon) ade By of aah ’ ris  & 4 : 
 4 ‘ * ; : ‘ ¢ pet 
 Se ea er a ee Eke : 
 e Mire oad yrs; mei” ws + oe ie : ara 
 te ‘ i roe ee P . ; 9 i | hl 
 La a pa ru : ; { ra i 4 
 2 he On ee nba fbr) 
 = oe ys 
 ae et ‘he has : ] } 
 7 ‘ ¥ 4 ° fa * . a 4 
 . ' « : 7 
 . + . of « 
 y 4b . ‘ 
 ‘ " i tos ‘ 
 - i - ’ - 
 roy "y > 
 Say a « 
 cg ‘ 
 Bt nea 4 2d ‘ 
 : ) tt 
 i. ' i > 
 - ie vi 
 1 
 i] 
 % wey ; , 
 LU / } ps - 
 = \! 
 s = d ~ 
 ‘ ‘ 
 “4 : > . ; 
 ie af 
 ‘ 
 ie i 
 ' 
 ' ¢ 
 : 
 * 
 ery i 
 ‘ * 
 ‘ * 
 ry ; ‘ 
 ; > ‘ ri 
 on 
 ‘ad “4 
 me 
 s al i 
 | aye ik, My . © 2 cAW, oes ¢ } 
 dubia ® ; : \ ¢ 
 © TWA rer ger ! 
 i ‘. fy iF ag ‘ 
 ~*~ 
 1 
 * 
 ‘. ’ 
 f 
 @ 1’ a 
 . r; 
 ¢ us a y 
 % ibis 
 * o> =. ie die 
 sy * oy 
 dehy heed 
 h x ) : 
 Sees 4 LT 2 
 
 
 
BF21 .P96 v.33 
 The effect of manual guidance upon maze 
 
 Theological inary—Speer Librar' 
 
 iN A 
 
 1 1012 00008 5433