HL1VEIH CINV XLJJVS WVNOILVdUOOO 80:! .LNSWEIADVNVIN HQIA ’VHaa ~ ’ A .-. DEPOSIT .4 . {-WV 2719.,1) /BEHA VIOR MANAGEMENT /FOR ' ' occUPA TIONAI. 1 SA FE TY AND HEAL TH U. 8. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service ‘ (fl ‘ Center for Disease Control /\ LNational Institute for Occupational Safety and Health Division of Training and Manpower Development—J i. . Ix ..; 7...... 1 1.». 3.5:»... u. . .... .. . . .. u. .... . .w .. .. a .. V .1 1 {“9 . . .5, ‘ . a . .3 «WM» 3“..qu 3. % 3.. “W13. «3 a? ... W 3 z ., Mn 3,. 4%. .w 3 .3 .. “I a»... . u“ I). i k :ll‘l .r14\(u.3&.rn.\k.rxt\ . .r C. .IIiIiIl \zph... . y 3].!!! .l.’. wtfvl 3 I; . . ,. I .Lti I 3 I . . .i.|.,.t..l<...l: LI ,1 \ .r ? a k.; 3 «x. Q?» m... ..I.....1.rx, f . . : 2,, .. . . E . .- .. :....i ! :11} 3 x. . < . 1.- ,i «a. I I . .‘l 7...! r g... , «3.3; I .J .b £31. a .. _ BEHAVIOR MANAGEMENT FOR OCCUPATIONAL SAFETY AND HEALTH Prepared by BillB Eopkins Bradley C. Conard Deborah L. Duellman The University of Kansas for The Division of Training and Manpower Development US. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Center for Disease Control National Institute for Occupational Safety and Health September 1979 xubo 5519 Paul. DISCLAIMER The opinions, findings, and conclusions expressed herein are not necessarily those of the National Institute for Occupational Safety and Health, nor does mention of company names or products constitute endorsement by the National Institute for Occupational Safety and Health. 11 ACKNOWLEDGEMENTS The authors appreciate the patience and skills of Deborah J. Graham and the assistance of Ronda McF all in the demanding typing preparation of this manuscript. Appreciation is also extended to Mr. Rogers and the entire PEP group for their support and encouragement. Gratitude is due to several colleagues and friends whose research and work made this manuscript possible. Special acknowledgement is appropriate for Rodney Conard and Gordon Fitch of the University of Kansas, Richard Dangel of the University of Texas, David Fox of the Lucky Mc Uranium Corporation, Judi Komaki of the Georgia Institute of Technology, Beth Sulzer-Azaroff of the University of Massachusetts, and Kent Anger and Michael Smith of the National Institute for Occupational Safety and Health. III /"7'/’/ =25.) 1:: (M CONTENTS Page Acknowledgements ........................................................ iii Introduction ............................................................. 1 UNIT ONE LESSON 1 ...................................................... BASIC CONCEPTS LESSON 2 .................................................. GOALS AND TARGETS LESSON 3 ..................................... DATA SYSTEMS: AN INTRODUCTION LESSON 4 ...................................... DATA SYSTEMS: IMPLEMENTATION UNIT TWO LESSON 5 ................................. HAZARD CONTROL: WRITTEN FEEDBACK LESSON 6 ......................... EXPOSURE CONTROL: TRAINING AND FOLLOW UP LESSON 7 .................. SAFETY PERFORMANCE CONTROL: GRAPHIC FEEDBACK LESSON 8 .............. SAFETY AND HEALTH LOSS CONTROL: INCENTIVE PROGRAMS INTRODUCTION Behavior management methods have been successfully applied to common work problems such as tardiness, absentee— ism, and production and quality con— trol. This module focuses on the use of these methods, with a particular emphasis on their application to oc— cupational safety and health. Virtually every significant safety or health problem in an industrial or business organization involves human behaviors. There are endless exam— ples of employees' behaviors that im— mediately or potentially contribute to someone's safety or health. Any— one who manages employees needs to learn how to control such behaviors. The behavior of human beings does not just happen; it is influenced and maintained by factors in the environ— ment. Likewise, behaviors can be al— tered by manipulating the environ— ment. Due to the growing concern for occupational safety and health, re— cent emphasis has been placed on needed behavioral changes of employ— ees. To effectively deal with, and alter employees' behaviors, an under— standing of the elements of behavior is needed. Unit I is dedicated to the study of these, and other factors important in behavior management. After a particular behavior has been initially elicited, it will remain a part of the persons repertoire only if its continuation is appropriately influenced. The consequences of a behavior are an important element in determining if that behavior will be maintained or changed. By recogniz— ing and applying rewarding and pun- ishing consequences, needed improve— ments can be effected. Lesson 1 explores the different possible con- sequences a behavior may have, and demonstrates how these consequences can be successfully manipulated and applied toward behavior management. Before any changes can be made, the improvements needed or desired must be identified and clearly defined. To accomplish this, good communica— tion between management and employees is essential. After it has been es— tablished what the desired improve— ments are, it must be determined how to go about making the changes which will yield those improvements. Lesson 2 presents the guidelines for establishing good communications, decifering what problems exist, and deciding how to go about making the desired improvements. To determine if improvements have been made, factual information which reflects any changes that have oc— curred must be acquired. Through the use of a data system, this type of information may be properly obtained and analyzed. Lessons 3 and 4 ex— plain the necessary steps for devel— oping and implementing a data system, including the collection, translation and evaluation processes involved. Unit II is comprised of four lessons which are dedicated to the summari— zation of studies in which a behav— ior management approach was applied to solve safety and health problems in business organizations. Each study is followed by an explaination section, designed to review the steps necessary in setting up safety and health programs similar to those used in the given study. If properly designed, implemented, and evaluated, a behavior management program can be an effective method to achieve the goals of safety and health within a variety of occupa— tional settings. UNIT ONE LESSON ONE: BASIC CONCEPTS CONTENTS Page A SHORT HISTORY OF BEHAVIOR MANAGEMENT ............................. 1—3 CONSEQUENCES .......................................................... 1—4 NA TURAL CONSEQUENCES ............................................ 1—6 ARRANGED CONSEQUENCES ........................................... 1 —7 EXTINCTION ............................................................. 1—8 SCHEDULES OF REINFORCEMENT .......................................... 1—8 SUMMARY COMMENTS .................................................... 1—9 BEHAVIOR MANAGEMENT OBJECTIVES After the completion of Lesson 1, the reader should be able to understand and discuss the following subjects: 1. 2. 10. (DWQCDUI The The The The The three—part sequence for analyzing behavior. importance of the effect of consequences on behavior. effect of reinforcers and punishers. three generalizations about the effectiveness of consequences. use and application of arranged consequences. Natural consequences as an uncontrollable influence on behavior. The use and application of extinction as a consequence. Schedules of reinforcement. The establishment of intermittent reinforcement to replace continuous reinforcement. The implications of reinforcement versus punishment. Terms and concepts the reader should be able to define, describe, and give examples of, l. 2. #0301an after reading this lesson: cue (stimulus) 8. natural reinforcer behavior (response) 9. natural punisher consequence 10. arranged reinforcer reinforcer 11. arranged punisher punisher 12. extinction reinforcement 13. continuous reinforcement punishment 14. intermittent reinforcement L2 BASIC CONCEPTS INTRODUCTION Behavior management enables manage- ment personnel to have productive and mutually satisfying relationships with their employees. To accomplish this, management needs to understand what controls employee behavior, and be able to arrange the controlling factors to effectively influence de— sired work practices. The importance of consequences and schedules of re- inforcement must be understood before the techniques of behavior management can be successfully applied. A SHORT HISTORY OF BEHAVIOR MANAGEMENT B.F. Skinner, a leading figure in psychology, began the work that has gradually evolved into behavior management. According to Skinner, every instance of behavior comes em— bedded in a three—part sequence as shown in Figure 1—1. Skinners' view Cue i Behavior Consequence Figure 1-1. Three-part sequence of the analysis of behavior. is that there will be a cue, (what— ever precedes the behavior in time and prompts it to occur), then the behavior, and subsequently the conse— quence(s) of that behavior. Figure 1—2 demonstrates this three—part se— quence using the following example: Joe is assigned to the task of grind— ing metal parts. The assignment cues behaviors such as his getting the parts which he needs to grind, moving 1-3 to the grinder, etc. Another behav— ior which might be cued, and one in which a manager concerned with safety should be interested, is Joe's put— ting on a pair of goggles. There are probably several consequences of this safe behavior. A safety consequence that should interest Joe and his man— ager is that Joe is not getting hit in the eye by a scrap of metal. How— ever, consequences such as the fact that Joe loses a little work—time while he puts on goggles, or that the goggles cause a bit of discomfort and some sweating at the temples, might also affect the occurrence of this safety behavior. Qua Assignment to Grinder Behavior Putson Goggms Consequence Dwst Get Metal In Eyes An example using the sequence. Figure 1-2. three—part In regard to this example, the cue (grinding assignment) can prompt a variety of behaviors, and several consequences can follow. If manage— ment personnel are to become good at managing behavior, they must learn how to understand the influence of cues and consequences. Although most management practices are concerned with the question "What can be done to one the behaviors a manager wants to occur?" Skinner's BEHAVIOR MANAGEMENT big contribution to behavior manage— ment was his emphasis on the impor— tance of consequences. Management personnel should become good at using cues, but the real payoff for behav- ior management involves understanding the importance of consequences and learning how to use them. That is how behavior management goes beyond other approaches to management. A few more examples might help to clarify the importance of conse— quences. The next example is split into the three—part sequence in Fig— ure 1—3. An employee responds to the 95 HR Monday at 7:30 am. Behaflor We Goes to Work Consequence Employee (Sets mm Figure l-3. An example illustrating the importance of consequences. fact that it is Monday morning by going to work. There are a lot of consequences of this response. For the moment, consider this question, ”Which is a more important determi— nant of the behavior, the fact that it is Monday morning or the fact that the employee gets paid?“ Clearly the consequence is the crucial ingredi— ent. Consider a slightly more subtle exam— ple illustrated in Figure 1—4. A boss gives a worker a routine assign— ment. The worker carries out the as- signment. What might be the conse— quences? It will depend on what the boss is like and the circumstances. Some bosses are not very interested in safety and health. Some bosses may be very appreciative of jobs done 1-4 well. Some bosses remember every as— signment given, check up frequently, and delight in reprimanding a worker at the first sign of any shortcoming. Cue Boss Gives Assignment Behavior Carry Out Assignment Consequence _____3_____ Figure l-4. An example illustrating the importance of consequences. It is clear that the consequences are much more important than the cue. Generally, the way someone responds to a cue will be determined more by the consequences of their behavior than by the cue initiating it. CONSEQUENCES In order to further the understanding of consequences, the following ma— terial will go a bit deeper into the way consequences function. First, note that consequences can function to strengthen or weaken the behaviors that procede them. This concept is illustrated in Figure 1—5. Suppose a supervisor sees an employee wearing goggles while working at a grinding wheel, and walks over to say, ”Thanks for remembering the goggles, I don't want you hurt!” The supervisor's comment, if given sincerely, is like— ly to increase the chances that the worker will remember the goggles the next time. On the other hand, a com— ment like, “It's about time you started wearing goggles, you dummy!” may make it less likely that the gog— gles will be worn the next time. Basic Concepts Reinforcers and Punishers If a consequence strengthens the be— havior that precedes it, it is a reinforcer. The process in which re— inforcers occur is called reinforce— ment. If the consequence weakens the behavior, it is called a punisher. Similarly, the process in which pun— ishers occur is called punishment. (See Figure 1-6). John works on a machine that process— es noxious chemicals. He puts on a respirator provided by the manage— ment. A supervisor walks by and no- tices that John is wearing his respi— rator. The supervisor approaches him and comments, "Good thinking, John, I see you are wearing your respira— tor.” The reinforcement is the act of the supervisor encouraging John to 4/, —e Trigger Behavior Figure 1—5. Trigger —_b —e first The effects of consequences on behavior. use the respirator. The reinforcer is the compliment given to John and should encourage him to increase or maintain his behavior of wearing the respirator. Anything an employee will work to obtain, anything he/she likes, enjoys or appreciates is likely to be effective as a rein— forcer for that worker's behaviors. Examples of things that commonly function as reinforcers are illus— trated in Figure 1-7. and include: recognition, pay, bonuses, time off, paid vacation, etc. A manager should make note of suggested reinforcers and try to add to his list. Rein— foreers are stock tools in the behav— ior manager's trade. Larry tries to take a shortcut to lunch by walking over a conveyor. He loses his footing, falls and ends ,. \ \ Strengthen \ Consequence l Weaken / / \\ / x, ,1. Behavior Consequence (Reinforce!) (Process is Positive Reinforcement) Trigger —__——£;> Behavior _.___{E> Consequence fl (Punisher) (Process is Punishment) ‘~ // fi/ Figure 1—6. The positive effect of a reinforcer, and the negative effect of a punisher, on behavior. 1-5 BEHAVIOR MANAGEMENT Recognition Pay Bonuses Time off Paid Vacation Extra break time Raises Promotions Gift certificates Prizes Awards Compliments Tickets to sporting events Tickets to movies Free drinks and snacks Vacation time Figure 1—7, Cnmmon reinforcers. Verbal reprimands Fines Loss of pay Demotion Pain Injuries Loss of priveieges Recognition of shortcomings Figure 1—8. Common punishers. up with a painful rear and wrist. The pain following the walking on the conveyor is punishment. Punishers are things that are disagreeable, displeasing or painful. Anything an employee will try to avoid, anything he/she dislikes or detests, is like— ly to be effective as a punisher. Examples of things that will usually function as punishers are illustrated in Figure 1—8 and include: verbal reprimands, fines, recognition of shortcomings, etc. It should be realized that behavior management does not recommend never using punishment. However, behavior management encourages all management personnel to consider using rein— forcement more than they are probably accustomed to doing. Reinforcement emphasizes the behaviors a manager wants to have occur. If a safety supervisor gives an employee a bonus for an unusually good, safety sug- gestion, it is clear to everyone what the supervisor is trying to promote. The more the manager uses reinforce— ment, the more people will want to work with him. There is probably nothing better a manager can do to promote cooperation with whatever he is trying to accomplish than to care— fully make sure that people are 1-6 reinforced for cooperation with him. Conversly, punishment does not teach what the manager wants employees to do — just what he does not want them to do. The more a manager uses pun— ishment, the more people will try to avoid working with him. A lot of people will react emotionally to being punished. These emotional re— actions can vary from simple anger, which is rarely productive, to deny- ing responsibility, blaming other people for problems, or retaliations. Of course, a manager will probably have to use punishment at times, but will have troubles if he relies on punishment more often than reinforce—~ ment. Natural Consequences Some consequences occur naturally. Natural reinforcers are those posi— tive consequences that just naturally and automatically occur following a behavior, and usually strengthen or keep strong that behavior. A manager does not have to do anything to make them occur. For example; Dave is hot and thirsty. He walks over to the water fountain and takes a drink. The cool and refreshing drink of water is a natural reinforcer. It Basic Concepts should encourage Dave to walk over to the water fountain whenever he gets hot and thirsty. Natural reinforcers can influence behavior but cannot be controlled. Managers or supervisors can learn to identify and understand many of the natural reinforcers but cannot con— trol their occurrence. Consider the example of an employee running from his work station to the time clock at the end of the work day. One natural reinforcer for running is the fact that the employee does not have to stand in line at the clock. Another may be getting out of the parking lot ahead of a traffic jam. Suppose an employee discovers a small fire. One reaction might be to grab a nearby extinquisher, put out the fire, and thereby remove the threat. The removal of the threat is a natural reinforcer. However, a different employee might react by leaving the plant through the near- est exit. Again, the action stops the immediate personal threat and functions as a natural reinforcer. It should begin to become clear that natural consequences can work for or against what a manager is trying to accomplish. Natural consequences can also be punishers. Natural punishers are those negative consequences that oc— cur automatically and weaken or keep weak the behavior they follow; Jack is careless while working on a power saw and cuts off his little finger; Joe runs carelessly down steep stairs and falls. The natural punishers are the pain, embarrassment and inconve— nience resulting from cutting off a finger and falling. They will prob— ably discourage and weaken the care— less behavior in each situation. There are natural punishers for all dangerous work behaviors. Management has tended to assume that natural reinforcers and punishers should be enough to deter dangerous work behaviors and cause people to behave in safe and healthful ways. Obviously this is not the case. Con— sider for a moment why the natural consequences are not enough. A hur— rying worker may run down a stairway twice a day for five years without 1-7 using the handrail. Each time she does this there are natural conse— quences, mostly reinforcers, for doing it; if one is young enough, it is more fun to run than to walk care— fully — having everyone see that they are still young enough to run is pro— bably a big reinforcer; the boss oc— casionally appreciates the hustle and enthusiasm; she gets where she is going faster. On the 2,243rd time down the stairs, she trips. Just look at all of the natural reinforce— ment that has been stacked against one punisher. The reinforcers are there every time, every day. The punishment is very chancy and infre— quent. Consider the natural consequences for wearing a respirator for protection from some toxic substance that either has very long—term, cumulative ef— fects or probabilistic effects. The immediate consequences for wearing the respirator are all negative; it is hot and uncomfortable; it rem stricts conversation; it has to be kept track of and cleaned. The pay— offs for wearing it are years away or they do not occur with enough cern tainty to be effective. If a worker became suddenly and painfully ill each time she worked around the toxic substance, wearing the respira- tor would probably become automatic. Three generalizations follow from these observations: 1) immediate consequences are generally more ef- g fective in controlling the behaviors they follow than are delayed conse— quences. Usually, the more delayed the consequence the less effective it is. 2) Consequences that reliably follow a behavior generally control that behavior more effectively than more sporadic consequences. (There is an important exception to this rule that we will discuss presently.) 3) It is very possible for relatively small consequences that occur immedi- ately and reliably after a behavior to exert greater control over the be— havior than greater but more delayed and more probabilistic consequences. Arranged Consequences Arranged consequences are those con— sequences that some person manipulates BEHAVIOR MANAGEMENT to control behaviors that are impor— tant to them. All organizations ar— range for employees to receive money for working, or at least for showing up at work. There is not enough nat— ural reinforcement for most kinds of work that much would get done without arranged consequences. Either punishers or reinforcers can be involved in arranged consequences. The deliberate scolding given to an employee who is seen on the job with— out safety-toed shoes is an example of an arranged punisher. Firings, time off without pay for a rule vio— lation, and demotions are other ex— amples of arranged punishers. When natural consequences are not enough to promote the needed safe and healthful behaviors, a manager can appreciate the possibilities for ar— ranging consequences, and should be able to see the implications of be- havior management. The major part of a manager's job in using behavior management will be to understand nat— ural consequences, then arrange the consequences needed to promote safe and healthful behaviors. A manager will have to be good enough at ar— ranging consequences to offset the effects of natural consequences that cannot be controlled and might pro— mote dangerous behaviors. EXTINCTION Another concept that will be impor— tant to a manager's work is extinc— tion. Suppose that Ben performs an act which amounts to his occasionally dancing a jig on the conveyor belt. The manager notices that a natural consequence of this act is the joking applause that Ben receives from the people who work around him. A super— visor might simply talk to the other workers and ask them to quit attend— ing to Ben's antics. If the workers agree with the idea, and if there are no other natural consequences reward— ing Ben's antics, he would not be ex— pected to continue the undesired be— havior. By removing the natural reinforcer that was maintaining the behavior, one can expect it to get weaker. This process is called ex— tinction. Extinction is the weaken- ing of a behavior that results from 1-8 the elimination of the reinforcement that had been maintaining it. If an employee, for some reason, no longer received his pay check as a maintain— ing reinforcer, his work behaviors would probably quickly extinguish. More subtly, if a boss who has gener- ally been interested in a manager's work, becomes engrossed in other problems and shows very little con— cern for what the manager is doing, the manager's act (behavior) of con— tacting the boss to tell him about what is being done is likely to ex— tinquish. SCHEDULING REINFORCEMENT Remember that there was an exception to the rule that reliable conse— quences are better able to control behavior than sporadic consequences? That exception will now be explored and restated. Some consequences re— liably follow certain behaviors. This is called continuous reinforce— ment. This does not mean that rein— forcement occurs continuously regard— less of whether or not the behavior occurs, but rather that the rein— forcement always follows the behav— ior. Having someone laugh at an employee and applaud every time he clowns around on the job would be an example of continuous reinforcement. As noted before, consequences are sometimes sporadic. Whenever someone tries to call home, they sometimes get a busy signal and sometimes no answer. Only part of the time do they reach whoever they are trying to call. This is an example of inter- mittent reinforcement. The rein— forcer sometimes, but not always, follows the behavior. Intermittent reinforcement can some— times be used as effectively as con- tinuous reinforcement to maintain a desired behavior. A manager must de- cide if he needs to give continuous or intermittent reinforcement to strengthen and maintain a new behav— ior. It is a good practice for a manager to try to reinforce desired behaviors as often as possible when they are new and weak. However, it is unreasonable to expect a manager to continuously provide reinforcement for a particular behavior. For Basic Concepts example a manager of a fiberglass plant has just implemented a program to encourage spray booth workers to direct the spray toward the ventila- tion system. During the first sever— al hours of the program, the manager and supervisors continuously rein— force this desired behavior. After it has become evident that the work— ers practice this behavior consis— tently, the frequency of the given reinforcement may be gradually de- creased until an adequate schedule of intermittent reinforcement is reached. It is important that an intermittant reinforcement schedule is established before workers come to expect continuous reinforcement. SUMMARY COMMENTS This lesson has emphasized that the 1-9 behavior of employees can by system- atically analyzed by the use of a three—part sequence (cue — behavior — consequence) identified by Skinner. While cues prompt the initial behav— ior performed by workers, the conse— quences of that behavior are vital in strengthening and maintaining its' occurrence. Whether the consequences strengthen or weaken a behavior de- pends on if they have reinforcing or punishing qualities. Though manage— ment can not control natural conse— quences, they can apply arranged re— inforcers and punishers to strengthen or weaken a particular behavior. It is advantageous to give reinforcement on a continuous basis until the de— sired behavior has become well estab— lished, at which time an intermittent reinforcement schedule should be applied. Basic Concepts REVIEW QUESTIONS Every problem in almost any business involves human behavior. True or False? Although cues are an important influence for behaviors, B. F. Skinner emphasized that consequences are more likely to control behaviors. True or False? Indicate which element of Skinner's three—part sequence corresponds to the different phases of the following examples: 1 2 3 Cue (or stimulus), Behavior (or response), Consequence a) worker is thirsty 1 cool and refreshing drink 3 walk over to the water fountain 2 b) icy pavement careless driving accident c) paycheck work schedule work all week d) leave the room relief from discomfort chemical vapor causing nausea a) Larry decides to start wearing his steel-teed boots. The supervisor gives him a compliment to show Larry he is pleased. Larry starts wearing his boots more often. Is the compliment a punisher or a reinforcer? b) At the end of the work day, Mary runs down two flights of stairs to the time clock. The supervisor verbally reprimands her for doing this. Mary just laughs and continues to run down the stairs at the end of work days. Is the reprimand an effective punisher? Fill in the blank with one of these two available answers: weakens, strengthens. a) Something is a punisher if and only if it the behavior it follows. b) Something is a reinforcer if and only if it the behavior it follows. Identify the natural consequences that might result from the following: An auto assembly line is moving fairly rapidly. A person who installs bumpers sees a loose nut on a bumper of a vehicle that is about to move away from his station. He steps in front of the vehicle and tightens the nut. 10. ll. 12. 13. 14. 15. BEHAVKfliMANAGEMENT Identify the arranged consequences that might result from the following: Contrary to company policy, an employee goes home without showering or changing out of her work clothes. Harry puts on his respirator but gets hot and uncomfortable so he takes it off—and experiences relief. Identify the natural punisher and reinforcer, and the behavior that each will probably affect. Natural punishers have been, and always will be, sufficient to prevent accidents. True or False? The most productive and effective way to present a punisher after a behavior has occurred is immediately (or as soon as possible). True or False? Immediate and reliable consequences are generally much stronger and effective than sporadic and delayed consequences. True or False? Arranged reinforcers are positive consequences planned by management to follow a behavior but are not controllable. True or False? An arranged punisher for dealing with the behavior of not wearing protective goggles at the grinder is: a) the worker is hit in the eye with a scrap of metal. b) the worker is fined $10.00. c) the worker completes his/her job faster. d) none of the above. Arranged consequences can help combat the inefficiencies of natural consequences for controlling safety behavior. True or False? Jane,much to the management's disapproval, put on a fireworks display once a day at her welding table. The workers in the nearby area usually applaud and cheer her efforts. The supervisor talks to the workers who watch her, and convinces them to not pay attention to her when she ”horses around.” The supervisor is trying to use which of the following: a) punishment b) reinforcement c) extinction Basic Concepts Continuous reinforcement is probably better than intermittent when a behavior is: a) strong b) weak c) none of the above Which of the following can be considered disadvantages of using punishment? a) it teaches only bad behaviors. b) it teaches only good behaviors. c) it produces negative emotional reactions. d) a and c. 10. 11. 12. 13. 14. 15. 16. 17. BEHAVIOR MANAGEMENT ANSWERS TO REVIEW QUESTIONS True True 3 d) 2 1 2 I-‘CDN a) reinforcer b) no a) weakens b) strengthens He trips and falls. He becomes entangled in the conveyor. He is hit by the moving vehicle. He slows down assembly line production. She is reprimanded by her employer. She is denied reinforcement given to other workers. She is fined for violating company policy. The natural punisher is getting hot and uncomfortable and affects the behavior of putting on the respirator. The natural reinforcer is the relief from discomfort and affects the behavior of taking off the respirator. False True True False (b) True (C) (b) (d) Basic Concepts BIBLIOGRAPHY Luthans, Fred, and Kreitner, Robert. Organizational Behavior Modification. Glenview, Illinois. Scott, Foresman and Company, 1975, pp. 33—52. Margolis, Bruce L., and Kroes, William H. (Eds), The Human Side of Accident Prevention. Springfield, Illinois. Charles C. Thomas, 1975. Miller, Lawrence M. Behavior Management —— the New Science of Managing People at Work. New York: John Wiley and Sons, 1978. Skinner, B.F. Science and Human Behavior. New York: The MacMillan Company, 1953. LESSON TWO: GOALS AND TARGETS CONTENTS Page IDENTIFYING NEEDS FOR IMPROVEMENTS .................................. 2—3 DEVELOPING GOALS ...................................................... 2—4 DEVELOPING TARGETS ................................................... 2—4 IDENTIFYING SOURCES OF INFORMATION .................................. 2—5 REINFORCING INFORMATION GIVING ...................................... 2—11 DEALING WITH COMPLEX LISTS OF TARGETS ................................ 2-12 SUMMARY COMMENTS .................................................... 2—15 After the completion of Lesson 2, BEHAVIOR MANAGEMENT OBJECTIVES discuss the following subjects: 1. 2. domed: 10. the reader should be able to understand and Recognizing problems and desires for improvement. Translating problems and desires for improvement into goals. Deriving targets from goals. The difference between goals and targets. The distinction between behaviors and products of behaviors. Identifying sources of information. Obtaining concrete and specific descriptions and information. Reinforcing information giving. Organizing complex lists of targets. Setting terminal and intermediate targets. Terms and concepts the reader should be able to define. describe, and give examples of after reading this lesson: 1. 2. goal target objective behavior product of behavior 6. 10. 22 sources of information concrete, specific description prerequisite target terminal target reinforcement information GOALSAbfl1TARGETS INTRODUCTION Lesson 1 proposed that occupational safety and health can be improved if management can learn how to effec— tively deal with people's behaviors. Lesson 2 describes the first few steps that can assist management per— sonnel in determining the specific behaviors or products of behaviors with which they should be concerned. IDENTIFYING NEEDS FOR IMPROVEMENT People who oversee employees can learn how to recognize problems or Q} sires for improvement and translate them into concrete action objectives or targets. The first steps in be— havior management are very similar to a management tool called management by objectives (Drucker, 1954; Raia, 1974). Both behavior management and management by objectives involve de— veloping specific targets or objec— tives. Also, both encourage measure— ment or data collection to determine if targets have been achieved. How— ever, behavior management goes fur- ther to provide suggestions of how to deal with targets that involve human behaviors (most of them do involve human behaviors), and develop a tech- nology managers can use to achieve their objectives. The importance of being able to translate relatively general problems and desires into concrete objectives can be easily illustrated. Suppose a department foreman tells his manager that workers have poor attitudes to— ward safety, and asks him for help in improving their attitudes. What can the manager do? Does he know pre— cisely what the foreman means by ”attitudes" and what kinds of changes would indicate to him that workers' attitudes have improved? Does the manager or foreman know how to change workers' attitudes? Both the manager and the foreman may have to give negative responses to these ques— tions. 2-3 Now suppose that a manager has become skillful in helping foremen to recog— nize and deal with their problems. A foreman can then tell his manager specific things about the workers' behaviors such as; they do not wear goggles when operating a bench grind— er; they leave solvent cans uncover— ed; they leave parts stacked in the aisles; and they fail to immediately clean up lubricant spills. Further— more, the foreman claims that workers never point out hazardous conditions to him. The manager now has a much better idea what the foreman is talk— ing about and might know some ways to generate changes. The first step necessary to effect an improvement in safety and health is to find out precisely what needs to be done. By implication, when a man— ager is considering making improve- ments in his company's safety and health programs, he is debating mak— ing some changes. How can people who oversee employees decide if changes are needed? Needs for change origi— nate in information. The information involves either known or Suspected problems or some desire to do things better. Information that might re— flect safety and health problems could include: 1) A report which shows that rising accident costs have led to increased insurance premiums. 2) Finding that annual frequency and severity rates are running well above the industry average. 3) A verbal statement of a supervisor about shop areas being cluttered with trash. 4) A safety and health inspection that cites hazardous conditions which need to be corrected. 5) An industrial—hygiene survey that indicates that workers in certain jobs are receiving unreasonably high exposures to styrene. 6) A supervisor's observation that workers who carry out a particular operation are frequently found to BEHAVIOR MANAGEMENT have some form of dermatitis. Organizations sometimes want to make changes not because conditions that would be considered problems exist, but simply because people want condi— tions to be better than they present— ly are; there is a desire for im- provement. Remember that what is considered a problem in one organiza— tion may be viewed as unobjectionable in a different company. One of the most important jobs of management personnel should be to make as many improvements in safety and health as their resources allow. Even if a company does not View a particular situation as a problem, a manager may still be able to capitalize on de- sires for improvement. Following are a few examples of in— formation about conditions that might not be considered problems in a par— ticular organization but for which there may still be desires for im— provement: 1) Though a plant's injury frequency and severity rates are below the in— dustry average, a manager sees sev— eral areas in which improvements can be made. 2) A production department has never had an injury causing lost work-time; nevertheless the department foreman suggests that a number of hazardous conditions exist in all of the work areas. 3) A manager watches a group of work— ers load a boxcar and notices several ways they could behave differently, thereby reducing the chances of hav— ing accidents and injuries. 4) A supervisor acquires the best spray booth the industry has to offer, but sees workers spraying in such a way that they fail to take advantage of the existing air currents. DEVELOPING GOALS The second step in applying behavior management to safety and health is to develop goals. A goal is a general written or verbal statement that re— flects the changes necessary to solve a problem or make acceptable situa— tions and conditions better. Goals that would reflect the changes needed to solve the earlier list of problems are: 1) To reduce accident costs. 2) To lower frequency and severity rates to at, or below, the industry average. 3) To clean up and keep shop areas clean. 4) To meet the criteria for passing the next safety and health inspection 5) To reduce unreasonably high expo— sures to styrene to an acceptable level. 6) To eliminate cases of dermatitis. Goals can also relate to the several changes needed to improve situations that are not considered problems. Goals for the earlier list of these kinds of changes would be: 1) To reduce the frequency and sever— ity rates below their current levels. 2) To indentify and eliminate the hazardous conditions in all depart— ments. 3) To induce the shipping department loaders to behave in specified ways so that they have fewer accidents and injuries. 4) To influence workers to spray so that they are utilizing the air-flow characteristics of the booth. You should be able to learn from the above examples that goals can be as broad as reducing overall measures of safety, or as narrow as changing the way one worker does a small part of her job. What makes a statement a goal is not how broad the statement is, but the fact that it is a general rather than a specific statement of what management is going to try to accomplish. DEVELOPING TARGETS Behavioral targets are specific and concrete behaviors or products of be- haviors that must occur for goals to be achieved. Suppose a safety audit has just indicated that a number of fire hazards exist in a packing de— partment. That is a problem. The department manager accepts the chal— lenge of overcoming the problem and sets a goal to eliminate the hazards. However, to go beyond this general statement of intent he has to be much Goals and Targets more specific. His list of specific targets might include: 1) Have workers smoke only in desig- nated areas. 2) Replace all frayed electrical cords on machinery. 3) Remove all accumulated paper trash. 4) Requisition and place trash con— tainers at six strategic locations. 5) Have the custodial department emp— ty the trash containers every evening. 6) Dispose of all trash on the floor at the end of every shift. Each of these targets is sufficiently specific that the manager can easily determine if it has been done or is regularly being done. Moreover, he now knows precisely what needs to be done and can communicate effectively with the department foreman, the maintenance and custodial depart— ments, and any other people involved. A distinction between two different kinds of targets will be useful. Some targets are behaviors. Others are products of behaviors. Consider the goal of reducing a plant's fre- quency rate of injuries. The plant manager might set one target to have every worker come by his office to check out a new set of safety glass— es. The worker's asking for new glasses is clearly a behavior. A different target might be to reduce the number of eye injuries to zero over the next six months. Having an eye injury, or having no such inju— ries for a period of time, is not a behavior but is a product of behav— ior. Injuries are results of workers behaving in certain ways or failing to behave in other ways. Picking up trash is a behavior. Having trash— free work areas and aisles is a pro— duct of behavior. Cleaning up spills is a behavior. Having no injuries involving slips and falls is a pro— duct of behavior. In summary, the first steps in behav— ior management are collecting infor- mation about problems and desires for improvements, translating these into general goals, and translating goals into specific targets which may be either behaviors or products of be- haviors. The remainder of this lesson will explain how to execute these steps. IDENTIFYING SOURCES OF INFORMATION Management personnel can obtain in— formation from a number of sources. The purpose of the following discus— sion is to provide suggestions and ideas as to where to obtain this in— formation. However, it is important to realize that the sources, and how useful they are, can vary from one organization to another. Management personnel may discover many sources and must decide how much to rely on each one. It is important that none of these sources are overlooked. Management personnel may find it ben- eficial to obtain information from the people who operate the machines and staff the positions to carry outL the basic work done in their organi— zation. These people are going to be in positions that make it likely that they will come into contact with un— safe conditions and hazardous prac— tices. Management should cultivate their trust and interest. If this can be"accomplished, line workers may be one of their best sources of in- formation. In return, management can attempt to make these employees' work conditions less hazardous. Some organizations have shop union stewards or union representatives who assume some responsibility for work— ers' safety and health. Management may be able to rely on these people for information. In certain cases, a worker may be reluctant to give use— ful information directly to a manager Instead he can give it to the manager through the union representative. Managers should try to establish good relationships with union representa— tives, supervisors, and formen so they will not hesitate to give infor- mation, or object to his talking di— rectly with the line workers. Such relationships will be significant be— cause these potential intermediaries may have responsibilities in addition to those concerning safety and health and, therefore, might not favor look: ing for ideas and problems or encour— aging workers to provide the informa— tion a manager needs. BEHAVIOR MANAGEMENT Foremen, supervisors and department and production managers, or their equivalents, will see many examples and conditions which need to be im— proved. Therefore, they can develop an appreciation for ways in which different parts and processes of the organization interact«-information which can be invaluable. For exam- ple, a particular hazardous condition may exist in one department because of maintenance problems, and the forman may need the plant manager‘s help to get that condition corrected. Suppose one department's opening of windows and doors in their area re— duces the effectiveness of ventila— tion in a second department. The manager may be the only person who can explain the importance of having the ventilation work at peak effi— ciency. As we move away from line workers and up the heirarchy in an organization, the information obtained from people often becomes more general. An exec— utive or administrator may respond to information about insurance costs, time or production lost due to inju- ries, or the safety and health re- cords of the organization after com- paring some statistics for similar operations. The personnel office may be a good source of records that will help to determine which people or areas are particularly likely to have problems. Engineering and mainte— nance departments may be among the best sources for obtaining ideas for improving hazardous conditions. Even people who deal with general informa—, tion can, at times, bring fairly spe— cific ideas and problems to the attention of a manger — if he cul— tivates their interest. Management should not assume that the people in their organization have a monopoly on all of the wisdom in the world. If a manager can get informa— tion from people outside his company he will be able to benefit from their experience and observations as well as his own. There are a number of ways this can be done. Managers can visit other operations that are simi— lar to their own, attend professional meetings and conventions devoted to occupational safety and health, and read trade and technical publications 2-6 Insurance companies may be able to provide different perspectives that are helpful. The National Institute of Occupational Safety and Health (NIOSH) has a Division of Surveil— ance. Hazard Evaluation and Field_ Studies which may be able to provide technical expertise. The NIOSH Division of Technical Services also provides publications which may be useful. Staff in the regional NIOSH office may be able to give valuable suggestions. Management should be— come familiar with relevant regula— tions set by the Occupational Safety and Health Administration (OSHA), and any other regulations which apply to their operations. If a manager re— quests assistance from OSHA, they may be able to provide good preventative consultation prior to any accidents or prior to being called in as a re— sult of an employee complaint. A manager can be one of his own best sources of information if he learns to leave his office and regularly tour his plant or operation and watch the processes involved. In addition to observing work conditions, he can determine the work behavior of his employees. One tip - if a manager ordinarily wears a suit and tie, he should replace them with clothes similar to those worn by the employ— ees. This will result in his being less noticable and, therefore, in— crease his chances of seeing things more as they are rather than as em— ployees would want a member of the management team to see them. If pol— icies permit, it can be useful for a manager to spend some time partici— pating in line work duties such as being an assistant to a line worker, stacking parts for a machine operaton occupying a position on an assembly line, or riding with a bulldozer op— erator. These kinds of investigative practices can give him a more accu~ rate appreciation for the problems and possibilities for improvements than would a coat and tie inspection tour. A manager should set aside time to expose himself to various sources of information. A manager may delete any sources he decides not to use, and add other potentially useful sources that occur to him. He should Goflsandngpu Information Schedule Source of Information Line Workers Supervisors Foremen Union Stewards Production Manager General Manager Outside Consultant Trade Publications 52.1525 Cam/radii 5m? New FAQ/mi” Tau/w Figure 2—1. often each should be contacted. also decide how often to expose him— self to each source and include that information on a form that might re— semble the one shown in Figure 2—1. In this example, a manager decides to visit with line workers, supervisors, foremen, union stewards and the pro— duction manager. Once a month, he regularly attends safety committee meetings which include representa— tives of the personnel and mainte— nance departments. The manager then decides to use the other sources as indicated, and will call or visit with the safety and health engineer of a similar plant six times a year. How Often Twice a Week Once a Week Once a Week Twice a Week Once a Week Once Every Two Months Twice a Year Once a Month 04L(L “/A7m£Zi r Omoz. Eve/L? Tw-u— M 0m a Wed: ,An example of a form for listing sources of information and how Managers should make appointments as necessary, and enter their schedule on a personal calendar. Scheduling time in advance makes it much less likely that a manager will overlook a source of information, and increases the chances that he will engage in the activities as he planned. A manager may want to make 3” x 5" cards like the one shown in Figure 2—2. He should keep them handy in his office, and take a supply with him as he contacts sources of infor— mation. As he receives each piece of information, he should fill in the BEHAVIOR MANAGEMENT Source Date Area Information Goal Targets Target Date Prerequisites Figure 2—2. An example of a card to be filled out as information is received. Source fl (12534511110: Date 822 Area Phi-1&1; {LCM ) Information 5243 239 m 0:3ffv' QXLLifi‘239 ? 1 Q flxL .%\ai¥xQ Goal _[ i jf‘lafrt‘ 39 23b: 5 1(2; pg 51m,~mj'_'fp_mv_p£gg (p “jiggl— Targets __ffi1¢“£4 Target Date Aj;[% Prerequ I Si 1:85 2 mm” [XIIMIIMQJMELUF‘L— Figure 2-3. An example of a completed information card. 2-8 Goals and Targets card. Here are a few examples of informa— tion that could be collected; On August 1 the manager visits with Vlad Kurzinski, a supervisor in the pro— duction department. Kurzinski men— tions that a number of his machines have frayed electrical cords, and that he has requested replacements by maintenance but has not received any service for over a month. The man— ager agrees to try to help him. He continues to talk to Kurzinski about the frayed cords until he is sure that a problem exists. The manager determines how many cords are in— volved, and knows what target has to be achieved for the problem to be solved. After leaving Kurzinski, the man— ager's card might be filled in as shown in Figure 2—3. He had remain- ed with Kurzinski until he was sure that he knew enough about the problem to specify the target. In fact, this problem is so simple that the goal and the target are easy to determine and are virtually the same; to re— place all frayed electrical cords. However, it is usually not as easy as it was in this case. Note that this target is a product of a lot of be— haviors rather than a single behavior. It is not always necessary that a manager know all of the behaviors needed to achieve the target. For example, the maintenance people should already know how to replace the electrical cords. However, it is the responsibility of the manager to get them to accomplish this task. The way to get people to do these kinds of things will be discussed in a later lesson. For now, let's as— sume that the manager has developed good cooperation with maintenance em— ployees. He visits them on the way back to his office and they replace the cords before the day is over. Notice that the information cards serve as a way to organize informa~ tion. When a card is completed it should list or imply what actions need to be taken. The presence of the cards will serve as a reminder to eliminate a particular problem or to not pass up an opportunity to make im— provements. Collecting information often can be much more difficult than implied in the earlier example with Supervisor Kurzinski's frayed electrical cords. Let's assume that Kurzinski is not always so effective at explaining what he thinks should be done. With some people, a visit may have been more like the following: Manager: "How are things going in production?” Kurzinski: ”Awful as always. They want twice the production we can get from our machines and I have two peo— ple off sick." Manager: ”Sounds tough. Do ou know of any safety problems we should be aware of?" Kurzinski: ”I know one. This whole place is gonna go up in a shower of sparks one of these days.” Manager: ”I don’t want that. some more about the problem.” Kurzinski: ”We're gonna have somebody electrocuted!” Manager: ”How will they get electro— cuted?” Kurzinski: ”They'll just accidentally touch one of the wires.” Manager: "Which wires?” Kurzinski: ”The electrical supplies to my machines.” Manager: ”Show me some of them.” Kurzinski: ”They are all bad. We're going to have some real troubles one of these days.” Tell me Manager: "Is the wiring bad on that press?” Kurzinski: ”Naw. It's O.K. But I've got a bunch that are bad.” Manager: ”Help me for a few minutes and let's inspect all of them.” (The manager and Kurzinski inspect the wiring to each machine in his area and find frayed insulation on six of the fourteen machines.) The significance of this story is to demonstrate that people are not al— ways good at explaining existing prob— lems. Managers need to become skillful at asking questions that will direct people in giving useful and detailed information. In the preceding example, Kurzinski was ready to tell the manager about all of his real or imagined troubles. The manager cannot help Kurzinski with all problems, so he must steer him to BEHAVIOR MANAGEMENT discuss only those problems related with safety and health. Then,Kurzin— ski quickly describes a disaster he wants to avoid, but is slow to relate the specific conditions or actions that might cause the disaster. Agaim specific questioning leads to the identification of the sources of the problems. Managers must continue to question people until a target be— comes specific enough that there are no doubts about the sources of con— cern. It should be recognized that Kurzin— ski overstated the problem by telling the manager that all of his machines had frayed wiring. Such overstate— ments are not uncommon. Questioning and personal inspection are necessary to obtain a concrete understanding of what needs to be improved. However, management personnel should not as— sume that everyone will overstate problems. Some sources of informa— tion will give only a hint of a prob- lem, and a manager may have to ques— tion skillfully to find out that one actually exists. Workers may imply that a problem is limited or of small scale. They should be questioned, and the described problem should be further inspected to insure that it is not wide-spread. There are several kinds of questions managers can ask in particular cir— cumstances to obtain needed informa— tion from people. Following are cir— cumstances which illustrate the kinds of questions a manager can ask em— ployees: 1) The source tries to talk about everything but safety and health. A question to steer the focus to safety and health considerations is needed. Source: ”If I don't get that machine going again, we are really going to fall behind.” Manager: ”Well, I certainly hope they get it going for you. How are your workers doing in the safety areas?" Such diversionary questions must, of course, be used judiciously. A source could be telling a manager that he needs to be working on the broken machine rather than talking to him. 2-10 Source: ”Well, I'll tell you, I'm going to go back and get that fish the first day I have off. Manager: ”I want to see a picture of him when you catch him. Say, do you know of any hazards that need to be corrected or have you seen any near accidents?” Again the question to get the source to focus on safety and health has to be used with judgment. Management should not be totally uninterested in a source's interests, but should want to get the job done. 2) The source is speaking in genreal— ities and isn't explaining specifi— cally what needs to be changed. Questions about what conditions or behaviors are involved in the general problem are needed. Source: ”Our people just don't mind living with dangerous conditions.” Manager: ”What do they do that makes you say that?” Source: ”We had bad brakes on the lift truck for a week before they told me about it, and they will walk right by poorly stacked skids and never bother to straighten them." A variation on this kind of question— ing might be: Source: "We have a lot of hazardous conditions.” Manager: "Tell me the conditions so I can make noteof them.” 3) A source is speaking in general— ities, or the manager is having trou- ble understanding a specific condi- tion or behavior. The source should be asked to describe, show or act out an example of the problem or the be— havior that he or she believes to be desirable. Source: ”There are several ways we could be more careful in handling the ore.” Manager: ”Describe for me the ways you could handle it better.” 0_R Source: ”A guy could just run the GodsandTaqpm machine a lot more safely." Manager: "Would you take a minute to show me how it should be run? I'm not that familiar with the ma- chine." 95 Source: ”Some of the intersections on the exit grade from the pit are dangerous.” Manager: ”What's wrong with them?" Source: "They just aren't designed right.” Manager: "Let me ride out on the next load with you and you show me the bad ones." ( — ~ and during the ride — — ) Source: ”There, that one's bad.” Manager: ”What would it take to make it safe?’ 4) The source has given concrete in— formation about a problem or possible improvement, but the manager is not sure if he has a complete list of targets. The source should be asked if anything else would be involved in solving the problem or making the im— provement. Source: ”I guess that is the last bad intersection.” Manager: ”Suppose we get those four intersections widened and the piles of dirt removed to make it easier to see; can you think of anything else that should be done to the exit grade?” Becoming skillful at asking these questions will make communications with sources much more precise and will increase the chances that the right things will be done to solve the problems or make the wanted im— provements. REINFORCING INFORMATION GIVING Consider the following descriptions of scenes that occur too often: The safety and health engineer of a plant is talking to a union representative. The representative suggests that the locker room needs a paint job and better lighting. The safety and health engineer goes away without thanking the representative, never mentions the suggestion again, and the paint job and brighter lighting are never done. How often is this particular union representative going to bother to make suggestions? A foreman complains to the manager that scraper operators drive too fast and too close together, and do not down—shift when descending grades. The foreman predicts that there is going to be a severe vehicle acci— dent one day unless something is done about the problem. The manager nods in agreement and walks away. He does not check with the foreman later to see if the problem is better. He does not try to prompt the foreman to come up with solutions. The manager does not even attempt to try to help him find a solution. Why should this foreman bring information about safe— ty and health to the manager's atten— tion again? Management must reinforce the giving of information if they want their people to continue to relay informa— tion. Managers may not always be able to produce what the sources want or solve every legitimate problem brought to their attention, but in some way, they must reinforce employ— ees for providing information. Suppose a production department has just had a small fire and the depart— ment manager is talking to an inter— ested worker. Mary: ”I think we should do something to make sure fires like that don't get out of hand.” Manager: ”What do you think would be good to do?’ Mary: ”I think we should have an au- tomatic sprinkler system.” (Themanager does not know if a sprinkler system is feasible. Still, the source of information has given a suggestion. The manager should try to encourage suggestions.) Manager: ”Maybe that would work. Let me check it out. Thanks for the idea. I'll get back to you.” (He talks to the department foreman and explains that he appreciates Mary's suggestion. He hopes the BEHAVIOR MANAGEMENT foreman will give Mary a kind word. The manager subsequently talks to the engineering department, and they promise to give him a quick answer on the feasibility of the proposal. The manager finds a minute to visit with Mary again.) Manager: ”I've talked to engineering about installing a sprinkler system. They are seeing what they can do about it. I sure appreciate your sug— gestion.” Mary: ”Well, I just don't want to have somebody get hurt, and I don't want us to burn the place up and have to shut down." Manager: ”I appreciate that. again.” Thanks (A day later the engineering depart— ment tells the manager that a sprink— ler system will not help because of several technical considerations. The manager goes back to see Mary again.) Manager: ”I need some more help from you. Engineering says we need some- thing other than a sprinkler system. The heat from the shrink—wrap machine would set off the sprinklers if they are set too low and we could have a bad fire going before the sprinklers would operate if we hang them high. What else can you think of that would help?” Mary: ”We at least ought to get more fire extinquishers and have regular fire drills.” Manager: ”That's a good idea. I'm sure we can pull that off. Where do you think we should locate the extin— quishers?” Some suggestions for reinforcing sources giving information are: 1) Express appreciation and thanks for suggestions. 2) Try to let other people who might talk to the source know about the information and that the sources's giving it is appreciated. 3) Follow up on suggestions. Use the information cards so as to not forget the suggestions and who gave them. 4) Let the source know about progress on carrying out the suggestion. 5) If an idea is not feasible, be honest with the source but do not tell them they are dumb for giving an impractical suggestion. Instead, thank them for giving it, explain factually why their idea will not work, and invite them to try again to find a way to solve the problem or accomplish the improvement. 6) Take every reasonable opportunity to give extra public recognition to the source of a good solution or im— provement. This can take the form of a letter of commendation, recognition at an awards ceremony, or perhaps a bonus if the organization uses them. Minimally, a few kind words can be said to the source while with their boss or fellow workers. The above approach to encouraging suggestions is, in part, based on a successful suggestion program devel- oped and tested by Robert Quilitch (1978). The approach applies regard- less of whether the source of the in— formation is the president of the company, a line worker, or a custo- dian. DEALING WITH COMPLEX LISTS OF TARGETS Because targets can sometimes be dif- ficult to achieve, information cards can be much more complicated than in— dicated in the previous examples. A manager may tell his supervisor to do something about all of the skin rash- es that are being reported in the formulation department. The goal is clear; elimination of dermatitis from the hands and arms of workers in the formulation department. The ultimate target is easy to deduce. The man— ager wants no new cases of dermatitis and elimination of the present cases. However, there are a lot of prereq- uisite targets that will have to be met to reach the terminal target. Let's see how you might go about de— ducing the prerequites. What is necessary to achieve the tar— get? The answer to this question is often clear, but how to accomplish what is necessary is more difficult to determine. In this example, the manager must find the source of the dermatitis and eliminate or control the source. However, the manager may quickly conclude that he has no idea what is causing the dermatitis; and if he does not know the cause he Goals and Targets cannot control or eliminate it. How can the cause be found? There may be several alternatives, but it is likely that many of them are not practical. The manager cannot safely vary the formulations to determine if some change in chemical composi— tion will eliminate the dermatitis. Without knowing the source, he cannot suggest engineering changes to con— trol it. The manager may possibly find the cause by bringing in an ex— pert consultant. Again, two ques— tions come into play; What is necessary to bring in an ex— pert consultant? The manager needs to hire an expert consultant. Know— ing that the regional NIOSH office can give recommendations for this position, he calls them to obtain their suggestions, and subsequently hires the chosen consultant. After consultation with the expert, itis determined that the likely cause of dermatitis is a particular powder that is stored in a bin in the for— mulation room. If this hypothesis proves correct, the source of the problem has been discovered. What is necessary to eliminate or control the culprit powder? The man— ager talks over the possibilities with the consultant and an engineer, and learns that the powder is neces— sary for the products being formu— lated, and that there are no promis— ing substitutes. Now the manager considers the second aspect of the question. How can he control skin exposures to the powder? He watches the workers using the powder in mix— ing a batch and observes the follow— ing: the powder is kept in a bin; the bin lid is often left open even when workers are not using it; they do not wear gloves as they handle the powder (in fact, they often use their bare hands to scoop up a little more powder or to shovel a bit out of the weighing container); they spill small quantities of the powder on the floor around the bin and do not clean it up; a cloud of powder flies into the air each time a container of it is crudely dumped into a batch; the workers do not bother to wash the powder from their hands and arms after 2-13 they handle it, even though a sink and paper towels are convenient. This creates a number of new targets for the manager. They include pro— curing gloves, scoops, mops and buckets and setting up a training program for the workers. The tar— gets of the training program are to keep the bin lid closed when it is not being used, wearing gloves when handling the powder, using scoops to get quantities of powder, wetting the powder before dumping it into a batch, rinsing powder off of the con- tainer and gloves between uses, wet mopping to remove any spills, and washing hands and arms after working with the powder. These prerequisites are too many and too complicated to enter on an infor— mation card, so the manager uses a sheet of paper to record the informa- tion. (See Figure 2-4.) Note that each of the letters and numbers in the outline are followed by targets that are prerequisites for the ter— minal target, eliminating the derma— titis. Further note that procuring gloves, etc. are prerequisites which are important for controlling expo- sure to the powder. Actually, the worksheet shown in Figure 2—4 may not yet be complete. Does the manager know how to set up a successful training program to train the formulation workers good work practice behaviors? Unless his answer is an unequivocal ”yes”, he should continue asking the what is and how can questions. That would generate more targets. Essentially, through answering the questions, ”What is necessary to reach that tar- get?" and ”How can you do what is necessary?”, a manager is required to determine the prerequisites for higher—order targets. As each target is generated, managers can ask them— selves if they know how to produce that target. If the answer is "yes”, they should list the prerequisites (what they have to do) and begin to work on them. If the answer is ”no", working through the what is and how can questions will help them reach the necessary prerequisite levels. If a manager is not sure how to set up a successful training program, he BEHAVIOR MANAGEMENT Target: Eliminate dermatitis of hands and arms of workers in the formulation department. What is necessary? A. Find the source of the dermatitis. B. Eliminate or control the source of How can you find the source? A. Get expert consultation. the dermatitis. 1. Get clearance for consultation. 2. Call NIOSH for recommendations. 3. Contract with the consultant. How can you A. Procure gloves, scoops, control exposure to the powder? mops and buckets. B. Set up training program to teach workers to: 1. Keep bin lid closed. 2. Wear gloves when handling powder. {.0 Use scoops to remove powder. 4. Wet powder before dumping. 5. Rinse powder from gloves and containers. 6. Wet mop to remove powder spills. 7. Wash hands and arms after working with the powder. Figure 2-4. complex list of targets. should ask himself what is necessary to set up a training program to achieve all the targets that have been specified. In the preceding example, the manager has identified a choice. He can sim— ply tell the formulation supervisors and workers what the desired work practices are or he can develop a more elaborate program to train the workers to carry out the work prac— tices. He may even think of other alternatives such as posting reminder An example of a worksheet used to record information about a signs or writing a training guide. There is no sure way to choose among the alternatives. This choice will be governed by the complex relation— ships among the importance of the problem, the confidence that the var— ious alternatives will work, and es— timates of the costs of the various alternatives. Generally, the more effective the approaches appear, the more time consuming, complicated and expensive they will be. However, if problems are of a very high degree of importance, probable effectiveness Goals and Targets may be the first consideration, and costs almost irrelevant. SUMMARY COMMENTS Using the behavior management ap— proach to effect needed changes for improvement, it is necessary to re— cognize the existing problem and de- velop goals and targets accordingly. The chosen targets must reflect the specific changes needed to achieve 2-15 the goal. To fulfill these targets, and consequently their goals, vari— ous sources of information should be utilized. It is important that peo- ple's providing useful information is appropriately reinforced. The most important aspect of obtaining pertinent information is that man— agement has a clear understanding of their goals, and an organized method to achieve them. 10. BEHAVIOR MANAGEMENT REVIEW QUESTIONS Needs for change originate in information. True or False? Which of the following could lead you to believe a problem exists: a) workers come to work on time. b) supervisors have a schedule for changing ventilation filters. c) an insurance premium for hospitalization has doubled in the last year. d) the amount of time lost due to accidents is well above the industry average. Organizations may sometimes want to make situations not considered a problem even better. True or False? Goals are usually stated to reflect the general needed for a situation to become better. Goals are usually developed with the intent to solve problems or improve non-problem situations. True or False? Deriving targets involves considering the and that must occur for a goal to be achieved. Give three examples of a behavior and the product it might produce. 1) Behavior — Product — 2) Behavior — Product - 3) Behavior — Product — Goals are more general in nature and targets are more specific. True or False? List three steps necessary to begin a behavior management program. 1) 2) 3) No sources of information should ever be ignored. True or False? 11. 12. 13. 14. 15. 16. Goals and Targets List the five possible sources of information. 1) 2) 3) 4) 5) Which of the following would be considered the best concrete description of an existing problem? a) a bad safety environment. b) careless workers. c) loading dock workers are lifting improperly and experience back strains. d) workers are not concerned about safety and health. It is not important to reinforce information giving. True or False? List the two questions to ask when organizing complex targets and their prerequisites. 1) 2) Targets should always be defined and described specifically and concretely. True or False? Choosing the behaviors and products of behaviors that will probably be the most contributing and practical targets to achieve the desired goal is clearly a judgment decision. True or False? WK‘IOUUI (D 10. 11. 12. 13. 14. 15. 16. BEHAVIOR MANAGEMENT ANSWERS TO REVIEW QUESTIONS True 0 and d True changes True behaviors, products of behaviors For behaviors, any type of action True 1) Collect information about problems or desires for improvements. 2) Translate these into general goals. 3) Translate goals into specific targets. True 1) Line workers 2) Supervisory and middle management 3) Upper management and staff personnel 4) People outside the organization 5) Yourself (C) False 1) What is necessary to reach the target? 2) How can you do what is necessary? True True GodsandThqgu BIBLIOGRAPHY Drucker, P. F. The Practice of Management. Harper and Row, 1954. Quilitch, H. R. Using a simple feedback procedure to reinforce the submission of written suggestions by mental health employees. Journal of Organizational Behavior Management, 1978, 1, 155—163. Raia, A. P. Managing by objectives. Scott, Foresman and Company, 1974. LESSON THREE: DATA SYSTEMS — AN INTRODUCTION CONTENTS Page USEFULNESS OF DATA SYSTEMS ........................................... 3—3 PROVIDE A BASIS FOR DECISION MAKING ............................... 3—3 INDI CA TE NEED FOR S UPPLEMEN TA TI ON ............................... 3—6 PROVIDE A LEARNING TOOL FOR MAKING IMPROVEMENTS ............... 3—9 E VALUA TE IDEAS AND SUGGESTIONS .................................. 3—1 2 WHAT TYPE OF DATA TO COLLECT ......................................... 3—14 SUMMARY COMMENTS. . . . . . . . . . . .; ....................................... 3—19 BEHAVKfliMANAGEMENT OBJECTIVES After the completion of Lesson 3, the reader should be able to understand and discuss the following: 1. 2. How a data system can aid in decision making. How a data system indicates if and where programs or interventions need to be supplemented. That a data system is a tool for determining which targets and efforts are effective to make improvements. That a data system provides a basis for determining if what works in one situation will work in different situations. That a data system can be used to evaluate ideas or suggestions that come from literature. That the data collected should provide the information needed to determine if the targets and goals have been achieved. That data may need to be collected on more than one target level. That data collected can illustrate the relationship between prerequisite targets, higher—level targets, and goals. Terms and concepts the reader should be able to define, describe, and give examples of after reading this lesson. 1. 2. data system 6. effective prerequisite targets 7. graphic feedback intervention 8. validated suggestions or ideas supplementation 9. quantifiable data evaluation 10. levels of data collection 32 DATA SYSTEMS: AN INTRODUCTION INTRODUCTION Lesson 2 proposed that deriving tar— gets can help make the changes needed to achieve desired improvements in an organization. However, once targets are set, how is management suppose to know if their efforts are effective and if those targets are achieved? Granted, in some cases, it might be relatively easy for a manager to as— certain the answers to these ques— tions by asking the people in charge of achieving the target, by checking on it himself, or by obtaining infor— mational reports the organization normally generates. But, consider the fact that some goals and problems are very complex and may involve many targets. The targets may be quanti— tative, making it hard for management personnel to judge if improvements are occurring. An organization may not be set up to routinely provide information about these situations. Therefore, Lesson 3 deals with these concerns. USEFULLNESS OF DATA SYSTEMS The purpose of this lesson is to il- lustrate what management personnel must learn in order to develop a data system that will provide a way to ob— tain the information desired. Man— agement will have to determine if targets are being achieved, if they are truly important prerequisites for goals, and if any additional efforts are needed to help achieve these tar— gets or to possibly generate new tar— gets. Data systems provide a basis to make decisions. There are several reasons why data systems are useful. One reason is that they provide a sound basis for making decisions, especially ones that might involve targets. Assume that data have been collected over the past year to get average, 3-3 monthly, injury frequency rates as illustrated in graphic form in Figure 3-1. Suppose the plant manager knows that the industry average is quite a bit lower than the average for his plant. This information may lead him to decide that something needs to be done. Let's say he decides to start a new safety program. He implements it and, after a year of collecting new data, the frequency rates are as shown in Figure 3—2. The manager now has to decide if the program was ef- fective. The data obtained clearly illustrate that the safety program significantly caused a decrease in the injury frequency rates. For this reason, he should decide that the program should be continued. In the absence of these data, the manager would not have been able to recognize the need for improvement nor the suc— cess of the program. Suppose a department manager is mon— itoring personal exposures to poly— vinyl choride (PVC) and that there is a long record of keeping personal time weighted averages (TWA's) of key mold operators below 1 ppm. Suddenly, the TWA's Jump to anexposure level of about 10 ppm (see Figure 3-3). The manager decides to conduct a sur— veillance program and calls in engi— neers and the maintenance crew in an attempt to track down the source of the problem. An engineer discovers that the system for pumping PVC is maintaining higher pressures than ex— pected. All of the staff members agree that this is the probable source of the problem, and decide to have the engineer replace the regu— lating valves on the system to main- tain a lower level of pressure. After a day, the new data obtained plots as shown in Figure 3—4. Obvi- ously, the valves were not the source of the problem because the TWA's are still about 10 ppm. An alternative source of the problem must be derived. Someone suggests that the supervisor may be forgetting to turn on the ven— tilation system. The manager then BEHAVIOR MANAGEMENT 1+0_ In 3 m “c 30.. >. U C ‘5’ 20 U ._ Q L. LI. 5 10d .3 E 00 44f f f f l L r I I r r l .1 Q 3 ‘Y 5' b 7 J? 4) to H [:1 /v77 Mow/15‘ Figure 3-1. Average, monthly, injury frequency rates before implementation of safety program. I }(---New Safety Program I I 3 “0— E U I m l 01 I I 330— : C I Q) I 3 I 320 i L I Ll. I > I l— I ? ‘IOfl I‘ P I —. I I 00 I l l l l I r 1’ rl I [I | j I 1| I! [*1 l I l l 11‘3+yc’7fvth11 23 Vye'7quv-m I977 197! /V7 o n+4 5‘ Figure 3-2. Average, monthly, injury frequency rates after implementation of safety program. Data Systems: An Introduction In ppm \n PVC Day; Figure 3-3. Daily exposures to PVC; showing the recent jump in the TWA's. 10 I Va} R I d /:e:’ ve ep ace l l E l Q I a I c 5 : "' l U : > I 1 l I I r I 1 J r I {Q 3 If r e 7 y anS Figure 3-4. Daily exposures to PVC after valve was replaced. I Valve Replaced l I r//”: .: Super. Reminded I l E I Ie////// I I Q. : I C l --s s s U l l i : :x I I W 1' ' I 0 r I l I I I_' I I I DAYS. Figure 3-5. Daily exposures to PVC after supervisor reminded to turn on ventilation. BEHAVIOR MANAGEMENT decides to implement a program to remind the supervisor to turn on the ventilation. After this is done, new data are collected and are as shown in Figure 3- 5. The graphed data leads the manager to decide this course of action has solved the prob— lem. In the absence of all of these data, the department manager would not have been aware of the initially low expo— sure level, that the exposure to PVC had suddenly increased, that deciding to change the pressure valve was not the solution, or that the supervi— sor's forgetfulness was the probable source of the problem and, therefore, the decision to try to improve his practice was the correct solution. The manager may now decide to install an interlock so that other machinery cannot be operated unless the venti— lation is on. Data systems can indicate whether or not management needs to supplement an already existing intervention or pro- gram that is only partially effective. Suppose a manager is interested in how reliably line workers use person— al protective equipment such as safe- ty glasses, respirators, hard hats, and gloves. During the next few weeks, he collects some data on the usage of such equipment. The data obtained are then transformed into graphs indicating the weekly percent of appropriate usage for each kind of safety equipment (see Figure 3—6). After viewing the graphs, the manager decides to start a training program to increase line workers' usage of safety equipment. He collects new data on usage for a few weeks after the training has begun, and trans— forms them into the graphs illustrat— ed in Figure 3—7. It appears that the program was successful to in— crease the appropriate usage of all the safety equipment except respira— tors. This leads the manager to de— cide that the program was effective and probably should be kept; but it needs to be supplemented in some way to increase the usage of respirators. In the absence of these data, the manager probably would not have been aware of the initial percent of use, decided that the program was partial— ly effective, or known that the training program for respirator usage needed to be supplemented. Suppose from the preceding example, the manager decides to turn over the problem of low percentage use of res— pirators to the supervisor of the de— partment where most of the respirator use should occur. Realize that res— pirators need not be used eight hours a day, but only at appropriate times. The manager collects data for a few more weeks and again graphically transforms them as shown in Figure 3—8. As he looks at the graph, the manager realizes that the problem ac— tually became worse. A line worker informs him that the supervisor was not appreciative when workers did use the respirators, unreasonably re— quired workers to use respirators even when they were not needed, and was not sympathetic to the fact that g I _ , l D 100 : TraInIng Program : Supervisor's Efforts ‘ 1 I L l I a 75 : : < i l I u l WI , 5 25 : I E .I I \‘W m 0 Respirators . . . ' m I r I r I I I I f i F I W I i l 3 i S‘ b 7 9 [0 ll IL 13 IV 15’ week 5' Figure 3-8. Appropiate uaeof respirators after supervisor's efforts. Data Systems: An Introduction 100 100 75 75 o :v; 50 W 50 B 25 25 /\ f.“ a Glasses Respirators o I I I I T I I I I E I 2‘ 3 '7" 5' I Z '3 5‘ 5- < Way! Week; u. o 100 100 § 75 75 '/\// g 50 W 50 25 25 Harleatsl . Gloves I _I _I I I I I l_ F T I Z 3 (f r I 2 3 '/ 5” Weeks chkj‘ Figure 3-6. Appropriate use of protective equipment before training program. I 100 'Trainin Pro ram 100 ' Trainin Pro . IV 9 9 I 9 9 75 If 75 : i I 3 so \/": 50_ I I» 25 I 25 /\I\/\*“ Q I I fl) l l ‘E O Glasses : O Respirators ' Q 'T I l I I I I I II IIIIII I I Fl 3 Il3¥§e7svlo Iz3ys‘67a’9/o '3 k g wee 5 week: '4— o I E 100 Egreining Program 100 E€:i;gl£iflglfrog. I I 8 75 I 75 /\/I a) I I no I I 50 W: 50 I I : 25 I 25 I I I 0 Hard Hats I 0 Gloves ' I I I I I I I I I I l I T f l r7 I II 11375—67J’q/o IzJ‘fS'I-7fqlo Ned“ wcer Figure 3—7. Appropriate use of protective equipment after training program. BEHAVIOR MANAGEMENT :1: “9 g 70__ I'D I U) c L 60 2 w — x H— L o o 50 3 m 3 o u — hO C m m L. "-‘ Q) G. 2 30-. _ S u— .- 20 O > —’ - E o m 10 2 ED —" Assembly Lfine Hockers r. . I . l l 3 ‘f b— w et/(‘S‘ Figure 3-9. Instances of unsafe behavior by assembly line workers before implementation of feedback and discussions. E Feedback 8 Discussion 3 7o_ : .2 3 : g I 60_ : C L l 2 w a u- f 50.. : ° a : 3 ‘Dh0_ : u _ u C I 13 3 30. l (n D. I E L ' o 20- l u— ._ . o > u E ‘0- i O (D I z m . I Assembly Lupe Workers J . I I , , , 1 I z 3 ‘f 5' c 7 J ‘1 lo Week! Figure 3-10. Instances of unsafe behaviors by assembly line workers after implementation of feedback and discussions. Data Systems: An Introduction repirators can be hot and uncomfort— able. This information, along with the data obtained, increased the man— ager's understanding of the problem. In the absence of these data, the manager probably could not have con— fidently decided that respirator us— age needed improvement, that the su— pervisor was ineffective to increase such usage, and that the supervisor had such a negative effect that res— pirator usage decreased. The man— ager's next decisions should be to try to get the supervisor to be more supportive of workers, and to search for a new solution to the problem. Data systems provide a basis to learn what is probably effective to solve problems and make improvements in an organization. After management discovers which programs are effective to make one improvement, their best response may be to use these same or similar courses of action to make other im— provements. Suppose a plant manager is faced with the problem of a rising number of accidents and their related costs. He is particularly interested in the unsafe behaviors of assembly line workers in the production de— partment. The specific behaviors on which he wants to collect data are ”horsing around” on the conveyor belt, stepping in front of moving parts, stacking parts not being work— ed on where they are in the way, and not putting on gloves to handle sheet metal. These are behaviors that the manager believes might cause acci— dents and injuries. He then develops a way to collect some initial data on the number of these behaviors occur— ing weekly. Data are recorded for a few weeks and then transformed into a graph (see Figure 3—9). The manager realizes that there is a problem of too many of these unsafe behaviors, and decides to improve the situation. His approach to try to solve the problem begins by providing the pro— duction department supervisor with the initial graph. The manager tells the supervisor that he will continue to take data and will update the graph every week. The supervisor is to post a feedback graph in a con- spicuous place and briefly discuss its meaning with the workers. The results are shown in the graph in Figure 3—10. These results give the manager a degree of confidence that the feedback graphs and discussions were effective to decrease the rate of unsafe behaviors. The plant manager then realizes that the problem of too many unsafe behav— iors might also exist for loading dock workers in the shipping depart— ment. Because he learned that feed— back graphs and discussions worked to decrease the rate of unsafe behav— iors of assembly line workers, he de— cides to use the same methods in an attempt to improve the behaviors of the loading dock workers. The spe- cific behaviors in which he is inter— ested are: lifting improperly; reck— less forklift driving; careless han— dling of boxes when unloading them from a truck; and making tall, un— stable stacks of boxes in heavy traf— fic areas. The manager now develops a way to collect data on these behav- iors, then collects some initial data for a few weeks. He transforms these data into a graph, provides the ini— tial graph and weekly updated graphs to the shipping department supervison and asks the supervisor to post these graphs in a conspicuous place and discuss them with the workers once a week. The results of these efforts are as shown in the graph in Figure 3-11. From these results the manager can conclude that posting and discussing feedback graphs also worked to de— crease the number of unsafe behaviors of loading dock workers, even though they were in a different department and exhibited different unsafe behav— iors than the assembly line workers. His further conclusion may be that he has discovered a way to affect unsafe behaviors that has general applica— bility. The techniques used may work in different areas and with different problems. The manager has begun to learn effective ways to make improve— ments and solve problems in his or— ganization. Suppose that in the previous examples with assembly line workers, the man- ager decided to take a different ap— proach. Assume the initial data BEHAVIOR MANAGEMENT Loadipg Dock Workers 1 v I Z 3 i’.‘ 0 :1 7°- “6 E '2 L 60_ D d.) u— ‘35 o o 50_ m 3 8 0 1+0 : F 1 m L C — L H— .‘3 2°_ 0 > - E o m 10 2 CD ‘ 6 ’7 a? c? 10 week S Feedback 8 Discussion Figure 3-11. Instances of unsafe behavior by loading dock workers after implementation of feedback and discussions. obtained are the same as before. In— stead of using posted feedback graphs and discussions, the manager decides to use only verbal feedback. He in— forms the production supervisor that assembly line workers exhibit too many unsafe behaviors, and that he (supervisor) is to give those workers verbal feedback on their behavior. For the moment, presume that the man— ager does not give the supervisor any instructions about what kind of ver— bal feedback to give or when to give it. The manager then collects addi— tional data for a few weeks after verbal feedback is begun. The re— sults are as shown in Figure 3—12. The manager concludes from the graph that the verbal feedback of the su— pervisor was ineffective and perhaps even harmful in effecting positive behavioral change. An assembly line worker suggests to the manager that the supervisor gave mostly negative feedback and that what little posi— tive feedback he gave was not sincere This example, coupled with the earli— er example in which a supervisor was expected to increase the proper usage of respirators, could lead managers to believe that verbal feedback by supervisors, without the proper tech— nique, can be ineffective for improv— ing safety behavior — and can possi— bly make it worse. Suppose that after the manager learns that the verbal feedback of the su- pervisor was ineffective, he decides to train the supervisor to improve his practice of giving verbal feed— back to workers. The manager in- structs the supervisor in how to give more positive feedback and very lit— tle negative feedback, how to be sin- cere when he gives each type, and how to tell when each should be given. The manager then sends the supervisor back out to apply this new feedback practice. More data are collected on unsafe behaviors. After a few weeks, the results are transformed into a graph (see Figure 3—13). From these data, the manager can conclude that this form of verbal feedback by the supervisor was effective to decrease the number of unsafe behaviors. By using a data system and making graphs, the manager has learned that the inital form of verbal feedback the supervisor gave was ineffective to decrease the number of unsafe be— haviors. He also discovered that the verbal feedback after proper training was effective. Data Systems: An Introduction 3 : Verbal Feedback 0 II I m o 50 I (U I l U1 I 5 3 no : ‘/’,,4————4,,,/r——__. .X 1 ‘ \4— L. I ° § 30 I In I w o I 8 " 20 I (U \- | u CD l ‘2 ‘L 10 l _ |_ l u— .9 .' o > Assembly qukers, _ . I _ _ . g I I r I I r I I I I g 33 I Z 3 ‘7‘ 5’ ‘0 7 a? c; m Figure 3-12. verbal feedback of supervisor. Assembly Workers KI WEf Verbal Feedback 2/ Instances of unsafe behavior by assembly line workers after After Training I I I I I l I l l I l I I I I l I I l I l I I l i’.‘ m :50 w 0 (III 2 LlIo D d) Iii o 030 3 U) 8820 I: IDI— 7.310 E —L 0 III—.— 0 > “I '.C O 0) 211! I | I I IL3‘/ I ‘7' | I I J I I0 week! V‘— GN— I I I I I I/ Figure 3-13. Instances of unsafe behavior by assembly line workers after supervisor was trained to improve his practice of giving verbal feedback. 3-11 BEHAVKHiMANAGEMENT Data systems can be used to evaluate ideas or suggestions that come from literature. Organizations sometimes receive or acquire literature such as pamphlets, booklets, safety journals, etc., that are loaded with unevaluated ideas and suggestions. Data systems are a re— liable method that can be used to evaluate and test those ideas and suggestions. As the persons managing an organization practice using data systems, they will become more crit— ical about accepting unevaluated ideas or suggestions. Even if man— agers decide to use some of these suggestions, a data system can be used to evaluate their worth. Suppose that the manager of a large, concrete—products manufacturing plant is concerned about the problem of a rising number of accidents that in— volve material batching and mixing workers. He receives a pamphlet in the mail one day that includes sever— al suggestions about how to reduce the number of accidents in plants like his. A few of the suggestions that catch his attention are: to post additional warning signs; to provide better personal protective equipment; to impress upon workers the need for constant awareness of hazards; and to hold weekly safety meetings to discuss safety problems with workers. The manager is criti— cal of these ideas because the pam— phlet does not include any proof that these suggestions have been tried and tested. However, he decides he would like to try a couple of the sugges— tions himself. The manager first tries the suggestion to post addi— tional warning signs and posters. He instructs the foreman responsible for supervising workers in the batching and mixing department to post some additional warning signs in this area The manager collects data for the next few weeks to compare with previ— ous data taken on the number of acci— dents. The previous and new data are then plotted on a graph (see Figure 3—14). The manager notices that the number of accidents, on the average, did not decrease after the additional warning signs were posted. He dis— cards the idea, or at least assumes that it must be supplemented in order 3-12 to produce significant effects. Next, the manager decides he wants to try the suggestion to hold weekly safety meetings for the batching and mixingworkersso they can discuss safety problems with their foreman. The manager instructs the foreman to organize these meetings and continues to collect data on the number of ac— cidents for the next few weeks. The data are illustrated in the graph in Figure 3—15. The manager finds that the number of accidents dropped con— siderably, so he decides that the safety meetings are a worthy sugges— tion for him. What this manager learned, and what all managers should learn, is to be more critical of unevaluated ideas and suggestions, and how to evaluate the ones they might want to try. By using a data system to evaluate and test ideas and learn what is ef— fective to solve problems, a manager is in the position to provide vali— dated suggestions to others who ask for his help. Suppose that you are the plant manager for a farm machin— ery and equipment firm. You have al- ready evaluated and tested the method of having someone post graphic feed— back and discuss it with workers, finding it to be very effective to decrease the number of unsafe worker behaviors in your plant. One day, the manager of a similar plant owned by the same firm calls you on the phone and, in the context of the con— versation, mentions that he thinks there has been a noticeable increase in unsafe worker behaviors in his plant. You tell him that you once had a similar problem and tried a particular method that was successful. You then suggest that he might give it a try and fill him in on the de— tails of how to do it, including how to make sure there really is a prob— lem. He agrees to try it and thanks you for your suggestion. A couple of months later, he calls you back and thanks you again. He explains that your method helped him to elicit more safe and fewer unsafe behaviors from his workers. Consider what has been learned. As the plant manager, you have learned Number of Accidents Data Systems: An Introduction Batch 8 Mix Workers I I I I / 7. 3 '+ 5’ G weflk/S' Additi onal Signs Posted 7 a? q l0 Figure 3—14. Number of accidents by batch and mix workers after additional warning signs were posted. Number of Accidents Batch 8 Mix Workers érfjjijional Signs I I l I I I l I I i i /13~/5' 1. Improper lifting. 2. Not cleaning up spills. 3. Not using eye protection. CC?» Figure 5-3. An example of an observation form that has been used to record the existence and location of hazards. BEHAVIOR MANAGEMENT practice sessions in which they re— peatedly identify hazards, and record data about them until a confident level of accuracy and agreement is reached. Having completed adequate training exercises, the observers should begin to record initial or baseline data. Baseline for a hazard control program can best be described as data re- flecting the hazardous practices and conditions that exist in a workplace before any intervention efforts are implemented. Collecting baseline da— ta provides a basis by which to judge the improvements or changes made after a group of employees is placed in the intervention condition. INTERVENTION After baseline has been obtained for a preselected period of time, the follow—up step is to implement the chosen intervention. Corrective ac— tions, such as arranged consequences, which have been selected to produce the necessary improvements to achieve targets and goals, should now be ini- tiated. If a feedback system is to be implemented, the appropriate feed- back, whether written or verbal or both, should be given following the predesigned feedback procedure. If, in addition or instead, other kinds of interventions such as incentive or reward programs have been chosen, these corrective methods should now be applied. The final step concerning the collec— tion of data is to obtain additional data after the intervention has begun If feedback is to be given based on the data recorded about hazards dur— ing observations of work areas, the data collection process should be de— signed to provide ongoing information about the hazardous practices and conditions that need to be corrected. A successful method, applied in both of the previously mentioned stud- iesl'2, is to present the completed form from the most recent observa— tions to the appropriate people, us— ually supervisors and foremen. Ver— bal feedback, especially praise and recognition for improvements, can easily be presented along with writ— ten feedback. This should enhance the effectiveness of the feedback system, and thus reduce the frequency of undesirable behaviors and products in many situations and settings. TRANSLATING THE RESULTS After data have been collected for a preselected period of time, the first step in translating the results is to derive numerical representations of the data. It is wise to begin this practice by counting the occurrences of behaviors or products about which data has been recorded. In most cases dealing with hazard control, this involves counting the number of hazardous practices and conditions that were cited and recorded during observations in work areas. For ex— ample, a manager looks at the obser— vation form shown in Figure 5—4 and counts a total of ten hazards. If a manager desires, other figures such as means of percentages can be derived from the totals obtained by counting occurrences. Suppose that, once a week, a manager conducts in- spections for hazards in a production department. He obtains baseline data for eight weeks. At the beginning of week nine, he implements the chosen intervention — weekly written feed— back to supervisors. He then col— lects additional data for eight more weeks. Assume he counts the occur- rences of hazards observed and ar— rives at the following totals: Number of Hazards Week 1 23 Week 2 18 Week 3 22 Week 4 17 Week 5 16 Week 6 24 Week 7 19 Week 8 21 Total = lgfi Mean frequency of hazards during Date (% Z i g '7; Observer {:00 Time A. Obstruction of Walking—Working Surfaces: 1. 2. Hazard Control: Written Feedback Waste on floor. Unused equipment in the way. B. Electrical Hazards: l. 2. 3. 4. Uncovered socket. Overloaded socket. Frayed electrical cords. Misuse of extension cords. C. Machine Guarding: 1. 2. Misplacement of guard. Guard removed. Sapim «ad 0.4 @ ® C“) \ (C3 Figure 5-4. An example of a completed observation form. BEHAVKHIMANAGEMENT baseline = Total number of Hazards Number of Observations Mean = 160 = —§— 20 Hazards Number of Hazards Week 9 15 Week 10 11 Week 11 12 Week 12 9 Week 13 5 Week 14 6 Week 15 4 Week 16 2 Total = 64 Mean frequency of hazards during 6_4 8 = 8. intervention = The manager then decides to compute the percent of decrease from the baseline condition to the feedback condition in the following way: Percent of Decrease = Baseline Mean — Feedback Mean Baseline Mean Percent of Decrease = 20-8 _ 12 20 _ 20 The next step, and one which is wide— ly practiced though optional, is to use tables and/or graphs to illus— trate the results produced by the hazard control program. Assume that data about hazards in a production department have been obtained daily and that weekly mean frequencies have been computed in the following way: = 60% Number of Hazards Cited Day 1 15 Day 2 18 Day 3 16 Day 4 12 Day 5 14 Total 75 Weekly Mean Frequency of Hazards = lg = 15_ The mean frequegcy of hazards for each week during baseline and inter- vention is as follows: Mean Frequency During Baseline (Weekly) Week 1 15 Week 2 18 Week 3 17 Week 4 14 Week 5 10 Week 6 19 Week 7 20 Week 8 16 Total 129 Mean Frequency during Baseline = 1%: 16.1 Mean Frequency During Intervention (weekly) Week 9 15 Week 10 9 Week 11 7 Week 12 10 Week 13 3 Week 14 2 Week 15 5 Week 16 _4 Total 5 Mean Frequency during Intervention= 55 _ —§ — 6.9 Hazard Control: Written Feedback These data may be further illustrated in graphic form (see Figure 5—5). EVALUATE THE HAZARD CONTROL PROGRAM After the data have been translated, management can begin to evaluate the hazard control program. The first decision to make is whether or not the chosen intervention was effective to produce the changes needed to achieve the targets and goals and, therefore, solve the problem. Compu— tations and illustrations of data (means, percentages, tables, graphs, etc.) can often be used to aid in this decision. Suppose a manager has a problem with hazard control. He sets a goal to improve hazard control, and more specifically sets a target to reduce the frequency of hazardous practices and conditions. He collects baseline data for twelve days. The interven— tion he implements after baseline consists of providing a feedback sheet to each of two foremen. The written feedback specifies the types and locations of hazards cited during the most recent observation. In ad- dition, the manager discusses with each foreman suggestions about how to correct these situations. The man- ager then obtains data for twelve days after the feedback practice is begun. The results are translated into graphic form and are as shown in Figure 5-6. Based on this graph, should the manager decide that the feedback method is effective to re— duce the frequency of hazards? Ob— viously, this intervention was effec— tive and, therefore, should be continued. Again, suppose a manager is concerned with the existence of too many poten- tial hazards. He obtains baseline, implements the arranged consequences (posting warning signs about potential hazards), and collects additional data. The results are graphed and are illustrated in Figure 5—7. After looking at the graph, should the man- ager decide to keep the intervention? Probably not, because the warning signs do not appear to have reduced the instances of hazards. Remember that arranged consequences can sometimes be only partially ef— fective and, therefore, need to be supplemented in some way. Assume that a manager in charge of four pro— duction departments is interested in improving hazard control. He sets a target to reduce the mean frequency of hazards cited in each of the four departments. The data obtained are illustrated in Figure 5—8. Should the manager continue the interven— tion? Yes, because it was successful to reduce the mean frequency of haz— ards in all four departments, espe— cially in two of them. However, the intervention needs to be supplemented because there is clearly room for further improvementin the other two departments. How is a manager to decide which sup— plementary efforts are needed? He can rely on his own ideas or ask his supervisors and/or foremen for sug— gestions. Line workers may even be able to provide insight into which efforts by management might further improve hazard control. Managers can also try to acquire booklets, pam— phlets, publications, etc. to obtain suggestions. These sources of ideas can often provide suggestions about arranged consequences that worked to solve similar problems. Even if the setting and kind of problem are en— tirely different, arranged conse— quences with general application can be revealed. After a manager decides to implement supplementary efforts, he should go through the same process of obtaining data and analyzing the results. There are some other adjustments that might need to be made. The data sys— tem may need to be changed. If haz— ards that were not initially identi— fied have surfaced during observa— tions of work areas, these hazards will need to be defined and included in the list of hazardous practices and conditions. Suppose the people who make observations have had unex— pected trouble agreeing on the iden— tification of hazards. In this case, the observational definitions of cer- tain hazards may need to be discussed again and refined. Further judgements related to hazard— ous practices and conditions in an BEHAVIOR MANAGEMENT 13‘ Baje’l'nf In‘I-crucn‘f‘ion Mann Fr‘equww O‘F Hazard: 1 III III [Z3‘ffé7d’alIOlllll3i‘f/J‘lé Wti KS Figure 5-5. An example of illustrating the mean frequency of hazards in graphic form. Hazard Control: Written Feedback (34 SC Ib1€ 20-4 19 - I?- I7 - IQ- I)" I3 - I1‘ I! - Io- I I I I I I l I q - I l I I I I | l Ffeed.£ac}< / 0*? Ha zal‘d! ;. 7. 6- 5‘. y- 3-I z-I la A/mméf" ‘ 1 1 III I I l I l I I 'l ,7, 3 Y 5' G 717/o/I Italy/{1517;371:110 Bayf Figure 5-6. The graphic representation of data obtained before and after implementing feedback as an arranged consequence. BEHAVIOR MANAGEMENT . I I] . 5" . I l L ‘ l nS BaSfi/ne “(/Ilmnmg j I I 10" I I lo“ I I 16" I I N" I I I I | I I I1‘ 043 Hazar‘alj‘ A} up” ECI‘ ~11 I I I 7— ‘ I I l I Illill IIIIIXIII ll 3 ‘fr 9 ‘7!«I/aIIIzIJI-f/rwnIf/qzc Figure 5-7. The results obtained before and after warning signs were posted. Hazard Control: Written Feedback Mean Frequency of Hazards During Baseline Department 1 18.3 2 20.5 3 30.9 4 15.1 During Intervention % or Decrease 2.1 89 4.3 79 15.6 50 9.2 39 Figure 5-8. A table illustrating the mean frequency of hazards during baseline and intervention, tion to the intervention condition. industrial setting may be made through the acquisition of more data. Statistics on safety and health, in— cluding the number of accidents (both major and minor), injury frequency and severity rates, and work—time lost due to accidents, are usually already recorded by each individual industry. Therefore, it would in— volve a minimal amount of time, cost and effort to obtain these kinds of data, and infer their relationships with the reductions of potential haz— ards. SUMMARY COMMENTS If a hazard control program is pro— perly developed and implemented the benefits gained should far outweigh including the percent of decrease from the baseline condi- the costs imposed. It usually re— quires only a little time and effort to adjust work operations to accom— modate hazard awareness and control procedures. If an incentive or re- ward program is implemented, there may also be monetary costs involved. However, because adequate hazard con— trol is a prevention practice, there can be other benefits in addition to reducing the frequency of hazards. These benefits might include lower insurance premiums for accident cov- erage, elimination of fines imposed by OSHA for hazard violations, reduc— tions in the amoung of work—time lost due to accidents, and a decrease in expenditures for workmen's compensa— tion. BEHAVIOR MANAGEMENT REVIEW Following are the questions a manager should consider setting up and implement— ing a hazard control program: 1. 2. 10. ll. 12. 13. 14. 15. 16. 17. 18. 19. Has the problem been accurately identified? Has a goal been set to correct the problem? Has a target been set to reduce the number of potential hazards that cur— rently exist? If necessary, have other targets been specified to make improvements in sit— uations or statistics related to hazard problems? Have all potential hazards been identified and accurately defined? Using operational definitions of hazards, can separate people agree on the existence of a hazard? Have the shop or work areas where observations will take place been speci— fied? Has it been decided how often to conduct observations? Has a form on which observers can record data been developed? Have all observers practiced recording data about hazards until a confident level of agreement has been reached? Have baseline data been collected for a selected period of time? Has the chosen intervention been implemented after baseline data were obtained? After beginning the intervention, have additional data been obtained? Have statistics been computed from the data? Have these statistics been transformed into the tables or graphs necessary to illustrate and further analyze the data collected? Have all arranged consequences that were applied been evaluated to decide if they should be continued or discontinued? If results show there is still room for improvement, have supplementary efforts been considered and decided upon? Have needed adjustments of the data system been made? If other areas that need improvement have been discovered, have new targets been generated? Hazard Control: Written Feedback REFERENCES Sulzer — Azaroff, B. Behavioral ecology and accident prevention. Journal of Organizational Behavior Management, 1979. Sulzer — Azaroff, B. and Consuelo de Santamaria, M. Cost—effective industrial safety hazard reduction through applied behavior analysis. Submitted for publication, Journal of Applied Behavior Analysis, 1979. 5-21 LESSON SIX: EXPOSURE CONTROL - TRAINING AND FOLLOW UP CONTENTS Page EXPOSURE CONTROL IN A FIBERGLASS PLANT .............................. 6—3 SETTING UP AN EXPOSURE CONTROL PROGRAM ............................. 6—6 DEVELOPING THE OBJECTIVES ............................................. 6—7 DEVELOPING A DATA SYSTEM ............................................. 6—7 INTERVENTION: TRAINING AND FOLLOW UP ................................ 6—9 TRANSLATING THE RESULTS .............................................. 6—9 EVALUATING THE EXPOSURE CONTROL PROGRAM ........................... 6—12 SUMMARY COMMENTS .................................................... 6—12 EXPOSURE CONTROL: TRAINING AND FOLLOW UP INTRODUCTION Engineering improvements have done much in recent years to reduce workers' exposures to toxic chemicals. These improvements have centered around technology such as enclosing chemical processes, designing venti— lation systems, and developing better personal protective equipment. There has been relatively less emphasis on getting workers to behave in ways that would minimize their exposures to such chemicals. It is generally assumed that it is safer to depend on reliable engineering mechanisms than on human behaviors to reduce exposures, attributable to the fact that a lot of people doubt that human behaviors can be controlled with any sufficient degree of preci— sion. Nevertheless, human behaviors can be important in exposure control. Examples in which human behaviors become important include instances of emergencies and equipment break— downs, at which time a worker may be subjected to an unnecessary and unsafe degree of exposure unless he behaves appropriately. Proper beha— vior is obviously involved in all examples of work—related personal hygiene, such as the correct use and maintainance of personal protective equipment. EXPOSURE CONTROL IN A FIBERGLASS PLANT The following application of beha— vioral management methods is based on research conducted by Conard, Hopkins, Fitch, Smith, Anger and Dangel (1979).1 Because many chemicals are only moderately toxic; they may be used in open processes. However, workers should still try to minimize their exposures to these chemicals. One such chemical used in the fiberglass— 6-3 reinforced plastics industry is styrene. Styrene is a polymerizing agent for resins. A styrene—resin mixture is reinforced with fiber— glass to make products such as boats, shower stalls, truck cabs and many other products. Typical operations include hand—spraying various mix— tures onto molds, hand-laying fiber— glass reinforcing pieces soaked in resin—styrene mixtures onto partial- ly formed products, and using metal rollers to remove gas bubbles from the plastic mixture while it is still soft. The current OSHA standard for sty— rene exposure is 100 ppm, TWA. This level of exposure can be met in industries if they employ and main— tain proper exhaust ventilation in the area in which the styrene is introduced into the manufacturing processes. However, because the chemical is used in open processes, many of which are done by hand, there are abundant opportunities for workers to reduce exposures by altering their behavior. The research team listed above accepted the challenge to reduce exposures by changing workers' behaviors in a production department of Labconco, a company that makes laboratory equipment such as fume heads and bactoriological glove boxes of fiberglass—reinforced plas— tics. The goal of the program was to reduce the workers' exposure to styrene. This would require the identification of jobs which involved higher concentrations of styrene, specifying behaviors which were likely to reduce exposures for workers in those jobs, and training the workers to engage in those behaviors. To develop targets, an industrial hygienist was hired to take air sam— ples in the work areas. It was de— termined that the grestest exposures to styrene occurred for workers BEHAVIOR MANAGEMENT who sprayed the wet plastic mixture, and those who rolled the gas bubbles out of the mixture after it had been sprayed. After observing the workers in action, the research staff developed a list of work behaviors important in con— trolling exposure levels. The beha— viors included such things as the spray—workers‘ turning on the exhaust ventilation before they began spray— ing, keeping their work close to the ventilation exhaust, and spray— ing towards the exhaust. Appropri- ate behaviors for the roll-out work— ers included working on the upwind side of parts,and wearing respirators whenever their work required that they place their heads inside parts. Altogether, twelve potentially impor— tant behaviors (target behaviors) were identified for each worker. The alterations of these specific behav- iors to reduce exposures were estab— lished as the prerequisite targets. The terminal target, to reduce the occurrance of exposures of particular individual workers, was presumed to be the product of these prerequisite targets or the changes in the target behaviors. Two completely independent data sys— tems were employed, one to measure the extent to which the selected worker behaviors changed, and a second to measure styrene exposures. The prerequisite target behaviors were measured because they were assumed to be crucial to exposure reduction and because it was, at that time, unknown whether or not such behaviors could easily be changed. Observers were employed to collect data on the workers' behaviors. Each observer was provided with a stop— watch and was given a list of written definitions for each of the desired work behaviors and data collection sheets much like those shown in Figure 6—1. Each of the symbols on the data collection sheets represented a par- ticular work behavior. The hori— zontal rows indicate fifteen—second time intervals. The observers were instructed to carefully observe a worker while keeping track of observation time, and to note on the data sheet each behavior that occurred during each fifteen seconds of elapsed time. They practiced recording the behaviors until two observers, working far enough apart that they could not copy from each other, would agree that particular behaviors did or did not occur. They were then assigned to observe the workers on a random schedule which insured that they would not obtain biased estimates of how often the behaviors actually occurred. Air samples were also taken on a random schedule to avoid biases. Sampling procedures, based on NIOSH recommendations, were used to mea— sure exposures to styrene. Bendix- BDX Super Sampler air pumps were calibrated to draw known volumes of air—per—unit—time. Fresh charcoal tubes were affixed to the air—intake tubes of the pumps. A pump was fastened to a workers' belt and the charcoal tube clipped to a workers' collar so that the open end of the charcoal tube hung just below the workers' chin. The operation of the pumps was monitored to keep the rates of airflow constant. At the end of each day, the charcoal tubes were sealed and shipped to a labora- tory which performed analyses to determine the quantity of styrene trapped by the charcoal per volume of air pumped. Both behavioral data and air samples were collected for several days to determine the baseline or pre—treat— ment levels. The intervention developed by the research team was simple and straightforward. One member of the team, Conard, would meet with one of the workers at the beginning of a shift and briefly describe each of the behaviors that was appropriate for that particular person. If the worker indicated a lack of under— standing about any behavior, Conard either demonstrated how it should be done or had the worker perform it as he understood it. This training session ended with the worker per— forming all of the behaviors select— ed for him or her while Conard pro— vided feedback regarding how Exposure Control: Training and Follow Up JOB DATE EMPLOYEE OBSERVER : ROLLOUT OBSERVATION CODE Turning mold T= Uncovered skin H = Head in mold U: Working N = Not working B = Booth R = Respirator on F: R = Rollout area 0 Fan E = Exposure to spray Downwind #5 D: Outside 15 60 AS 30 15 60 30 W H F W H F W H F W H.F W H F W H F W H F W H.F W H F W H F O U R O U R 0 U R O U R O U R 0 U R 0 U R 0 U R 0 U R 0 U R RDT. RDT RDT RDT RDT RDT RDT RDT RDT RDT BNE BNE BNE BNE BNE ENE BNE BNE BNEBNE n W H F W H.F W H F W H.F W H F W H F u H.F W H.F W H F W H F O U R O U R 0 U R O U R O U R O U R O U R O HVR O U R O U R RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT DUNE BNE BNE BNE BNE BNE BNE BNE BNEBNE I Wnnr. WHHE. Wnnr. W Hpr W Hpr W Hpr Wunc. W Hr. Wunc. W H—r O U R O U R O U R O U R 0 U R 0 U R 0 U R 0 U R O U R O U R RDT RDT RDT RDT RDT RDT RDT RDT RDTRDT BNE BNE BNE DUNE BNE BNE BNE BNE BNE BNE -WHF WHF WHF WHF WHF WHF WHF WHF WHF WHF OUR OUR OUR OUR OUR OUR OUR OUR OUR OUR RDT RDT RDT RDT RDT RDT RDT RDT RDT RDT BNE BNE BNE BNEBNE BNE BNE BNE DUNE BNE 2 .h ,b Ru 0 2 b. ,b RU 0 1 «I .l .I .l 2 1 W H.F ”Wunr. W H.F Wnflcu W H.F ”WunF. WHHE. WHHc. W HUF W H F OUR OUR OUR OUR OUR OUR OUR OUR OUR OUR RDT RDT RDT RDTRDT RDT RDT RDT RDT RDT BNE BNE BNE BNEBNE BNE BNE BNE BNE BNE 1 W H F W H.F W H F W H.F W H F W H F W H F W H.F W H F W H F 0 U R 0 U R 0 U R O U.R 0 U R O U R O U R O U_R 0 U R O U R RDT RDT RDT RDTRDT RDT RDT RDT RDT RDT BNE DUNE BNE BNE BNE BNE BNE BNE BNE BNE IWHF WHF WHF WHF WHF WHF WHF WHF WHF WHF O U R O U R 0 U R O U R O U R O U R O U R O U R 0 U R 0 U R R D T R D T R D T R D T R D T R D T R D T R D T R D T R D T BNE BNE BNE DUNE BNE BNE BNE BNE BNE BNE I W H.F W H.F W H F W H.F W H.F W H F W H_F W Hsr W H.F W H F 0 U R n.U R 0 U R 0 H.R 0 U R 0 U R 0 U R 0 U R 0 U R O U R RDT RDT RDT RDT RDT RDT RDT RDT RDTRDT BNE BNE BNE BNE BNE BNE BNE BNE BNE BNE 1| 3 5 7 9 .I 3 5 7 9 1 1 l I I An interval form for recording data about a rollout employee. Figure 6-1. BEHAVIOR MANAGEMENT correctly the behaviors were per— formed. Once a day for the next three weeks, Conard visited each of the workers for a few minutes to observe how the behaviors were being performed, and to provide feedback. He was careful to avoid feedback that might be punishing. If a worker was observed doing something wrong, such as spray— ing away from, rather than towards, the ventilation exhaust, Conard simply explained to the worker what he was doing incorrectly, what result that would probably have on exposures, and what could be done to alter his work behavior. For example, a typical conversation would have been as follows: Conard: "Hey, D.L. do you have a second to talk?” Worker: ”Sure.” Conard: ”You were just then spraying into the wind. That will cause the styrene vapors to blow right back in your face.” Worker: "1 see what you mean.” Conard: ”How about working from that side and spraying back this way? Then the vapors will blow away from you.” Most of the feedback Conard gave to workers was positive: ”I see you remembered to turn on the vent fan. That's good.” ”I saw you wore your respirator while you were rolling out that hood liner. That saved you from getting a big dose of styrene." ”Everything's looking good." The data collected on workers' behavior after training reflected positive results. Most of the workers began to reliably engage in the desired work behaviors much more consistantly after training than they had initially. Some of the workers reliably performed some, but not all, of the desired behaviors before they were trained. For example: one of the workers who sprayed the resin— styrene mixture always remembered (even before training) to activate the ventilation system for his spray booth. However, among the behaviors that did not occur at acceptable levels before training, and there were several such behaviors for each 6-6 worker, there was only one behavior that failed to change as desired after training. One worker declined to wear a respirator under condi- tions in which it would have been appropriate for exposure control. Data on styrene exposures also continued to be collected after training. The mean decrease in styrene exposures of individual workers ranged from 36 to 57 percent Thus, the training procedures were generally effective to change the selected prerequisite behaviors as desired; the substantial reductions in styrene exposure were correlated with the changing behaviors. The procedures used to change the behaviors have been referred to as training. That is an oversimplifi— cation if one were to consider train— ing to mean only those things that were done during the initial 10 min— utes Conard spent with each worker. Recall that there was an important training component involved in the daily follow—up feedback visits whenever Conard noticed that a worker was not behaving as initially instructed. Also notice that the feedback visits provided Conard many opportunities (through his positive comments) to reinforce the target behaviors that were occurring. The distinction between the changes in behavior due to the initial training and the procedures used in the follow—up visits is crucial. Some behaviors changed as desired immediately following training. These quick changes may have been largely due to the training. How— ever, some behaviors changed slowly over several days. These changes might be attributed to the follow—up activities more than to the ini— tial training. SETTING UP AN EXPOSURE CONTROL PROGRAM Because exposure control is an important aspect of any organiza— tion using toxic chemicals, an updated list of all such chemicals used in each process should be kept. Most organizations have such a list, but if one has not been made, a Exposure Control: Training and Follow Up manager should construct one. To initiate a successful exposure con— trol program, management must demon— strate its concern about keeping em- ployees healthy in the face of threat posed by the chemicals. DEVELOPING THE OBJECTIVES The goal for any organization insti— tuting an occupational health program such as exposure control is to keep employees free of disease and symp— toms that might result from work practices or conditions. The parti— cular goal for those organizations concerned with exposure control, such as the Labconco Fiberglass Plant, would be to minimize exposures to toxic chemicals, therefore keeping all exposures within OSHA standards. It is recognized that various engi- neering controls are pertinent in keeping exposures low. However,engi- neering considerations are beyond the scope of this guide. Our immediate focus is on producing behavior,that will minimize exposures and contri- bute to the good health of employees. The particular target behavior a man— ager will need to alter to achieve the goal, will depend on the chemical substances from which the worker should be protected, and the physical facilities (including engineering controls) with which the employees have to work. Because each industri— al Situation is different it is im— possible to list all of the factors a manager should consider in developing a successful control program. How— ever, there are general basic proce— dures pertinant to any exposure con— trol program; these are presented below as a general guideline. DEVELOPING A DATA SYSTEM As is true with any control program, a data system is needed to moniter the achievement of targets and goals of an exposure control program. To do this effectively, several systems may have to be implemented. One sys- tem should be designed to measure changes in the specific target behav- iors related to exposure control for that particular organization. A man— ager should begin by developing an 6-7 operational definition of each target behavior or product. The definitions should be stated so that observers will be able to consistently apply them to decide whether a particular behavior or product occurs. Consider a few of the definitions of the tar— get behaviors specified in the ex— posure control study described ear— lier. (See Figure 6—2.) Another data system should be design— ed to measure exposure to the chemi— cal or any other substance about which a manager is concerned. For recommendations about how to measure the amount of exposure to which a worker is subjected, a manager might find it useful to contact a qualified consultant. An industrial hygienist can provide suggestions about what kind of equipment can be used to ob— tain accurate exposure measurement, and how to use this equipment. Data for both of these systems can be obtained at the locations of the jobs for which exposures are a concern. A manager should have already identi— fied the work areas where target be— haviors and exposures will occur, and, therefore, would not have any trouble specifying where observations and exposure measurements should take place. Obviously, exposures should be measured in the same place where the target behaviors or products will occur. A manager's next step should be to decide how often to collect data. Usually, observations of target be— haviors or products that are presumed to reduce exposures, and measurement of exposures should be carried out simultaneously. As a general rule, a manager should select random times during the chosen frequency of data collection (daily, weekly) to obtain a representative sample of what regu- larly occurs. Whoever is in charge of data collection should develop a schedule of when observations and measuring expsoures are to be con— ducted. A manager may have several available alternatives in deciding who will re— cord data about target behaviors or products and measure exposures. The manager might choose from his own BEfliAJfNDIIIAIUNIHGEEAEEWT OBSERVATION DEFINITIONS ROLLOUT I) Booth: (8) : Will be scored any time the rollout person is inside the spray booth. 2) Rollout area: (R) : Will be scored any time the rollout person is within the rollout area. The rollout area will be considered an area extending 10 feet from the forward edge of the spray booth. 3) Outside : (0) : Will be scored any time the rollout person is outside the rollout area (forward of the rollout area, not inside the spray booth). A) Working : (w) : Will be scored any time the rollout person is rolling Out a part, using his roller, a brush, or his hands. During pauses of 60 seconds or less, the rollout person's behavior will continue to be scored, unless the rollout person signals that he is stopping rollout verbally or by turning off his air pump. 5) Not working : (N) : Will be Scored any time the rollout person is not working, up to 60 seconds. After 60 seconds, data will not be scored until the rollout person resumes rollout. EXCEPTION: As long as the rollout person is inside the spray booth, data will be taken, whether the rollout person is working or not. 6) Downwind rollout: (D) : Will be scored any time the rollout person's face is oriented more away from the ventilation than toward the ventilation while rolling out a part. Face orientation directly perpendicular to the ventilation will not be scored as downwind rollout. If in doubt as to whether face orientation is perpendicular or downwind, consider it to be downwind. 7) Uncovered skin: (U) : Will be scored if any part of the rollout person's body below the neck is not covered by clothing while he is rolling out a part, or if the chest is exposed more than six inches below the chin. 8) Head in mold: (H) : Will be scored any time a rollout person's head is placed into a female mold beyond the chin while rolling out. 9) Exposure to spray: (E) : Will be scored any time a rollout person is in front of the spray gun while it is operating inside the spray booth. Any position forward of a line drawn perpendicular to the nozzle of the spray gun will be considered to be in front of the spray gun. 10) Turning mold: (T) : will be scored whenever a mold is turned at least 900 by a rollout person, other than when parts are moved into or out of the spray booth. ll) Respirator on: (R) will be scored any time a rollout person is wearing a respirator. A particle mask will not be scored as a respirator. l2) Fan: (F) : Will be scored whenever a fan is placed forward of the rollout area, directed toward the ventilation in the spray booth. A fan must be within 20 feet of the rollout area to be scored. If more than one Fan are operating in the above manner, write the number of such fans above the symbol in the first interval of each minute. Figure 6-2. An example of definitions observers can use to record data on the chosen behaviors or products. Exposure Control: Training and Follow Up personnel such as supervisors and foremen. However, if these people do not have experience or are not familiar with recording data about behaviors, the manager might find it helpful to Visit with a consultant who does have such experience. Peo- ple not currently working in the or— ganization can be hired and trained to record the data needed. A manager should always remember that whoever is chosen must have the available time and be capable of recording ac- curate, reliable data. observations of the behaviors and pro— ducts, observers should probably use an interval form because they can then accurately record if, when, and how long a particular behavior or product occurs. Remember from Lesson 3 that an interval form divides an observation period into minutes. The minutes are then divided into shorter intervals, generally 15, 20, or 30 seconds in length. Interval forms should generally also include the following: a list of the behaviors and products involved and correspond— ing symbols which will be used to re- cord an occurrence; the date; the time when the observation begins; the observer; the worker being observed; and the job being performed. In the pilot study for exposure control, symbols for all the selected behav— iors and products were placed in each interval. This practice makes re— cording data easier because the ob— server only has to mark the appro— priate symbol whenever the corre— sponding behavior or product occurs. A manager should also make certain that a form is designed for recording data about exposures. While conducting exposure related Observers should always practice re— cording the occurrence of the chosen behaviors or products before any data are actually collected. The follow— ing procedure is probably one of the best methods to make sure observers are capable of recording data on which a manager can rely: First - all observers should watch the same worker and record data as a group, frequently checking to see if they all record the same data and subsequently settling any differences 6—9 of opinion. Second — when it appears that the ob— servers agree consistently, they should all watch a particular worker and record data separately for a period of time without comparing data until the observation period is over. Third — the data each observer re— corded should be compared,and again all differences should be worked out. These practices should continue until a confident level of agreement is reached. Practices for recording ex— posure data may not be necessary be— cause the equipment used will gener— ally provide readings which only need to be written on a pre—designed form. The next step is to obtain baseline data. A manager should decide for what period of time to collect base- line. In the exposure control study at Labconcoi the research group de— cided to collect baseline on one group of workers for three weeks. As was true in the study,and as a gener— al rule, baseline data should provide a basis to assess present situations and to eventually determine the amount of improvement made after an intervention is implemented. INTERVENTION After baseline data is obtained, the chosen intervention efforts should commence. Any arranged consequences should now be appropriately applied. In the pilot study for exposure con— trol, the intervention was begun by training workers how to perform the behaviors that would reduce exposures to styrene. The workers were then verbally reinforced (arranged conse- quence) when an appropriate behavior was observed. Generally, managers should make sure that the employees behaviors or products of their behav- iors which reduce exposures are strongly reinforced. Additional data should be obtained during interven- tion in order to determine if there are any changes in target behaviors or products and exposures. TRANSLATING THE RESULTS Often it is imperative that managers BEHAVIOR MANAGEMENT :22 u “5'3 : l) 2 m t:- >\ so * NI“) + 3': ’3‘% UN 70 180 q. (.0 L“ ”66 Q So o-E 41, ‘3. g Bo ECHO 0‘ 2.9 , Q~ IO (2%20 I i; I L; I 7 CL. ' K I l I f X ‘1 I + s '7 0" IL 3 -/ 5 a 7 (P Don/5 0 pe rcen‘l‘ 0+- In+erValS (A I? 5 E Cky l 30 30 .56 f5 2 if #5 e3 cm 3 r77 93 2/ 9G 4 £3 36 57! 3’7 5 7C: 40 512 7‘7 Figure 6-3. An example of a table and graphs that might be used to illustrate collected data, gas: line « Training anal Vef‘liml FCt’Ilél-Ick I00“ 5K ¢-<) 0 \ PCPCWJL 01C In‘lflvali O‘F Defu‘ed geliaWar‘f I l l l w— l l 40‘ I | 1‘ l 20 I I 1 liiltlrilllri‘r.1 [2 3 ¥f67j>qmumnnhrlo Day; Figure 6-4. A graph illustrating the effectiveness of training and verbal reinforcement. 6-10 Exposure Control: Training and Follow Up 6a Sell'nc In+€f U (“+7 6'1 /. ; \ 5‘: m 00 2' O 0 I l I I \4 D | I I I I I I I I I I I I I I Percen+ 0+: In‘I‘Prva/S o4: Off/re) ficfiauiar; I I I 1 fl 1 I I - I 11‘3‘45‘6 736/0III213/1/r1b anf , l 54 Sf /Inf I p Inc — [KI-’1‘“? v rev/Io n w qd‘ I z 490‘ I ‘* \/\ l k 70‘ )\ \f- 60' I vx 50' l 96 #0” l r 30‘ I \ - — I Q. 10 I Q— 10 I l t I ._ 5 . ' r r I l I I r ( I I 2. 3 ‘I 5 4 7 J ‘1 lo I! /1. I} [‘1‘ I)‘/(. 00.7} Figure 6-5. Two graphs illustrating the relationship between an increase in desired behaviors and a decrease in exposure to styrene. BEHAVKHIMANAGEMENT begin translating data immediately after the intervention is implemented in order to apply the appropriate re— inforcers or punishers. However, a manager may decide to begin trans— lating data after it has been col— lected for a specific period of time after intervention. Data recorded for exposure levels will probably need very little translation. Trans— lation of data recorded during obser- vations should begin by counting the occurrences of the chosen behaviors or products. After the number of occurrences and intervals have been tallied, percent of intervals figures can be computed for each behavior or product as follows: number of occurrences total number of intervals Remember from Lesson 4 that the dura— tion of the occurrence of certain be— haviors or products can also be de— rived from data recorded on an interval form. In the pilot study, a particular product of behavior for which duration was computed was that of uncovered skin (u), because it was important for the researchers to learn how long workers left their skin exposed to styrene. After the necessary computations have been per- formed, the data collected can often be illustrated using tables and/or graphs. For example, consider the tables and graphs in Figure 6—3. EVALUATING THE EXPOSURE CONTROL PROGRAM After performing all the needed com— putations and illustrations of the data, a manager should evaluate the exposure control program. He might begin by evaluating the effects of the arranged consequences on the tar— get behaviors and products. Consider the graph in Figure 6—4. A manager could confidently conclude from this graph that the training and the fol- low-up procedure of verbal reinforce- ment was effective to increase the rate of desired behaviors. Another evaluation should be to decide if any changes in the chosen behaviors or products was related to any changes in exposures. Consider the graphs in Figure 6—5. A manager could safely assume that increasing the rate of these particular behaviors caused a decrease in exposure to styrene. After analyzing the graphs and tables, a manager must decide if the chosen intervention should be con— tinued, or possibly supplemented if it was only partially effective to make the needed improvements. A man— ager should also determine if any ad— justments are needed in the data sys— tem itself. After data have been collected and results translated, a manager might decide to: conduct ob_ servations more (or less) frequently; shorten or lengthen the intervals; add or delete certain behaviors or products; redefine a particular be— havior or product. SUMMARY COMMENTS As demonstrated in the pilot study conducted in the Labconco Fiberglass Plantl, exposures to chemical sub— stances used in an industrial organ— ization can often be controlled by altering workers' behaviors. Expo— sure control programs focus on cor- recting those work practices or con— ditions that are detrimental to the health of employees. The changes that need to be accomplished by an exposure control program are reflect ed in the targets and goals set by management. If a data system is ad- equately set up to measure both the changes in exposures to a particular substance, and the changes in behav- iors presumed to be related to expo— sure control, the relationship be- tween the two can be determined as well as the effectiveness of any in— tervention efforts. By developing and implementing an appropriate expo- sure control program, an important benefit to be gained is the improve— ment of work practices and conditions that are vital to the health of em- ployees. This factor alone can jus— tify the costs involved in conducting an exposure control program. Exposure Control: Training and Follow Up REVIEW Following are the questions a manager should consider when setting up and imple— menting an exposure control program: 1. 2. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Have potentially harmful substances used in work operations been identified? Have the jobs where exposures are a crucial problem been identified? Has a goal been set to improve the health conditions related to exposures? Have terminal targets been stated to reflect the specific changes desired in exposures? Have prerequisite targets been stated to reflect the specific changes needed in work practices? Have all behaviors or products that need to be measured been identified and specifically defined? Have techniques for measuring exposures been appropriately set up? ) Has it been specified in which work areas observations and exposure measure— ments will be taken? Has it been decided how often observations will be conducted and exposures measured? Have people been chosen to record data? Have forms which observers can use to record data about exposures and em- ployees' behaviors been developed? Have all observers practiced recording data until a confident level of agreement and accuracy was reached? Have baseline data been obtained for a pre-selected period of time? Have intervention efforts been initiated after baseline data were collected? Have data been obtained during intervention? Have the necessary statistics been computed from the data? Have these statistics been transformed into the tables or graphs needed to illustrate the collected data? Have all arranged consequences that were applied during intervention been evaluated to determine if they should be continued? If the results show that there is still room for improvement, have supple- mentary efforts been considered? Have needed adjustments of the data system been made? BEHAVIOR MANAGEMENT REFERENCES Conard, R.J., Hopkins, B.L., Fitch, H.G., Smith, M.J., Anger, W.K., and Dangel, R.F. A strategy to validate work practices: An application to the reinforced plastics industry. Submitted for publication; Journal of Occupational Medicine, 1979. LESSON SEVEN: SAFETY PERFORMANCE CONTROL — GRAPHIC FEEDBACK CONTENTS Page SAFETY PERFORMANCE CONTROL IN A FOOD MANUFACTURING PLANT ........ 7—3 SETTING UP A SAFETY BEHAVIOR REINFORCEMENT PROGRAM ................ 7—5 DEVELOPING THE OBJECTIVES ............................................. 7—5 DEVELOPING A DATA SYSTEM ............................................. 7—6 iNTERVENTION: FEEDBACK GRAPHS ....................................... 7—7 TRANSLATING THE RESULTS .............................................. 7—8 EVALUATING THE SAFETY PROGRAM ...................................... 7—8 SUMMARY COMMENTS .................................................... 7—10 SAFETYPERFORMANCECONTROL:GRAPHRJFEEDBACK INTRODUCTION It has become more evident in recent years that the cause of most acci- dents is due to human error. Indus— trial organizations have, therefore, become interested in setting up safe— ty programs that deal with the safe— ty—related behaviors of workers. These programs usually focus on im— proving work practices of employees, in an effort to solve problems re— lated to poor safety performance. Many safety programs have relied heavily on the method of training em— ployees how to behave safely in rou— tine circumstances, and how to react in emergency situations. Though training is an important componant in a safety program, the positive ef— fects of training can wear off if an adequate follow—up procedure is not developed and properly carried out. Rather than relying on the method of punishing workers for unsafe prac— tices, many programs have taken a more positive approach by designing fol— low—up procedures whereby workers are rewarded for improving their safety performance. Lesson 7 presents a summary of a study in which graphs were provided as positive feedback to workers for achieving and maintaining a desirable level of safety perfor— mance. Following the summary is a description of the steps necessary to set up a safety program that will involve similar methods to directly, or in some cases indirectly, influ- ence employees to behave safely at work. SAFETY PERFORMANCE CONTROL IN A FOOD MANUFACTURING PLANT A study was conducted by a research team to assess the effectiveness of positive reinforcement to improve safety performance in a wholesale bakery.1 The research team was con— tacted because plant personnel had become concerned about a dramatic in- crease in the plant's injury frequency rate and a corresponding rise in workmen's compensation premiums. The researchers accepted the challenge to help the company set up a safety pro— gram that would solve this problem. Two departments, make—up and wrap— ping, were identified as the major concern due to the high proportion of accidents that had occurred in these areas during the previous year. The goal of the management was to set up a positive safety program that would improve the safety performance of employees. The terminal target was to reduce the injury frequency rates and premiums for workmen's com— pensation. A prerequisite target, presumed to be an important influence on the terminal target and goal, was specified as the increase in the rate of safe behavior in the two selected departments. A data system was then set up so that information, which could be used to determine if the prerequisite target was achieved, could be obtained. The first course of action was to identi— fy the safe and unsafe behaviors or products about which data needed to be collected. Past accident records were reviewed to determine which un— safe actions by workers had resulted in accidents. Also, suggestions from supervisors were incorporated to help identify any other behaviors or pro— ducts which might need to be measured The behaviors and products selected were then specifically defined so that observers would be able to re— cognize them consistently. It was easily determined that the work areas in the make—up and wrap— ping departments were where observa— tions should be conducted to record data. Observations were scheduled to be carried out on four days of the work week, at random times during each of those days. After observers were chosen, they were trained to re— cognize the chosen behaviors or pro— ducts. A form was designed for re— cording data. Each observer was BEHAVKHiMANAGEMENT instructed to use the list of the chosen behaviors and products (in— cluded on the form) to record these items as either safe, unsafe, or not observed. The observers then toured the designated work areas and prac— ticed recording data until a satis— factory level of accuracy and agree— ment was reached. Baseline data on the rate of occurrence of the chosen items were then obtained for both of the departments. After baseline had been collected for an adequate period of time for each department, the in— tervention was implemented The beginning of the intervention consisted of a training component, whereby employees attended a thirty minute session devoted to educating them about safe and unsafe practices. A slide presentation of workers in action was given. After viewing an example of an unsafe practice, em— ployees were asked to describe why it was incorrect. The same incident was then shown performed in a safe manner. At the end of the presenta— tion, conduct rules were reviewed, and workers were shown a graph, (sim— ilar to the one in Figure 7-1) that illustrated their present (baseline) safety performance. After the training was completed, the follow-up procedure, notably the most important component of the interven— tion, was begun. Graphs and a list of safety reminders (as described above) were posted in a conspicuous H o . “3.x ”1 Bafflm€ 36: :4 ao- H .fi CL W 90‘ o i L 0 'fi— L+ $0 ‘ t L 14 w tk m d Q. 20 place in each department. The graphs served as feedback to the workers on their safety performance. The re- sults on each graph was, of course, slightly different for each depart— ment. As additional data were ob— tained during the intervention condi— tion, observers frequently updated the graphs to give the workers con— tinuous feedback on any changes in their safety performance. Another arranged component of the in- tervention was for supervisors to make specific comments when they ob— served a worker performing safely. These supervisors were instructed to keep records on how often they gave positive comments to workers. Ob— servers continued to record data on all pertinent safe and unsafe inci— dents for several weeks during inter- vention. The data collected had to be immedi— ately translated in order to contin— uously update the feedback graphs. The number of incidents performed safely and the total number of inci— dents observed were counted, enabling the percent of incidents performed safely to be computed. The resulting figures were then plotted on graphs with the percent of incidents per— formed safely on the vertical axis, and number of observation sessions on the horizontal axis, as previously illustrated. The results obtained indicated very l I l l x I l | I Z 3 '7’ r c 7 3 Observation Sessmr/s Figure 7-1. An illustration of baseline safety performance. 7-4 Safety Performance Control: positive effects of the follow—up procedure. The percent of incidents performed safely in the make—up de- partment increased from a baseline mean of 77% to an intervention mean of 99%, and in the wrapping depart— ment, the mean of 70% during baseline improved to 96% during the interven- tion condition. To insure that it was the feedback graphs, and not just the presence of observers, that caused workers to im- prove their safety performance, a re— versal phase was instituted. During this phase, the previous graphs and list of safety reminders remained in each department, but the observers no longer updated the feedback graphs. Data collected during this reversal phase indicated that the performance of workers in the wrapping depart— ments returned to the baseline mean of 70%, and those in the make—up de— partment fell to a mean of 72%. Based on the results obtained during intervention, the research team and the plant personnel concluded that the posted feedback graphs had been effective as an arranged consequence to improve safety performance. In addition, the results obtained during the reversal phase further convinced the research team that the change in safety performance was due to the feedback graphs — not just a reaction to the presence of observers. The safety program, combining the train— ing, the feedback graphs, and super— visors comments, substantially con— tributed to the achievement of the targets and goal, therefore solving the problem. SETTING UP A SAFETY BEHAVIOR REINFORCEMENT PROGRAM To begin setting up a program to deal with behaviors related to safety, such as that employed by Komakil, the areas or situations where safety per— formance needs to be improved must first be determined. Several sources of information can be utilized to identify problems related to the ex— istence of unsafe behaviors. Safety and health statistics are one source which can reflect poor safety perfor— mance by employees. Another source is any employee in the organization — from line workers to company execu— Graphic Feedback tives. These people may recognize the existence of safety problems and suggest to a manager that improve— ments are needed. During their everyday activities, supervisors and foremen may come into contact with practices and situations that are detrimental to the safety of their workers. If management believes (or is led to believe) that corrective actions are needed, they should spe— cifically define any problems, avoiding using vague descriptions. If an organization employs a large number of people and/or consist of several departments, management may find it more practical to focus their initial efforts towards those depart— ments or groups of workers for which safety problems are the most evident and critical. DEVELOPING THE OBJECTIVES After the need for safety improve— ments has been accurately identified, management should specify appropriate goals. Goals should be stated to re— flect the general changes necessary to improve safety performance. A common goal for all organizations concerned with safety should be to set up a positive program which will encourage workers to behave safely. In addition to setting overall goals, management may also need to set goals for individual departments within the organization. Targets should reflect the more spe— cific changes that need to be accom— plished by the safety program. As a general rule, terminal targets should be set to improve any statistics that are related to safety performance. For example, Komaki set a terminal target to reduce injury frequency rates and premiums for workmen's com— pensation. Any prerequisite targets set should deal more directly with safe or unsafe behaviors. One such target that is very important in a safety program is to increase the frequency of safe behavior by workers which means a corresponding decrease in the occurrence of unsafe behavior. For example, increasing the frequency of proper lifting behaviors should decrease the frequency of improper lifting. Komaki set a prerequisite target to increase the observed BEHAVIOR MANAGEMENT occurrence of the safe behaviors and products (resulting from safe behav— ior) that were presumed to be an im— portant influence on the terminal target and goal. If a manager sets a terminal target to decrease the injury frequency and severity rates, his prerequisite targets should in— volve increasing the occurrence of the safe behaviors that will reduce the chances of an accident, and de— creasing those unsafe behaviors that might lead to the occurrence of an accident. DEVELOPING A DATA SYSTEM Following the specification of goals and targets, management should set up a data system that will measure changes in the safety—related behav— iors or products that are the focus of the targets and goals. Most or- ganizations already record data about routine safety statistics that might be involved in terminal targets. Therefore, a measurement procedure may only need to be developed to measure changes in the behaviors or products which are involved in pre— requisite targets. If a manager has set a prerequisite target to in- crease the percent of incidents per— formed safely, he must develop a measurement procedure, in which all possible safe and unsafe behaviors or products (related to his target) are identified. He may find it useful to consult plant personnel in supervi— sory positions to aid in the identi- fication of the safe and unsafe be- haviors and products that presently exist. The supervisor in charge of a certain group of workers may be the best authority for identifying the safe and unsafe behaviors that are being performed by those workers. More often than not, accidents are caused by unsafe acts of workers. Therefore, past accident records should also be reviewed; they may contain information about the parti- cular behavior or unsafe act of a worker that resulted in an accident. After adequate research and inquir— ies, management should select those specific behaviors or products on which they want to obtain data. In certain circumstances, it might be OPERATIONAL DEFINITIONS Ladder — recorded as safe whenever an employee is using the right size and type of ladder. — recorded as unsafe whenever an employee is using a box or Machine Guard — chair in place of a ladder. recorded as safe if guard is in proper place. — recorded as unsafe if original position has been altered or guard has been completely removed. Lifting recorded as safe when an employee bends knees to such a degree that both leg and back muscles are used. — recorded as unsafe when an employee bends straight overlagdfiafiis extra strain on back muscles. 1-w+ In: Cutting action away from self. recorded as safe if employee is directing cutting — recorded as unsafe if employee is directing cutting action towards self. Figure 7-2. Operational definitions of possible chosen behaviors and products. Safety Performance Control: best to measure a product rather than the behavior(s) that produced it. For example, a manager decides to measure whether the guard on a par— ticular machine has been removed (product removed guard) instead of the behavior of removing the guard. His decision is based on the fact that an observer may not be present when the guard is actually removed. the selection process, op— definitions of the chosen or products should be de— veloped. These definitions should be specific, and should be supplemented with instructions on how to record data on the particular behavior or product. (See Figure 7—2.) Following erational behaviors The departments, or groups of workers which are the focus of targets will have already been identified, either when a need for improvement was re— cognized or when the targets and goals were set. Therefore, whoever is in charge of developing the data system should have little trouble specifying the work areas where ob- servations will take place. A sched- ule will need to be designed in such a way that observations are carried out at random times to obtain a re— presentative sample of the regular safety performance. Observers will need to be selected and can often be chosen from plant personnel. After measurement procedures have been developed, management should design an interval form or checklist that can be used to record data about the safe or unsafe incidents observed A manager must design an interval form, described in Lesson 4, if he wants the data to reflect the length of time a particular behavior product occurs. Komaki designed a form which included a list of the behaviors or products that observers could use to record each of these items as being either safe, unsafe, or not observed. To begin the implementation of the data system, baseline data must be obtained. Baseline on routine safety statistics can usually be obtained by consulting past company records. In most cases, a manager should specify a period of time for which baseline will need to be obtained. Before anv baseline data are collected for pre— Graphic Feedback requisite targets, observers should be trained to recognize each safety— related behavior or product. They should then be required to practice recording data until a sufficient level of accuracy and agreement is reached. Baseline data on safety— related behaviors or products should always reflect the safety performance of employees before any intervention efforts are implemented. INTERVENTION After baseline has been obtained, the intervention efforts of the safety program should begin. If needed, training sessions similar to the ones in the Komaki studyl should be con— ducted to educate workers about safe and unsafe practices. It is essen— tial that management explain to work— ers what makes a particular work practice unsafe, and demonstrate how to perform safely in that circum— stance. Workers should be encouraged to participate in these demonstra— tions and contribute information about safe or unsafe practices that are overlooked. This training component can be accom— panied by a follow—up procedure de— signed to reinforce workers for prac— ticing the desired safe behaviors taught during the training session. A follow—up procedure which was suc— cessful for Komaki involved posting a graph displaying the safety perfor— mance of a selected group of workers. The graph should be posted in a place where the workers involved will be able to view it frequently. When a graph such as this is initially post— ed, only the baseline data should be plotted. The graph should be updated as additional data is obtained. Re— member that data collected after the intervention has been initiated will eventually enable management to judge the effectiveness of the safety pro— gram. The practice of updating the graph enables workers to see any changes (hopefully improvements) in their safety performance. This type of graph can function as an arranged consequence (in the form of feedback) to workers on their safe or unsafe behaviors. Additional graphs, dis- playing other safety statistics, can also be posted. BEHAVIOR MANAGEMENT Another arranged consequence, which has the potential to be effective for improving safety performance, in— volves immediate verbal feedback. To carry out this practice, supervisors or foremen can present positive com— ments (especially specific ones) to a worker immediately following the ob— servation of a safe behavior or pro— duct. Management should realize that supervisors and foremen may need to be instructed on how to give sincere, effective, positive comments. TRANSLATING THE RESULTS If feedback graphs are presented dur— ing the intervention condition, data will need to be translated immediate- ly after they are obtained. All com- pleted observation forms, which have been used to record data about the safety performance of employees, should be collected, and the recorded occurrence of safe or unsafe inci— dents should be counted. The totals obtained from this counting process can then be utilized to make the nec- cessary computations for evaluating the program. In the Komaki studyl, the number of incidents observed were counted. The percent of inci- dents performed safely was then com— puted in the following way: % of incidents performed safely = number of safe incidents observed total number of incidents observed An example of data transformed in this manner is given in Figure 7—3. If an interval form has been used to record data, the percent of intervals in which a behavior or product oc- curred, and respective duration fig- ures, can be computed (refer to Les— son 4 under Translating Data). After the necessary computations have been performed, the data should be graphically displayed. Consider the following data: % of Incidents Performed Safely Baseline 1 ....... 6O Observation 2 ....... 63 3 ....... 72 Sessions 4 ....... 68 5 ....... 65 % of Incidents Performed Safely Intervention 6 ....... 75 Observation 7 ....... 82 8 ....... 79 Sessions 9 ....... 86 10 ....... 90 These figures can be plotted on a graph as shown in Figure 7—4. Such visual depictions of data are often more easily understood by employees than the presentation of a list of numbers. If data have been obtained on any other items or statistics related to safety performance. Similar computa— tions will need to be performed, and, for clarity, the resulting figures should be displayed in tables or graphs. Due to the number and vari— ety of safety—related statistics re— corded in industrial organizations, the precise computations that may be needed are beyond the scope of this lesson. EVALUATING THE SAFETY PROGRAM The first steps for evaluating a safety program designed to reinforce safe behavior should be directed to— ward determining if the methods ap— plied during intervention were suc— cessful to achieve the targets and goals. For most programs, this will involve deciding if the practices of training workers and providing feed— back graphs were effective to make the needed improvement in the safety performance of the chosen workers. Consider the graph in Figure 7-4. Based on the information provided by the graph, it would be concluded that the training and feedback practices were successful to improve safety performance, and should therefore, be continued. In addition, it would be decided that the feedback graph served as an effective consequence to encourage workers to behave safely. This evaluation practice should be carried out for any other arranged consequences that were applied during the intervention condition. If any of the graphs illustrate that there is still room for improvement, sup— plementary efforts should be Safety Performance Control: Graphic Feedback # of Safe Incidents Total # of Incidents Observed Observation —_—___—_ Session 1 12 20 2 16 25 3 18 21 4 14 19 5 13 18 % of Incidents Performed Safely Observation Session 1 12/20 = 60% 2 16/25 = 64% 3 18/21 = 86% 4 14/19 = 74% 5 13/18 = 72% fiwm73.WmflwmumflfiflmmhmhMonmkprhmmma L‘ * l é >\ Baseline . Irw+€rver1fion “ “ ”0 E ‘¥ #0 :r”’h~“(///’,d H v: 30 I 70 r”’t///‘\\\“\‘* l at .5 Go I ; §o : 1- Z to I 5 Lt' 30 I 2 c: 10 I Q] I Q... 10 I C t I l ( Y I i r ”r— I 2 3 $ 5' e 7 a7 4 la OLSt’Pua‘hm j‘PSI/ony Figure 7-4. A graph on which translated data have been plotted. BEHAVIOR MANAGEMENT considered. Remember that if an arranged consequence is ineffective, or produces undesirable results, it should be discontinued. In any control program, it is impor— tant that the sufficiency of the data system(s) used is regularly evaluated. Management may want to determine if the measurement procedures are ade— quate, if observers still appear to be qualified and capable of recording accurate data, and if more, or less, data should be collected. Another factor which must be considered is the cost of the program being imple— mented. As emphasized in previous lessons, the benefits gained from a safety program should be weighed against the costs. The costs of a program such as the one described in this lesson would be minimal, con— sidering that it is based on graphic feedback reinforcement - an inexpen— sive intervention. A program de— signed to maximize safe work behav— iors and conditions through the use of graphic feedback can result in a substantial increase in the safety performance of employees. SUMMARY COMMENTS As discovered by Komakil, a positive reinforcement approach is important for solving safety-related problems. After the appropriate goals and tar- 7-10 gets of a safety program are speci- fied, a data system can be developed and carried out to obtain baseline data on the safety performance of employees. If these data indicate the need for workers to be educated about particular safe or unsafe prac- tices, a training program can be implemented. Once it has been ascer— tained that workers are aware of proper safety practices, interven— tion procedures must be implemented. The practice of using graphic feed- back as a reinforcement tool, (exem- plified by Komakil) proved to be a successful method for improving the safety performance of employees. The need for, and success of, such a follow—up procedure can be measured by an appropriately developed data system. After data have been col- lected and translated, prior to, during, and after intervention, the effectiveness of applying arranged consequences, such as feedback graphs, can be evaluated to deter— mine if those consequences induced workers to increase their rate of safe behavior. A feedback graph is generally an effective consequence which is not only inexpensive, but also easily carried out by plant personnel. If positive methods are applied, and negative sanctions avoided, employees will react favor— ably to this kind of safety program. Safety Performance Control: Graphic Feedback REVIEW Following are the questions that should be considered when setting up and imple— menting a safety program: 1. 10. ll. 12. 13. 14. 15. 16. l7. 18. 19. 20. Have all significant problems related to poor safety performance been iden— tified? Has a goal been set to develop a positive reinforcement program that will solve those problems? Have terminal targets been stated to reflect the improvements needed in safety statistics? Has a prerequisite target been set to increase the occurrence of safe behav— iors? Have all safety-related behaviors or products, on which data need to be ob— tained, been identified and defined? Have work areas where observations will take place been specified? Have observations been scheduled to obtain a representative sample of the employees' safety performance? Have observers been selected and trained to recognize the chosen behaviors and products? Have forms, on which observers can record data on safety performance, been designed? Have observers completed a practice session to assure consistency in data collection? Have baseline data on the safety performance of employees been obtained? Has it been determined if the employees are aware of the safe behaviors they should be practicing? If employees are not aware of appropriate behaviors, have training sessions been conducted? Have follow—up procedures been developed and initiated following training sessions? Have data been obtained and translated prior to, during, and after interven— tion? Have feedback graphs been frequently updated? Have the results obtained been transformed into tables or graphs, so that the safety program can be evaluated? Has it been determined whether arranged consequences and other interventions are effective and should be continued? Have supplementary efforts been considered to produce further improvement in the safety performance of employees? Have all other needed adjustments orchanges in the safety program been made? 7-11 BEHAVIOR MANAGEMENT REFERENCES Komaki, J., Barwick, K.D., Scott, R.L. A behavioral approach to occupa— tional safety: pinpointing and reinforcing safe performance in a manu— facturing plant. Journal of Applied Psychology, 1978, §§. LESSON EIGHT: SAFETY AND HEALTH LOSS CONTROL -— INCENTIVE PROGRAMS CONTENTS Page SAFETY AND HEALTH LOSS CONTROL AT AN OPEN PIT MINE .................. 8—3 SETTING UP AN INCENTIVE PROGRAM ...................................... 8—6 DEVELOPING THE OBJECTIVES ............................................. 8—6 DEVELOPING A DATA SYSTEM ............................................. 8—6 INTERVENTION: TANGIBLE REINFORCEMENT ............................... 8—7 TRANSLATING THE RESULTS .............................................. 8—11 EVALUATING THE INCENTIVE PROGRAM .................................... 8—11 SUMMARY COMMENTS .................................................... 8—11 SAFETY AND HEALTH LOSS CONTROL: INCENTIVE PROGRAMS INTRODUCTION Training programs have been a suc— cessful method to teach workers the proper behaviors for improved safety. Though workers learn these behaviors through such programs, they often do not practice them routinely. It is, therefore, often necessary to devel— ope a supplementary program to influ— ence employees to practice the de- sired safety behaviors. It has been illustrated that reinforcement plays a determining role in the success of safety programs. As exemplified in Chapters 5 and 7, written and graphic feedback can be successful in influ— encing particular behaviors. Chapter 6 demonstrated the importance of ver— bal feedback. There is yet another method by which desired safety behav- iors may be induced. This involves the use of tangible reinforcers as incentives. If managed carefully,the delivery of tangible reinforcers can provide workers with the incentive needed to promote and maintain safe and healthful work practices. The success of such a program is illus— trated in the following study con— ducted at a Wyoming mine. SAFETY AND HEALTH LOSS CONTROL AT AN OPEN PIT MINE David Fox faced what must have seemed to be an almost impossible task. The open pit Lucky Mc Mine at which he worked was a very unsafe place. It was his responsibility as Safety and Health Engineer to do something about the problem. Some details will give a picture of the enormity of the problem. The mine site is about sixty miles out of Casper, Wyoming on a high plane. Winter conditions can be tough with temperatures as low as —40° F and constant winds which can push the chill factor twenty to forty degrees lower. The winters are long and snow depths of two feet are common. The mine is an enormous hole about a mile wide and up to 400 feet deep. Huge shovels and lZO—ton trucks are used to remove the overburden. Scrapers and bulldozers do the finer removel work. Finally, small shovels and dump trucks remove the ore. All of the overburden and ore hauling is done Via gravel roads which spiral up the sides of the mine pit. During the summer, the work condi— tions are difficult. During the win— ter they can become almost impossi- ble. Imagine that you drive one of the 120—ton overburden trucks. You pull up beside a shovel that loads your truck with 15 cubic—yard scoops. You spiral up the icy road, having to stop occasionally at intersections for other equipment coming in on lat— eral roads. Once you reach the rim of the mine, you drive over more slick roads for two miles to the dump sites. Then you return to the shov— el for another load. Snow is blowing over the entire route and dozers are constantly operating to keep the roads open. The members of the work crew of which you are a part are widely scattered, and you see them only as you pass on the narrow roads. Difficult conditions exist even in the maintenance shop. A 120—ton truck represents an enormous invest— ment that has to be kept in operation as much of the time as possible. Maintenance will often be done while diesel engines continue to run. A misstep in the cab of such a large rig can send a person sailing to the shop floor far below. When Fox first faced this problem, the mine's severity rate was running from 980 to 1,300, the frequency rate from 35 to 42.5, and the number of compensation cases from 73 to 104 per year. The costs of damages to vehi— cles were averaging $255,000. per year and compensation costs $22,000. per year. The mine work—force was 'BEHAVIOR MANAGEMENT gradually expanding and prospects were that safety problems would only become worse. The manager asked Fox to do something — quickly. Fox first set up training programs to teach supervisors about safety regulations, accident investigations and hazard prevention planning. Then supervisors were helped to be— gin mandatory, weekly, “tool—box” (on the job—site) safety meetings in which supervisors and workers would discuss hazards, possible im— provements, and ways in which work— ers could do their work more safely. Now it could be assumed that the miners knew how to behave safely. However, there continued to be many accidents and injuries. Fox con— cluded that his miners simply were not sufficiently motivated to behave safely. Realizing that an incentive program was needed, Fox set goals to reduce accidents and injuries, and the cost associated with them. He did not set specific numerical or quantita— tive targets. However, it was clear that his com— pany expected substantial reductions in all major measures of injuries and accidents and the costs of these injuries and accidents. The company was already keeping data records on the number of hours worked by all employees, the total number of injuries reported, the number of injuries given only first— aid treatment, the number of cases involving compensation, the number of lost—time injuries, the number of days lost, the frequency and sever— ity rates, the number of accidents involving equipment and vehicles, costs of compensation, and equipment and vehicle damage costs. The data systems Fox needed were therefore, already operational. Using the information provided by these data systems, Fox set up an in- centive (reinforcement) program for his mine. First, be defined a work unit to be all of those employees re— porting to a particular supervisor. Next, Fox set up what he called haz— ard groups. He analyzed the com— pany's safety records to determine how dangerous, on a per worker — per hour basis, each of the various job classifications had been. Hazard Group I included administrative, secretarial and clerical staff. Group II was made up of foremen, shift supervisors, lab technicians, field engineers and surveyors. Group III included bulldozer, truck, and shovel operators, mechanics, laborers and maintenance people. Scraper operators, fuel and lube people and electricians comprised Hazard Group IV. Fox selected trading stamps as the reinforcers he would use in his in- centive program. The stamps could be redeemed at a center to which all employees had access. A part of his rationale in using trading stamps was that they could be exchanged for a very large variety of reinforcers. Employees could select those items that would be particularly reinforc— ing for them. Each individual who had not experi— enced injuries or caused vehicle or equipment damages for a month, re— ceived a set number of trading stamps — with that number determined according to their hazard group mem— bership: Group I — 300 stamps Group II — 400 stamps Group III — 500 stamps Group IV - 700 stamps In addition, each individual in a work unit received bonus stamps each month if every member of that unit had avoided having accidents and in— juries. Bonus stamps were given as shown below: Group I — 200 stamps Group II — 300 stamps Group III — 500 stamps Group IV — 700 stamps These bonuses were, of course, de— signed to encourage people to look Safety and Health Loss Control: Incentive Programs out for the welfare of other members in their particular work unit. In addition, it was expected that such an incentive would increase the social presures applied by members of a work unit to behave safely. Fox hoped that the families of the workers would also provide social reinforcement. Because the stamps could be traded for merchandise which might benefit all family members, their interest in the workers‘safe behavior should be heightened. Special bonuses, award bonuses, of various denominations were also pre— sented for several additional kinds of safety efforts. These included: safety suggestions that were adopted by the mine; acts that prevented serious injury or death of another worker; acts that prevented damage to company property; and other special efforts that were determined to be deserving of special reinforcement. Fox's program also included punishment (or at least specific penalties) for accidents and injuries. The punish— ment consisted of workers losing po- tential stamp awards. An individual who suffered an injury that caused a loss of time from work was penalized according to the number of days lost: Number of Stamps Lost Days Lost for 1 to 2 1 month 3 to 4 2 months 5 to 6 3 months 7 to 8 4 months 9 to 10 5 months 10 plus 6 months An employee would never lose his or her individual stamp award for more than 6 months,regardless of the num— ber of days lost from work as the re— sult of an injury. An employee who caused an accident resulting in property damage lost the individual stamp awards for a number of months — determined by dividing the costs of the damage by $2,000. Thus an employee who wrecked a pickup truck valued at $6,000. would loose the individual awards for 3 months. Again, the individual awards could be withheld for a maximum of only 6 months. Finally, any individual who knowingly failed to report an acci— dent or injury would lose the indi— vidual stamp award for 1 month. David Fox was fortunate in that he had accurate data prior to the im- plementation of his incentive pro— gram which utilized trading stamps as reinforcers. This allowed for a relatively clear determination of the effects of the program. Those effects were good, indicating a sub— stantial improvement of safety sta— tistics (see Figure 8—1). These statistics indicate that the reinforcement or incentive program was apparently effective to produce a great reduction in injuries and accidents. The value of reducing human suffering was incalculable. However, in monetary terms, the com— pany saved 9 to 10 dollars (in re— duced damage, compensation and medi— cal costs), for every dollar they invested in trading stamps. The costs of administering this kind of reinforcement program were negligi— ble; the reinforcers had to be de— livered only once a month, and these stamp awards were based on data the mine routinely collected. There are often doubts that a safety program that has been effective in one setting will work as well in a different setting when run by dif— ferent people. David Fox received an opportunity to determine if his safety program would work at other settings. The parent company that owned the Lucky Mc Mine also oper— ated a coal mine in New Mexico. Fox was asked to help the safety engi— neer at the coal mine to set up a reinforcement program similar to the one he implemented at his ore mine. The Navajo Coal Mine in New Mexico was clearly differnt than the Lucky Mc Mine; it was larger, the work force was quite dissimilar, and the climate in no way resembled that in Wyoming. However, the Navajo Mine was also plagued by many accidents and injuries, and the resulting high costs. The Safety Engineer at the Navajo Mine set up and managed a program modeled precisely after Fox's. Again, the effects were not immediate. However, within a year, BEHAVIOR MANAGEMENT Incentive Program Started Here '- 211332 Frequency Rate 35.0 19.5 Severity Rate 980 311 Days Lost Time 668 276 Total Damage Costs $220,000 $138,000 Figure 8-1. Mc Mine. all of the safety statistics had dropped to a fraction of what they had been before the incentive program was implemented. This clearly indi— cates the success of the incentive program. SETTING UP AN INCENTIVE PROGRAM A safety incentive program, such as the one developed by Fox, can be eas- ily set up and managed. The following will provide information about each of the basic steps that need to be taken. DEVELOPING THE OBJECTIVES The objectives of improved safety and reduced related costs, are inherent in an incentive program. Such a program is not aimed at an isolated problem or particular division of a plant, fac— tory or mine.Rather, it is designed to involve all employees in all divi- sions. The goals of an incentive pro— gram are,therefore,to reduce the num— ber of accidents and injuries, and the costs associated with them,through the provision of incentives and reinforce— ments. DEVELOPING A DATA SYSTEM Most companies will already have an adequate data collection system either as a result of private initiative, or because a regulatory agency such as OSHA requires that such data be kept. 193».wa 4.7 3.2 3.0 175 130 61 161 111 60 $45,000 $30,000 $40,000 The statistical effects of the incentive program at the Lucky This data should include information on monthly and cummulative yearly figures on the number of all inju— ries reported, the number of employ— ees given only first aid, the num- ber of cases for which medical care is required, the number of injuries requiring time lost from work, the number of days lost from work as a result of injuries, the number of injuries for which workmen's compen- sation is provided, compensation costs, number of instances of equip— ment damange, and costs of equipment damange. Such records usually also include information on the total number of hours worked by employees. One useful convention is to calcu— late what the total number of lost— time injuries would have been if there were one million hours of ex— posure per year (the frequency rate), and the total number of days lost given one million hours of exposure per year (the severity rate). These calculations allow for easy compar— isons among companies of different sizes. Statistics for industry aver- ages usually employ such a conven— tion. There may be additional infor— mation that would be useful in each particular setting. All data should be brought together in one location. They must be regu— larly up-dated just before reinforce- ment is to be given to the employees. The data will be used as the basis Safety and Health Loss Control: Incentive Programs for delivering reinforcers, and to monitor the effectiveness of the pro— gram. It is useful to maintain graphic dis— plays of all safety measures. This type of feedback should be regularly (once per month) up—dated. For ex- ample, on one graph, one data point would represent the number of lost— time injuries for a particular month. It is important to remember that such measures will vary considerably from month to month, particularly in smaller companies. However, the monthly graphs will quickly reveal new problems and will allow for the detection of seasonal trends. It may also be useful to keep graphs on which each data point represents to- tals or averages over longer periods of time, such as three months or one year. These illustrations will re- move some of the variability, and will provide clearer pictures of the long— term effects of the program. INTERVENTION Recall that Fox set up a training program before he implemented the in— centive program. That training pro— gram is still in existence. Though the training program, alone, was not sufficient to produce large improve— ments in safety, it is possible that the incentive program would not be nearly as effective if training had not already been ongoing. The train— ing program may have been a necessary ingredient for success even though it, alone, was not sufficient to produce the desired improvements. This same possibility should be considered by anyone developing a program such as Fox's. Before a reinforcement program is ini— tiated, it should be determined if em— ployees know the proper behaviors for safety. If there is any doubt, a training program should be started prior to, or at least simultaneously, to the reinforcement program. The reinforcement program itself may be a bit more complicated than it looks at first glance. Fox was not reinforcing behaviors very frequently, but was, instead, reinforcing products of behaviors (the absence of accidents and injuries) only once a month. Re- inforcement should not be delivered any less often than once a month. It is doubtful that a delivery schedule like those involved in an- nual safety awards programs would be meaningful to many workers. A week— ly or bi—weekly delivery schedule would probably be more effective than a monthly one if the time nec— essary to administer the program is available. Awards other than trading stamps and the prizes for which stamps can be exchanged could be used as interven— tion reinforcers. However, the stamp—prize system has a number of important features. First, they have real value; they are not just meaningless bits of paper. If rein— forcement is to be delivered just once a month, the reinforcers must be substantial. The stamps can be exchanged for a large variety of items. Hopefully, this allows every employee to obtain something that is personally valu— able. If an award system had only a few possible redemption prizes, for example cigarette lighters, picnic jugs and electric blankets, the per— son who does not smoke, rarely goes on picnics and already has an elec— tric blanket might not find the monthly awards at all reinforcing. It is important that there is a va— riety of meaningful reinforcers available for everyone. If a new incentive program is being developed, a questionnaire should be sent to employees to determine what items would be most rewarding for them. The idea that money should not be given as a prize, is based on the assumption that a small money bonus is likely to go into the family gen— eral funds and become lost in pur— chases of tennis shoes and groceries, while the chosen specialty merchan- dise is more likely to be enjoyed and remembered. Using trading stamps rather than setting up a com— pany prize store has one additional practical advantage; the trading stamp company takes care of the nec- essary buying, stocking and redemp— tion work. Once a decision has been made about the kinds of reinforcers to use, the BEHAVIOR MANAGEMENT next reasonable step is to decide whether to give all employees the same amount of reinforcement for safe performances or to graduate the pay— offs according to the dangerousness of jobs (as done in the Fox study). If most of the employees in a plant do similar kinds of work, such as en- gage in small parts assemblying, lo— gistics will be simpler if awards do not discriminate among job assign— ments. However, if there are clear or suspected differences in hazards associated with different assign— ments, fairness, and perhaps poten— tial effectiveness, would argue that the magnitude of awards should be proportioned according to the haz— ards. The data should be analyzed to determine the extent to which var- ious job classifications are associ— ated with accidents and injuries. It is important that even the workers in the safest jobs receive a sufficient magnitude of awards that are likely to be effective as reinforcers. Awards for the other job classifications should be scaled accordingly. This will, of course, involve the estab— lishment of one or more hazard groups. Changes in the hazard groups can be made if data indicate that such al— terations are needed. Fox original— ly had only three hazard groups but established Group IV when data indi- cated that the employees in the rele— vant job classifications were not re- sponding as well as he had hoped. There have,in the past, been success— ful reinforcement programs that have employed only individual or only group incentives, but not both. How— ever there is again some real bene— fit in using a combination of the two. Using only individual incen- tives tends to reduce the extent to which workers talk to each other about safety, and pressure each other to behave safely. Using only group incentives could create an undesir— able situation; once any one employee in a work unit had an accident or in— jury, there would no longer be an in- centive for the rest of the unit mem— bers to behave safely for the remainder of the month. Fox defined his work units to include all of the employees reporting to one supervisor. This was done to maxi— 8-8 mize the opportunities for social contact. Other arrangements may work equally as well,as long as the prac— tice of maximizing contact is fol— lowed. A practice which would prob— ably reduce the effectiveness of such a program would be to provide insig— nificant and infrequent reinforcers. The reinforcer that occurs only once a year, and then only amounts to a $1.98 prize, with no variety in the prize to suit the individual, is not likely to generate an increase in safe behavior. Another important step in implement— ing this kind of safety incentive program is to prepare a proposal to convince the company that the costs of the incentives and the adminis- trative efforts will be worthwhile. (See Figure 8—2.) The person making, the decision about this investment would reasonably want to have a good estimate of the benefits to costs ratio. However, this is impossible to determine because the exact reduc— tions in accidents, injuries, and as— sociated costs to be obtained in a particular setting are yet unknown. Nevertheless, approximations of the benefits-to—costs ratio can be easily calculated. (See Lesson 4). First, assume that your employees have absolutely no accidents, inju- ries or damage costs for one year. Calculate how many stamps (or other reinforcers) would be awarded in this ideal state of affairs. Determine the costs of that number of stamps. The resulting dollar amount is the maximum possible cost of the program for a year. Second, compute the average direct costs over the last few years. In most organizations this will include the costs of medical care, compensa— tion costs, and damage costs. In companies such as public transporta— tion firms, direct costs would also include liability insurance premiums or average costs of claims if they are settled directly by the company. In some companies, direct costs will include known expenses for hiring substitute help to replace people who lose time from work due to injuries. Once the average of total direct costs is calculated, the amount of Safety and Health Loss Control: Incentive Programs INCENTIVE PROGRAM PROPOSAL AND JUSTIFICATION The attached graphs show the extent of our safety problems and costs over the last few years. I propose that we begin a safety Incentive program, as described below to try to improve our safety record.ThepK@mn1k estimated to cost approximately $ per year and will more than pay for itself given a percent reduction in injuries and accidents. Incentive Program Description - I propose that we give trading stamps to workers who go for periods of time without injury or accident or otherwise contribute to safety. I have analyzed our safety data to determine the most hazardous jobs and propose that we provide greater incentives for workers Who have greater risks for injuries and accidents. In addition, we should give bonus awards to every member of a work crew if no member of that crew has an injury or accident. Stamp awards, bonuses and penalties will be given according to the following schedules: HAZARD JOB MONTHLY INDIVIDUAL MONTHLY CREW GROUP ASSIGNMENTS AWARDS BONUS AWARDS I stamps stamps ll stamps stamps III stamps stamps INDIVIDUAL AND CREW DAYS LOST TIME BONUSES LOST FOR month months months months months months DOLLARS PROPERTY DAMAGE = NUMBER MONTHS BONUS AWARDS LOST $ ( month maximum) In addition to the above, I propose that the safety committee be authorized to give bonus awards to employees deserving of special consideration, such as those who make safety suggestions that are adopted and those who prevent injury or damage by some meritorious act. To help our employees appreciate the value of the incentive program, I propose that we begin the program by giving stamps to each Figure 8-2. An adaptation of the proposal written by Fox for the plant manager. 8-9 BEfliAVHCfliLLAN1H3ENflfl¢T employee who has not sustained an injury or been charged with equipment damage during the last year(s). Safety data, on which these awards will be made, are already kept by my office. Stamps should be included in pay envelopes at the end of the month. This can be done either by my office or by the payroll clerk. Incentive Program Costs - The following cost breakdown is computed assuming zero injuries and damage costs. Therefore, these figures are maximums. Hazard Group l: employees at stamps per month ( stamps for individual awards + stamps for crew bonus awards) times 12 months. stamps Hazard ll: (ditto) stamps Hazard Group III: (ditto) stamps Safety Committe Bonus Awards stamps TOTAL STAMPS/ONE YEAR Stamps will cost us per 5,000. Therefore, the yearly dollar costs would be S Cost Justifications - I estimate that we can reduce accidents and injuries percent over the first year we have the incentive program in effect iit typically takes several months for the full effects of such a program to be felt). After that first year, we would be spending a maximum of S per year on the incentive program while reducing the costs of injuries and accidents 5 per year. This would represent a net savings of $ per year. These figures represent conservative estimates. l have determined our current costs of injuries and accidents based only on hard dollar figures while our industry ordinarily assumes that hidden costs of injuries and accidents make a substantial contribution to total losses. In addition, the estimated costs of the incentive program are maximums which will probably never be realized. Finally, I believe there will be a considerable improvement in general morale and motivation that will result from this program. We will be demonstrating by a concrete investment, that we are concerned about safety while placing a greater responsibility for safety on the shoulders of our employees. Figure 8-2. (cont.) Safety and Health Loss Control: Incentive Programs savings that will result from various percentages of decreases in costs should be determined. In most incen— tive programs, a 10% - 20% reduction in accidents, injuries, and related costs, will compensate for the costs of implementing the program. If the proposal is approved, the fin- al steps are to obtain the necessary reinforcers, inform the employees about the reinforcement procedures, and begin the reinforcement. It is pertinant that a list be compiled of all the safety statistics for each employee, indicating the amount of reinforcement awards they are to re— ceive. The indicated number of trad— ing stamps can easily be inserted in— to each employee's paycheck envelope. TRANSLATING THE RESULTS The data system employed in a safety incentive program will provide an on— going record of all safety—related statistics. Therefore, reductions that occur among the various measures following the intervention can be easily detected. To determine the effectiveness of the incentive pro- gram, measures of the desired safety behavior must be taken at regular in— tervals. The data collected should be graphically illustrated, or put into table form for easy comparison. It is important to realize that the data plotted each month, (at the on— set of the program), can be highly variable. To ensure reliability, it is therefore imperative that conclu- sions about possible improvements in safety behaviors not be drawn until an adequate amount of data has been collected and properly translated. Computations such as percentages of safety behavior increases or de- creases, daily and weekly production averages, and the cost—benefit ratio (see Lesson 4) can aid in the trans— lation of data. EVALUATING THE INCENTIVE PROGRAM Once sufficient data have been ob- tained and transformed for compari— sons, the incentiJe program can be objectively evaluated. Comparisons must be made of the data collected at the various intervals throughout the intervention. To determine the full effects of the incentive program, these data must also be compared to those obtained prior to, and follow- ing, the implementation of the pro— gram. The differences in these sta— tistics will indicate the success of the incentive program to influence an increase in safety behaviors. To arrive at the most accurate conclu— sions and evaluations, it is neces— sary to continue to collect, trans— late, and compare data for several years following the initiation of the incentive program. SUMMARY COMMENTS Rather than frequently reinforcing instances of safe behavior, it is possible to periodically provide re— inforcement—contingent on the pro— ducts of behaviors. Many organiza— tions routinely keep statistics on the products appropriate for the re— inforcement of safe behaviors. An easily implemented system of rein— forcers can be based on the bestow— ment of trading stamps. The result- ing safety program is quite simple to administer because the data on which reinforcement is based should already be available, and the logistics of purchasing and distributing stamps are simple. The basic ingredients of an incentive program involve the provision of re- inforcement to employees, based on individual safety performance, and on the performances of other members of a work unit. There can also be par— allel penalties, such as losses of awards, for injuries and accidents — the products of unsafe behaviors. The obsolute and relative sizes of awards can be easily tailored to fit the conditions existing in each com- pany. Special awards can be present— ed to accommodate particularly out— standing safety performances. Through the use of an incentive pro— gram, specific desired safety behav— iors may be successfully elicited and maintained. BEHAVKNKMANAGEMENT REVIEW Following are the questions which should be considered when setting up and im— plementing an incentive program: 1. 10. 11. 12. Has a safety training program been established to insure that all workers know how to behave safely? Has the safety data been analyzed to determine known costs for injuries and damages over the last few years? Has the safety data been analyzed to determine the extent to which various classifications have been involved in injuries and equipment damage? Have employees been divided into hazard groups based on the analyses of the safety data? Have tentative values been set for the awards and penalties? Has the amount of money to be saved assuming,various percentages of reduc— tions in direct costs,been determined? Has the maximum cost of the incentive program, assuming that there are no injuries or accidents, been determined? Has the estimated savings that will result from the incentive program been calculated? Have the values of the awards and penalties been adjusted according to the estimated cost—benefit ratio? Has a written proposal for the incentive program been submitted? Have the safety data been monitered, and the graphs maintained and updated monthly? Have adjustments been made in the schedule of awards and penalties based on recent data? Safety and Health Loss Control: Incentive Programs REFERENCES 1. FOX, D.K. Effects of an Incentive Program on Safety Performance in Open Pit Mining at Utah's Shirley Basin Mine, Wyoming. Paper presented at the Midwest Analysis of Behavior Association Meeting, Chicago, May, 1976. SUGGESTED MATERIAL Industrial Loss Control Through Behavior Management. Film distributed by the Division of Training and Manpower Development, The National Institute for Occupational Safety and Health, Cincinnati, 1979. 8‘13 trusepo: 1979—657-061/5829 GENERAL UBHAHY-ILC BERKELEY MIMIMIIIMM BUDDHWBSHL