Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The Modification Of Maladaptive Behavior Of A Class Of Educationally Handicapped Children By Operant Conditioning Techniques
(USC Thesis Other)
The Modification Of Maladaptive Behavior Of A Class Of Educationally Handicapped Children By Operant Conditioning Techniques
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
This dissertation has been microfilmed exactly as received 67-6500 HOTCHKISS, James M erel, 1930- THE MODIFICATION OF MALADAPTIVE BEHAVIOR OF A CLASS OF EDUCATIONALLY HANDICAPPED CHILDREN BY OPERANT CONDITIONING TECHNIQUES. University of Southern California, Ph.D.t 1966 Education, psychology University Microfilms, Inc., Ann Arbor, Michigan Copyright by James Merel Hotchkiss THE MODIFICATION OF MALADAPTIVE BEHAVIOR OF A CLASS OF EDUCATIONALLY HANDICAPPED CHILDREN BY OPERANT CONDITIONING TECHNIQUES by James Merel Hotchkiss A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Educational Psychology) September 1966 UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 8 0 0 0 7 t This dissertation, written by . ..................JAm.e.a.Mfijr.eL.HQtchkisa.................. under the direction of his.....Dissertation Com mittee, and approved by all its members, has been presented to and accepted by the Graduate School, in partial fulfillment of requirements for the degree of D O C T O R O F P H I L O S O P H Y Dean Date . S a p . t . e mbe r.. 1 . 9 . 6 6 . DISSERTATION COMMITTEE i . Chairman TABLE OF CONTENTS Page 1 The Purpose The Hypotheses General Hypothesis Singular Hypotheses Scope and Delimitations Limitations of the Instruments Definitions of Terms Organization of the Remaining Chapters Chapter I. THE PROBLEM II. REVIEW OF THE LITERATURE.......................15 Research on Behavior Therapy Studies using Reciprocal Inhibition (Counterconditioning) to Treat Anxiety or other Maladaptive Behaviors Research on Operant Conditioning Techniques Studies using Operant Conditioning Techniques Studies using Social Consequences as a Reinforcer Studies using Operant Conditioning Techniques in the Classroom Summary ii Chapter Page III. METHODOLOGY....................................... 65 The Subjects Subjects in the Experimental Class Subjects in the Contrast Class Description of Rating Instruments used The Activity Check List The Burks Behavior Rating Scale The Teacher Observation Form The Sound Level Meter Other Sources of Data Parent Information Impressions and Opinions of Administrative and Guidance Personnel Pre-Experimental Period Procedure J udges Classroom Observation Instructions for Experimental Class Teacher Instructions for Contrast Class Teacher Daily Procedure in the Experimental Class Physical Arrangement of the Classroom Selection of the Experimental Time Period First Day Instructions to Children Procedure during the Experimental Hour Daily Procedure in the Contrast Class Post-Experimental Period Observations Method of Data Analysis Significance Level iii Chapter Page IV. EXAMINATION OF THE EXPERIMENTAL RESULTS.......................................... 106 The Hypotheses Hypothesis I Hypothesis II Hypothesis III Hypothesis IV Supplementary Findings Comparisons of the Two Classes V. COLLATERAL INFORMATION ........................ 129 Changes Occurring during the period of the Study Schoolwork Reduction in Self-Centered Thinking Fighting and Aggression Behavior on the School Bus Post-Experimental Period Behavior Home Behavior Appearance of New Undesireable Symptoms or Behavior Contrast Class Data VI. SUMMARY AND DISCUSSION..........................139 Summary Purpose Methodology Data Collection and Analysis Subjects iv Chapter Page Major Findings Supplementary Findings Additional Findings Collateral Findings Dis cussion Conclusions REFERENCES..................................................157 APPENDIXES Appendix A. Activity Check List Records Raw D at a..................169 Appendix B. Sound Level Meter Records Raw Data..................186 Appendix C. Teacher Observation Form Raw Da ta............................ 191 Appendix D. Recording Forms for Rating Instruments .............. 195 Appendix E. Letters from Redondo Beach School District Personnel . . . 201 Appendix F. Instructions to Experimental C l a s s ...............................215 v LIST OP TABLES Table Page 1. Summary of Experimental Class Data.......... 80 2. Summary of Contrast Class Data............... 8l • i 3. Combined Ratings of Three Judges of Student’s Classroom Activity on the Activity Check L i s t .............................107 Individual Ratings of Student's Classroom Activity on the Activity Check List— Judge No. 1...........................108 5. Individual Ratings of Student’s Classroom Activity on the Activity Check List— Judge No. 2..........................109 6. Individual Ratings of Student's Classroom Activity on the Activity Check List— Judge No. 3...........................110 7. Combined Judge's Ratings of Student's Classroom Activity on the Activity ^ Check List (Contrast C l ) ........................Ill 8. Individual Ratings of Student's Classroom Activity on the Activity Check List— Judge No. 1...........................112 9. Individual Ratings of Student's Classroom Activity on the Activity Check List— Judge No. 2...........................113 10. Individual Ratings of Student's Classroom Activity on the Activity Check List— Judge No. 3...........................11^ vi Table Page 11. Statistical Analysis of Sound Level Meter Data obtained in Pre and Post-Experimental Periods ............ 117 12. Comparison of Teacher Reinforcement of Adaptive and Maladaptive Subject Behavior in Pre and Post-Experimental P e r i o d ............................................119 13. Teacher Evaluations on Burks Behavior Rating Scale .......................... 121 lb. Comparison of the Activity Level Average of the Experimental Class with the Contrast Class......................... 125 15. Between-Class Comparisons of Sound Level Meter Data in Pre and Post- Experimental Periods ........................... 126 16. Between-Teacher Comparison of Teacher Reinforcement of Adaptive and Maladaptive Pupil Behavior .............. 127 17. Between-Class Comparisons on Burks Behavior Rating Scale .......................... 128 18. Summary of Teacher’s Evaluations of Individual Academic Performance during Study ................................... 131 vii LIST OF FIGURES Figure Page 1. Arrangement of the Experimental Room .... 94 2. Arrangement of the Contrast Room............ 95 3. Graphic Illustration of the Relative Position of Mean Sound Level Meter Readings of the Experimental and Contrast Rooms in the Pre and Post- Experimental Periods............................. 116 viii CHAPTER I INTRODUCTION Recently In the United States there has been an in creasing interest in the identification and treatment of emotionally handicapped children. While there is agreement the desirability of having and maintaining special class es for these children there is wide and vociferous disagree ment about the handling and treatment of the children while in the classroom (Quay, 1963: Morse, Cutler, & Fink, 1964; Burks, 1965). For example, Morse, Cutler, & Fink, in re porting research findings based on a survey of all emotion ally handicapped classes in the United States in 1962, state that: Goals were stated very generally, and little dif ference in program types existed among the many kinds of children served. In planning the development of their programs, very few of the responsible agents sought to draw upon the experience of prototype pro grams in other areas and many found solace by complain ing that the experts could not agree on what was to be done . . . Neither the profession of mental health nor that of education offered enough solid conceptualiza tion of psycho-educational problems to allow for a basic common framework of goals, structures, and imple menting activity to be developed. (p. 9) . . . In some instances, the difference in point of view was so great that diplomatic relations had been broken off, and one or the other point of view thoroughly dominate the program . . . The basic problem however, remains. Neither orientation alone offers a 1 2 sufficiently meaningful way of dealing with the pro blems of the child operating in the total psycho-educa tional complex. . . The rare program set its goals in terms of a method which had been presented as success ful in the literature. (pp. 9-10) Their findings revealed the existence of seven dif ferent types of programs for emotionally handicapped child ren currently in use in the United States: 1. Psychiatric Dynamic: Major emphasis was on dynamic therapy and pupil acceptance, with educational as pects played down or secondary. Individual therapy was expected or required. Parental therapy was stressed. There was heavy psychiatric involvement in diagnosis, decision making, treatment processes, consultation, and evaluation. Emphasis was on ac ceptance, use of interpersonal relationship, and overall tone. 2. Psycho-Educational: Psychiatric and educational emphases were balanced with joint planning and interweaving-equality of two emphases, educational and clinical. Educational decisions were made with a consideration of underlying and unconscious moti vation. Educational aspects stressed creative, pro ject type work, individual differences, and a benign but not permissive atmosphere. Clinical participa tion was apparent, but not omnipresent or decisive in day to day actions. 3. Psychological Behavioral: This series was based in systematic psychology of learning theory, with em phasis on diagnosis of learning potential capacities and relationship to specific remediation techniques. It Involved the use of associative learning and for mal habit. It contained a nonpunitive structure with emphasis on changing symptomatic responses through specific techniques on a planned, ego level. 4. Educational: Emphasis was on formalized, accepted educational procedures such as routine drills, work books, inhibition of symptomatic behavior, and at tention to skill training and dri]1. Little use was made of group processes. Emphasis was on control with restrictive handling seen as corrective. Atmo sphere was nonhostile. These classes relied largely on extension of traditional educational procedures without much systematic attention to the theoretical design. 3 5. Naturalistic: The teacher operated on a "green thumb" naturalistic basis without organized approach or any specific design. The work was dominated by ad hoc responses to individual problems (academic- behavioral) as they appeared. Frequently the teach er assumed a benign, kind but demanding mother- teacher role. Various procedures were used without any well developed plan. Sometimes the method of control involved joking. The teacher interaction was pervasive in interventions and decision making as the process evolved, but there was not much depth or fundamental consistency to the interventions. 6. Primitive; There was an overall coarseness evi denced in both rationale and handling procedures. Sometimes the teacher was aloof and cold. Control was maintained by establishing limits through a "no monkey business" approach by domination and fear. The class was essentially a holding company opera tion, with a lack of sensitivity in the overall tone. Emphasis was on surface compliance for its own sake. 7. Chaotic: Here, impulsive behavxor broke through continually and any semblance of order was momen tary. This might have been a consequence of ex treme passivity and permissiveness or an inability to cope with the situation and a lack of adequate back up or removal. There also may have been, in some instances, a belief or rationalization regard ing the therapy of permissiveness. (pp. 28-30) Their general conclusion was that there is at pre sent no overall direction in the goals, terminology, meth ods of selection or conduct of programs for educationally handicapped children and that: Beyond this, research reveals an amazing lack of specific pattern and uniformity in approach, much more than had been anticipated. Approaches are much less systematic and much more intuitive than had been expect ed. In part, this lack of systematization is a reflec tion of the situational responses to a broadly varied set of local situations. But, it also reflects a con fusion on the part of both educators and clinicians about how to proceed to solve this most trying educa tional and social dilemma. Here researchers and prac titioners together have a heavy obligation to provide a n solid conceptual system for the understanding of psycho- educational problems, and such a system manifestly does not exist at the present time. Cp* 130, Italics mine) Reger (1965), in discussing trends of special edu cation, places the problem in an even more general context of the educational philosophy of the program. There is, he believes, an almost complete lack of explicit philosophical direction around which to build a curriculum. This is seen as perhaps the most fundamental and important problem in all of special education. He says that more research with out more philosophy is an approach lacking sophistication and that all research should be placed in a philosophical context before it has meaning. Reger says that, first of all, we must define what we mean by education. For example, Wild (1955) suggests that "this transmission of knowledge is the task of education." Reger (1965), on the other hand, defines education as "nothing less than the changes made in human beings by their experiences" (p. 119)• Reger feels that without exploring such a basic issue as a definition of education (and studying the curriculum and program implica tions involved), the result is likely to be a continued frantic search outside the field of education for direction and guidance. Quay (1963) agrees with Reger and he too, warns of the danger of the rapid growth of programs for emotionally handicapped children before treatment methods are developed and experimentally proven. As Morse, et al. (196*0, have also noted, classes for emotionally handicapped children presently have questionable utility in the treatment of these children. Kounin, Friesen, & Norton (1956) concluded as follows: Although authorities recognize that emotionally disturbed children are present in schools, and that the classroom is important in the world of the child, we know of no definitive research on what the classroom teacher can d£ in programing for or in managing the be havior of emotionally disturbed children in regular classrooms. (p. 1) In a similar vein, Rabinovich (1959) has recently stated that it is high time that the school did assume the responsibility of educating emotionally disturbed children who were formally treated in agencies. But, as he points out, it is not yet clear how they are to be taught, and "The commonest, single immediate cause for referral to our guidance clinics today is behavioral, academic or social difficulties experienced by children at school" (p. 857). Morse, et. al. (1964) reported that behavior control was the major problem reported by teachers, and motivation for school work was a distant second. Shannon (1961), in a review of various traditional treatment approaches, reached the following conclusion: Research indicates that no group or profession has demonstrated the ability to effectively deal with de viant behavior; research shows that treatment results in no greater improvement that that which accrues by simply leaving persons with a behavioral problem alone. In this case the cry for more money, for a saturation approach, is indeed uncalled for. (p. 35) 6 We are still faced with the task of converting the many theoretical propositions, observational reports, lit erary humanism, and appealing speculations into concrete manipulable experimental terms. This may not be as fore boding a task as it first appears. One might note, for example, that certain physiological problems with complex etiologies (e.g., myopia, benign tumors) can be solved by rather simple corrective procedures. (Schwitzgebel, 1965) One of the more recent and popular techniques used by experimental psychologists is a method known as the "free operant" (Perster, 1953; Sidman, 1956; Skinner, 1938). Skinner (1953) has held that principles derived from this method are sufficient to explain and to design entire human cultures. Even without raising ethical issues, it would appear judicious to have considerably more empirical evi dence regarding the applicability of such principles to limited (but common) behaviors of human beings in their natural environments. One such "natural environment" is the public school room. There is currently an increasing interest in the im plementation of operant techniques to the classroom. To date, however, there are no published studies in which these techniques or principles were applied to an entire classroom. Two recent studies have addressed themselves to the urgent need for such research. Whelan & Haring (1966), concluding their report on research at the Children's Rehabilitation Unit of the University of Kansas Medical Center, state the following: Staff members are investigating and demonstrating the efficiency of utilizing behavioral modification techniques with individuals, small class groups, and classes of 15 to 20 children. Validity of these tech niques will be achieved only if they can be practically applied to groups of children by one teacher. It is most unreasonable to expect school districts to staff special classrooms with a teacher, assistants, and expensive automated equipment. (p. 285) They also stateT Data reported from laboratory experiments and a few studies with small groups of children have demonstrated a high degree of reliability. However, these data must be validated in regular and special classroom situa tions. This validation may or may not be forthcoming; it has yet to be demonstrated. (p. 288) Quay, Werry, McQueen, & Sprague (1966) reached a similar conclusion: The economics of public schools obviously require the development of techniques that will allow children to be handled in a group situation by as few adults as possible. Most of the techniques of behavioral reme diation have been developed for use on an individual basis and it seems crucial at this stage to attempt to extend these techniques to group situations . . . be havior techniques . . . are nevertheless likely to re main economically unfeasible, unless they can be adapt ed for use in a group setting such as the classroom. (p. 513) The Purpose The purpose of the present study was to explore the applicability of operant conditioning techniques, now in intensive use in the therapy of individuals, to class groups. This experiment was to determine whether operant procedures could be feasible with a class as a whole, and whether a whorthwhile decrease in the occurrence of maladap tive behaviors and an increase in socially acceptable be haviors would result. These purposes can be expressed in the form of questions, as follows: 1. Can the general method of operant conditioning be applied to a class-as-a-whole? 2. Can controls over the entire class be exercised by employment of operant procedures? 3. Can improvements in behavior be brought about by the use of operant techniques? a. Will there be a decrease in the occurrence of maladaptive behavior? b. Will there be an increase in adaptive be havior? c. Will there be a decrease in the activity level of the classroom? Will any new or different behavioral symptoms appear during or after the experimental period? 5. Will any change in behavior observed in the classroom generalize to outside situations? In other words will maladaptive behavior be reduced and adaptive behavior increased outside the classroom? The Hypotheses General Hypothesis Operant conditioning techniques, similar to those now in intensive use in the therapy of individuals, can be extended to an entire class of educationally handicapped children and can improve classroom behavior. The above hypothesis is general in nature and was not specifically tested. It formed a basis for the for mulation of singular hypotheses which were resolved in the manner to be explained in Chapter IV. Singular Hypotheses Hypothesis 1. The use of operant techniques will reduce the occurrence of maladaptive behavior in a class room of educationally handicapped children. Hypothesis 2. The use of operant techniques in a classroom for educationally handicapped children will re duce the background noise level generated by the pupils. Hypothesis 3. The use of operant techniques by a classroom teacher will result in less teacher reinforcement of maladaptive student behavior of educationally handicap ped children. The next hypothesis was developed to assess the generalization of classroom behavior to the total school setting— from the time the children left home in the morn ing on the school bus to the time that that same bus drop ped them off in the afternoon. 10 Hypothesis 4. The use of operant techniques based on behavior therapy will reduce the occurance of maladap tive and hyperactive behavior In the total school environ ment . Scope and Delimitations This study was limited to the effects of operant techniques upon the behavior of students enrolled in one of three classes for educationally handicapped children located at Franklin Elementary School in the Redondo Beach (California) Elementary School District. The experimental class was composed of nine Caucasian boys, eight of whom completed the experimental period. One boy was withdrawn when the family moved to another school district. The boys were between the ages of 8-8 and 10-6 with average or above intelligence quotients (88 - 119). They came from middle and lower middle-class backgrounds. The study was limited to classroom behavior as rated by three judges observing the children in the class room and rating them on an Activity Check List; by teacher ratings on the Burks1 Behavior Rating Scale; and by the sound or noise level of the classroom as measured by a General Radio Type 759 Sound Level Meter. Teacher utiliza tion of operant techniques was rated on the Teacher Obser vation Form by three judges. Further evaluation of the program included a Parent Opinion Questionnaire, and 11 evaluations from various administrative and guidance per sonnel in the Redondo Beach School District who visited the class before, during, and after the experimental period. The study was further limited to the school year in which it was done and in particular to the 22 day ex perimental period. Finally, this experiment was limited in that no provision for follow-up study in future years was made. Limitations of the Instruments The conclusions of this experimental study are de pendent upon the following assumptions: 1. The Activity Check List was a valid means of recording various types of rraladaptive behavior observed in children. The dependent variable in this rating was an "activity score" for each child derived by dividing the number of maladaptive activities noted by the total number of observations. This Activity Check List was selected for this study because it had been developed for use as a rat ing instrument for maladaptive behavior in children in the experimental school at the University of Oregon by G. R. Patterson who had previously used it extensively in operant conditioning studies with children (Patterson, Jones, Whittier, & Wright, 1965). See appendix D. 2. The Burks1 Behavior Rating Scale as scored by the classroom teacher yielded a valid estimation of classroom behavior of children. The Burks* Scale was chosen because it is a basic instrument utilized for the selection and placement of children in special classes in the Los Angeles County Schools. Also, school personnel in volved in the study were familiar with the instrument and teachers had had much previous experience in rating child ren with it. The scale was first developed by H. F. Burks of the Los Angeles County Schools for the use of teachers to gain an estimate of behavior which might in part spring from organic pathology of the central nervous system, as well as maladaptive behaviors (Burks, 1955). See Appendix D. 3. The General Radio Type 759 Sound Level Meter accurately registered the level of noise in the classroom. The instrument was secured from the Physics Department of the University of Southern California and had been cali brated immediately prior to the beginning of the experi mental period. This instrument was manufactured by the General Radio Corporation in Boston, Massachusetts. 4. The Teacher Observation Form provided a valid indication of positive and negative reinforcement of both adaptive and maladaptive behavior. This form was prepared for this study. See Appendix D. 5. Valid information of the child's behavior in the home and community during and following the experimental period was provided by the Parent Questionnaire and by 13 parent interviews. See Appendix D. 6. The impressions of guidance and administrative personnel who observed the class are valid and reliable. These impressions are in the form of letters to the Univer sity advisor. See Appendix D. Definitions of Terms Adaptive Behavior. Any behavior that helps the child meet environmental or personal demands. Educationally Handicapped. These are children who by reason of marked learning problems, or behavior al problems or a combination thereof, cannot receive the reasonable benefit of ordinary education facilities; but who are not also handicapped by mental retardation or obvious physical disabilities for which special pro grams are already provided elsewhere in the Code. (Education Code of the State of California, Section 6750, p. 336). Administrative Code Section 221 further defines an Educa tionally Handicapped minor as a minor described in Education Code 6750 whose learning problems are associated with a behavioral'disorder or a neurological handicap or a combination thereof, and who exhibits a considerable discrepancy between ability and achievement, (p. 118.15) Emotionally Disturbed and Emotionally Handicapped. In this investigation the terms "Emotionally Disturbed" and "Emotionally Handicapped" are synonomous with the term "Educationally Handicapped" (see "Educationally Handicapped above). 14 Free Operant Technique. The operant technique used in this study Is described by Ferster's (1953) definition: The use of the free operant is a method of wide generality; it refers to any apparatus that generates a response which takes a short time to occur and leaves the animal in the same place ready to respond again. The free operant is used in experiments when the main dependent variable is the frequency of occurrence of behavior. Nearly all the problems of a science of be havior fit this paradigm when the questions are of the form: what is the likelihood of a piece of behavior occurring on this particular occasion; how strong is the tendency to behave on this occasion relative to another occasion? The free operant has advantages in this respect because it removes restrictions on the frequency with which a response can occur and permits the observation of moment to moment changes in fre quency. (p. 263) Hyperactive Behavior. Behavior emitted at a rate much faster or greater than would be considered normal for the situation in which it occurs as determined by the classroom teacher. Maladaptive Behavior. Behavior that is considered inappropriate by those key people in a person's life who control reinforcers. Such maladaptive behavior leads to a reduction in the range or the value of positive reinforcement [or re ward] given to the person displaying it. (Ullman & Krasner, 1965, p. 20) Operant Behavior. The totality of operant respons es in the behavior repertoire of the child. Operant Conditioning. In this investigation operant conditioning is an experimental technique wherein a pupil is rewarded after engaging in operant behavior of a 15 selected kind. The successful conditioning of an operant response is inferred from an increase in its rate of occur rence as a function of reward (reinforcement) administered by the experimenter. Reward. A satisfaction-yielding stimulus or stimulus object that is obtained upon the successful performance of a task: e.g., a piece of candy or a penny for being quiet; teacher praise for a correct paper; peer-group approval. (English & English, 19^8) Socially Acceptable Behavior. Behavior that is considered appropriate by those key people in the pupills life. In this study socially acceptable behavior is synonomous with adaptive behavior. Organization of the Remaining Chapters Chapter II presents a review of the literature which relates to the problem and nature of this investiga tion. Chapter III describes the experimental and contrast pupils, the rating instruments and materials used, and the data analysis. Chapter IV presents the findings of this investiga tion with an evaluation in terms of the research hypotheses. Chapter V presents additional information that accrued to this investigation as a result of the experimen tal methods used. Chapter VI presents a summary of this study and relates the general conclusions which can be made from the findings. CHAPTER II REVIEW OP THE LITERATURE This chapter reviews the literature which pertains to the problem and nature of the investigation. Basically, this review can be divided into two sections: a review of literature pertaining to behavior therapy and literature pertaining to operant conditioning methods. The literature dealing with behavior therapy is limited to that based primarily on reciprocal inhibition (or counterconditioning) which has been used with human subjects. This section is further divided into two sec tions: a discussion of the theory of reciprocal inhibition and a review of specific studies implementing this theory which are relevant to this investigation. The second general section is concerned with the development of the concept and application of operant con ditioning techniques with humans. This section is also divided into two sub-sections. The first sub-section dis cusses the development of operant conditioning techniques with humans and the second reviews studies in which the operant technique"has beer} used with humans. Studies in this group fall into three main areas. The first section contains studies using operant conditioning techniques in 17 the laboratory or hospital setting. These studies are with psychotic patients, autistic children, adolescents and children with deviant behaviors. The second group of studies is concerned with the use of social, or verbal, re wards administered by adults or by the peer group to con trol and change behavior. The final group is composed of studies using operant techniques in the classroom situation with individual children. Research on Behavior Therapy This section will review the applications of learn ing theory to behavior therapy that are germane to the pre sent investigation. There has been an increasing use of learning principles with humans in the last few years in the treatment of specific behavior disorders (Bandura, 1961; Reyna, 1965). More specifically, Grossberg (1964, p. 73), in a review of the literature says "behavior therapy is derived from the rejection of traditional per sonality theories, and consists of the application of the principles of modern learning theory to the treatment of behavior disorders." Bandura says that on the whole, the evidence, while open to error, sug gests that no matter what the origin of the maladaptive behavior may be, a change in behavior brought about through learning procedures may be all that is neces sary for the alleviation of most forms of emotional disorders. (1961, p. 157) Reyna described the viewpoint of the behavior therapist towards maladaptive or neurotic behavior as 19 learned behavior: . . . conditioned emotional, verbal, and motor re sponses [which] have resulted from a history of aver- sive events and are maintained by immediate reinforce ment of behaviors that are instrumental in preventing extinction of conditioned emotional responses. These responses may constitute a small or large part of the individual's repertoire, depending on whether acquired and generalized anxiety responses have led to fresh aversive experiences. In describing the behavior called neurotic as learned, the emphasis is on indicat ing that this behavior is initially the result of various external operations such as reinforcement, generalization, and contiguity rather than on postu lated unobservable inner forces. Accordingly, no under lying disease state or state on neurosis is assumed to be present of which behaviors are the symptoms. Rather, behavior that is ineffective, useless, incapacitating, and persistent and that includes stron aversive emo tional components may be called "neurotic” or "unadap- tive.” (Reyna, 1965, p. 171) Wolpe (1958, p. 32) defines neurotic behavior in much the same terms: "any persistent habit of unadaptive behavior acquired by learning in a physiologically normal organism." He believes that adaptive consequences of behavior take the form of progress toward the satisfaction of a need and the avoidance of possible damage or deprivation. On the other hand, unadaptive consequences would be the expenditure of energy or the occurrence of damage or deprivation. He further hypothesizes that environment or stimulus events present when unadaptive behavior occurs become conditioned by stimulus generalization to the behavior and consequently become anxiety producing in themselves. He defines anxiety as "the autonomic response pattern or patterns that are characteristically part of the organism's response to noxious stimulation" (p. 3*0. Thus, for example, a box in 20 which a cat is shocked becomes on sight, anxiety producing to the cat although the box itself may be harmless with the electric grid long removed. He postulates that "free- floating anxiety" or fear in humans is developed in much the same way. Wolpe's method of treatment is based upon his form ulation of reciprocal inhibition as a therapeutic principle: If a response antagonistic to anxiety can be made to occur in the presence of anxiety-evoking stimuli so that it is accompanied by a complete or partial sup pression of the anxiety responses, the bond between these stimuli and the anxiety responses will be weaken ed. (Wolpe, 1958, p. 70) Thus, maladaptive behavior caused by fear or anxiety in a particular situation could be removed by causing the subject to have pleasant and satisfying experiences over a period of time in the same situation. Since pleasure and anxiety are not compatible, as the pleasant feeling grew, the anxiety is extinguished. In a typical study, Wolpe (1962) reported on a woman subject with a severe phobia of traffic situations. She had been in an accident and was afraid of cars approaching from any direction. Deep muscle relaxation was used, and she was encouraged to imag ine herself in various traffic situations, starting with very non-threatening episodes and gradually progressing to very dangerous ones. After 57 sessions treatment was dis continued. A follow-up in 1962--2 years later— revealed no new symptoms and a complete cure. The subject also 21 reported that her interpersonal relations with her husband had improved. Wolpe gives a number of similar case studies which are not reviewed here (Wolpe, 1958; 1965). He reports that his use of the reciprocal inhibition method (or counterconditioning) with 212 cases has resulted in favor able outcomes in 90 per cent of the cases, with 39 per cent "apparently cured," 50.5 per cent "much improved" and 7.2 per cent "slightly to moderately improved" (1958, p. 216). Haring and Phillips (1962) attempted to systemati cally apply this concept in working with a classroom of emotionally disturbed children in a public school. They reasoned that if children for whom school was anxiety- arousing, threatening, and associated with failure, could be made to experience pleasure and success in the school setting, the former behavior and attitudes would extin guish. They found this to be the case and obtained signi ficant improvement in their S's. Studies using Reciprocal Inhibition (Counterconditioning) to treat Anxiety or other Maladaptive Behaviors The earliest study using counterconditioning was by Mary Jones (1924). She reported on Peter, a boy, 2 years and 10 months old who was frightened of furry objects and particularly afraid of rabbits. Peter was placed in a pleasant play or eating situation with other children and 22 the rabbit was brought in for short periods of time. At first it was kept at long distances from Peter. Gradually it was brought closer until the boy was quite at ease with it. As he became more at ease with the rabbit he also began to lose his fear of other furry objects. A follow-up showed that his acceptance of the rabbit had generalized to all other furry animals and objects with whom he had come into contact and that he no longer showed any fear at all. In describing her procedure, Jones writes: The hunger motive appears to be the most effective for use in this connection. During a period of craving for food, the child is placed in a high chair and given something to eat. The fear-object is brought in, starting a negative response. It is then moved away gradually until it is at a sufficient distance not to interfere with the child's eating. . . . While the child is eating, the object is slowly brought nearer to the table, then placed upon the table and finally, as the tolerance increases it is brought close enough to be touched. (p. 113). Lazarus (I960) was also one of the earliest to apply this technique to children. He reported that 18 phobic cases ranging in age of 3-1/2 years to 12 years have been treated by these techniques with grati fying results. These cases had all either recovered or were much improved . . . follow-up studies conducted over periods of 6 months to 2-1/2 years revealed that none of the children had relapsed in any respect. (P-120) In one experiment, a nine year old girl who was suffering from separation anxiety, night terrors, and psychotic symptoms in the absence of her mother, which necessitated her being absent from school, was given relaxation train ing. When fully relaxed, the child was asked to imagine 23 that she would be separated from her mother for a period of five minutes, then of fifteen minutes, and finally up to a week. After a period of 10 days in which she received 5 treatments she was returned to school. A follow-up 15 months later indicated that her progress had been main tained. Lazarus (1961) studied the application of Wolpe's reciprocal inhibition theory to group therapy sessions. He treated 18 subjects by desensitization methods and re ported 13 recoveries. Of the 13 recovered, a follow-up 9 months later showed that only three had relapsed. With a more traditional form of interpretive group therapy with 17 subjects he found that only two patients were symptom free after 22 sessions and that both of these subjects had attended meetings in which relaxation and countercondition ing was employed. Next, Lazarus took the 15 subjects who had shown no improvement under interpretive therapy and treated them by group counterconditioning. Ten of these were recovered after a mean of 10.1 sessions. His subjects were all phobic patients whose phobias "imposed severe limitations on their social mobility, jeopardizing their interpersonal relationships or hindering their constructive abilities" (p. 506). Breger and McGaugh (1966) criticize this study, and rightly so, in that Lazarus was in both situations the therapist, the reporter, and the evaluator of the results. 24 Bentler (1962) used counterconditioning with a year-old infant girl who was frightened of water. Treat ment consisted of four parts and lasted one month. First, she was encouraged to play with toys in the bathroom and near the empty tub. After she was able to do that with no fear she was encouraged to move nearer to the tub and re move toys from it. Then she was placed by the sink which had water in it and gradually was required to get her feet, and hands wet in order to reach desired toys. As she adjusted to the water she was then washed in the sink while she played with her toys. Finally, she was returned to the tub for bathing where after two days she adjusted quickly. She became quite fond of water and was thereafter anxious to be in the tub. A follow-up when the child was 42 months old indicated a permanent recovery and no new symptoms. Landreth and Read (1942) reported the countercon ditioning of a boy who was frightened of the barber and of having his hair cut. The procedure was very similar to that of the previous study: the child was first given barber play toys in a sandbox where he was then led to "play-barber" and allow other children to cut his hair. He was taken "accidentally" to the barber shop on several occassions for a few minutes at a time and introuuced to the barber. He was then allowed to watch the barber cutting other people’s hair and later he sat in the seat himself. Shortly thereafter he had his first haircut. This 25 procedure required a total of four days and there was no evidence of any new symptoms in a follow-up period. In another study in which counterconditioning was used in a different way, Lazarus & Abramovitz (1962) used "emotive energy" in removing specific anxieties and fears in nine children aged seven to fourteen years. They de fined "emotive imagery" as classes of imagery that aroused feelings of self-assertation, pride, affection, mirth, and similar anxiety-inhibiting responses. The treatment determined the extent and range of the child's fears, and the clinician determined who the child's hero- images were. The child was then asked to imagine a se quence of events which were close enough to his everyday life to be credible, but within which was woven a story concerning his favorite hero or alter-ego. When this was accomplished, the least threatening of his fears was in troduced. Gradually, as each successive fear was desensi tized, or extinguished, the next higher was then intro duced until all previous fears were tolerated without dis tress. Seven of the nine children recovered In a mean of 3.3 sessions and follow-up inquiries up to 12 months later revealed no relapse or symptom substitution had occurred. Counterconditioning has been used in other ways as well. Raymond (1956) used it with aversion therapy with a patient who had a fetish for handbags and perambulators. 26 Treatment consisted of showing the patient a collection of handbags, perambulators, and colored illustrations just before the onset of nausea produced by injections of apo- morphine. This was repeated every two hours, day and night, for one week plus two additional sessions, one eight days later and one six months later. The patient reported the fetish successfully eliminated and also reported an im provement in other aspects of his social and legal life. Bandura (1961) reported that counterconditioning has been used with nauseant drugs, especially emetine. It has been utilized as the unconditioned stimulus in the aversion treatment of alcoholism (Thirmann, 19*19; Thompson & Bielinski, 1953; Voegtlen, 19*10; Wallace, 19*19). Usually eight to ten treatments in which the sight, smell and taste of alcohol is associated with the onset of nausea is suffi cient to produce abstinence. Of 1,000 or more cases on whom adequate follow-up data are reported, approximately 60 per cent of the patients have been totally abstinent following the treatment. Voegtlen (19*10) suggests that a few preventive treatments given at an interval of about six months may further improve the results yielded by this method. Research on Operant — Conditioning Techniques The principles of instrumental, or operant, con ditioning were first systematically investigated by 27 Edward L. Thorndike who concluded, on the basis of his ex periments, that behavior which was followed by a ' ’satis fying” state of affairs was more likely to occur again than behavior which was not. He further concluded that behavior which was followed by a "dissatisfying" state of affairs was less likely to occur again (1911, p. 225). While Thorndike pioneered the use of operant con ditioning, to B. F. Skinner goes the credit for the de velopment of its systematic theory. In his book, The Be havior of Organisms (Skinner,1938), he described the re sults of experiments of bar-pressing with rats. He was the first to make a clear distinction between respondent, or classical behavior, which he defined as behavior which oc curred in response to a stimulus, and instrumental, or operant behavior, which is contingent upon its consequences. (Operant behavior has also been defined as the "totality of operant responses in the behavior repertoire of the animal" [Verplanck, 1957].) Skinner was interested in studying behavior in a "free responding" situation as op posed to the trial-by-trial procedures which had been de rived from classical conditioning procedures and were used by most of his contemporaries (Kimble, 1961, p. 29). He was also the first psychologist to be particularly inter ested in the effects of partial reinforcement upon be havior and to study it systematically. One of these approaches, the free operant technique, 28 is the subject of this study. Ferster, a student and later a co-worker with Skinner, has characterized the free operant method as follows: . . . it refers to any apparatus that generates a response which takes a short time to occur and leaves the animal in the same place ready to respond again. The free operant is used in experiments when the main dependent variable is the frequency of occurrence of behavior. . . . The free operant has advantages in this respect because it removes restrictions on the fre quency with which a response can occur and permits the moment-to-moment changes in frequency. (Ferster, 1953, p. 263) In operant conditioning, then, the fundamental requirement is an increase in the occurrence of rate of response of some behavior as a result of its being followed by some event which, if the change or increase occurs, is called a reinforcer. Originally the response is not under the con trol of any antecedent stimulus. The experimenter waits until the specific behavior occurs and then presents a reinforcer, or, in other words, rewards it. In the case of occurrence of behavior that the experimenter wishes to re move, he determines the stimulus that is maintaining, or reinforcing, the behavior and withdraws it when the unde sired behavior occurs, thus extinguishing it. Interest in the use of operant conditioning techni ques with subhuman behavior has grown steadily since the publication of The Behavior of Organisms and many reports of the successful application of these techniques with animals have appeared (Keller & Schoenfeld, 1950, Skinner, 1959). 29 While operant techniques have been developed and perfected with infrahumans animals, it has been the convic tion of psychologists working in the area that the princi ples being developed would be applicable to human behavior as well. Skinner, in discussing the application of these principles, said: The use of this material often meets with the objec tion that there is an essential gap between man and the other animals, and that the results of one cannot be extrapolated to the other. To insist upon this dis continuity at the beginning of a scientific investiga tion is to beg the question. Human behavior is dis tinguished by its complexity, its variety, and its greater accomplishments, but the basic processes are not therefore necessarily different. Science advances from the simple to the complex; it is constantly con cerned with whether the processes and laws discovered at one stage are adequate for the next. It would be rash to assert at this point that there is no essential difference between human behavior and the behavior of lower species; but until an attempt has been made to deal with both in the same terms, it would be rash to assert that there is. A discussion of human embryology makes considerable use of research on the embyros of chicks, pigs, and other animals. Treatises on diges tion, respiration, circulation, endocrine secretion, and other physiological processes deal with rats, hamsters, rabbits, and so on, even though the interest is primarily in human beings. The study of behavior has much to gain from the same practice. We study the behavior of animals because it is simpler. (Skinner, 1953, p. 38) Recent reviews of the literature indicate the greatly accel erating use of operant techniques in all areas of human be havior and will all types of human subjects (Krasner, 1958; Lawrence, 1958; Salzinger, 1959, Rubenstein & Aborn, I960; Greenspoon, 1962; Eysenck, 1964; Eysenck & Rachman, 1965; Bandura, 1961; Krasner & Ullman, 1965; and Ullman & 30 Krasner, 1965). Studies using Operant Conditioning Techniques This section will review selected studies in three general areas. The first area will include studies gener ally concerned with the alteration of behavior in either laboratory or hospital settings. The second area will re view studies which are primarily concerned with social reinforcement administered by either peer groups or adults in social situations. The third area will review studies which are concerned with the use, or report the use, of operant conditioning techniques in the schoolroom situa tion. Studies concerned with Behavior Modification in Laboratory or Hospital Settings. Studies with Psychotic Patients. The first group of studies to be presented deal with the use of operant techniques to alter the behavior of psychotic patients. In the first paper, Isaacs, Thomas, & Goldiamond (i960) used operant conditioning to reinstate verbal behavior in hos pitalized mute schizophrenics. One patient, 40 years old and classified as a catatonic schizophrenic had been com pletely mute for 19 years. Treatment was by successive approximations starting with the rewarding of an eye move ment with gum, then gradually moving to the reinforcement 31 of lip movements and finally for the elicitation of sounds. As the patient improved he was able to verbalize requests and would do so upon request of other staff members. The authors report that the major problem in the follow-up period was to secure the cooperation of other patients and staff who continued to reinforce non-verbal requests. The second patient, also schizophrenic, was 43 years old and had been mute 14 years. He was treated by a dif ferent experimenter in private and group therapy sessions. The same procedure was used— the patient being rewarded for successive approximations of verbal behavior. This patient finally was able to speak words in group sessions and answer questions but would never speak on the ward. They concluded that while it was extremely difficult to eliminate behavior in the laboratory because it had usually been maintained by partial reinforcement schedules, it was possible to Increase responses rather easily. They felt that treatment should be aimed at whatever normal operant behavior occurred, and an attempt made to increase its occurrence through reinforcement. They viewed normal and abnormal behavior as reciprocally related: "Normal behavior probability can be increased by decreasing pro bability of abnormal behavior, or abnormal behavior can be decreased by the controlled increase of normal behaviors" (p. 11). A series of studies by Ayllon utilized operant techniques with other types of psychotic patients and situations. Ayllon (i960) reported on two female schizo phrenic patients. One, 53 years old, would eat only if the nurse led her to the dining hall, secured a tray, silverware, and then urged her to eat while occasionally spoon-feeding her. Treatment consisted of first discon tinuing all help. The subject was not led to the dining room nor was food allowed to be brought to her. The first four days of the experiment she sat alone and missed all meals. She then started going to the dining room and with in a few weeks was eating without aid. Ayllon reports that the critical factor in the shaping process was that the subject found that she did not eat when she failed to perform adequately. — A second report on a 60 year old schizophrenic woman who would not leave the dining room unless forced or coaxed by nurses gave the same results. In this case, treatment consisted of totally ignoring the subject while in the dining room after she should have left. When she left at the appropriate time she was reinforced with candy and social attention. Within two weeks this patient was leav ing at the appropriate time virtually 100 per cent of the time without being coaxed or forced out by the nurses. Ayllon concluded the following: In these two cases a combination of extinction for social attention and reinforcement contingent upon the desired unit of behavior brought about the desired behavioral changes. Two generalizations from the 33 results appear possible. First, food appears to be the appropriate and most effective relnforcer to use to eliminate behavior deficits associated with eating. Many of the behavior difficulties patients display in the course of eating may be traced to the social rein forcement that attendants, unwittingly to be sure, shower upon such behaviors. (p. 76) Ayllon & Haughton (1962) extended his former study to a group of schizophrenic female patients. The S’s were 32 females seven of whom were regarded as eating problems. Before admission to the experimental ward these patients had been spoonfed, tubefed, or given electro-shock therapy when they refused to eat. The remaining 25 were all long standing schizophrenic patients. The median age was 54 years. The procedure was simple: patients were no longer coaxed, reminded, led, or escorted to eat. They were not spoonfed, tubefed, or subjected to electroshock therapy. Nurses kept completely away from them at mealtimes. If the patient was to eat, she had to go to the dining room within a 30-minute interval following the announcing of mealtime. The door to the dining room was left open for the Interval of five minutes by the 11th week. All patients learned to eat by themselves as soon as it was obvious to them that they would not be fed otherwise. In the next phase of the experiment the same S's were taught to operate a coin-slot machine to gain entrance. A further phase required that they cooperate with each other to get a coin to operate the coin-slot machine. They reported that no eating problems or other problems were seen to appear after the experiment. In another study, Ayllon & Michael (1959) trained ward nurses to administer operant reinforcement to psycho tic patients. Of the 19 patients used, 14 had been classi fied as schizophrenic and five as mentally defective. All except one had been hospitalized for several years. No S’s were receiving any other treatment at the time of the study. S's were systematically observed to determine base lines for maladaptive behaviors. After the observation period nurses were given instruction on the treatment pro cedures. Behaviors that were to be eliminated were either extinguished by withdrawing all approval or reinforcement or attention when they occurred, by strengthening weak or incompatible behavior, by extinguishing by escape and avoidance conditioning, and by a combination of the above, depending upon the patient. Results required from six to 11 weeks. In five cases of strong behavior that were treated by extinction, three were successful, and two partially so. In one S, strong behavior was successfully eliminated by strengthen ing a weak behavior that was incompatible. In two patients weak behavior was strengthened by avoidance and escape con ditioning. These two cases were women who would not feed themselves. Nurses were instructed to spill food on their dresses as they fed them and also give social reinforce ment when S's fed themselves. One S was released from the 35 hospital shortly after the successful experimental period since her inability to eat had been the reason for her admittance. The other patient relapsed after 5 weeks. Four male patients who had hoarded magazines for several years previous to the experiment were "cured" by the nurses who simply refrained from taking the magazines away from them while providing plenty of magazines to steal and socially reinforcing all non-hoarding behavior. Other patients were treated with mixtures of the above methods. Ayllon & Michael report that much care needed to be used in instructing the nurses, particularly those who were clini cally oriented. This study has much value in the consider ation of teaching school teachers how to apply operant conditioning principles in the classroom. In the final study by Ayllon (1963), he applied operant conditioning techniques to extinguish three types of psychotic behavior in a 47 year old chronic schizo phrenic woman. In reviewing the literature of previous studies, Ayllon says that from the point of departure of Lindsley’s (1956) study of operant conditioning of humans, the "behavior of mentally defective individuals (Orlando & Bijou, I960), stutterers (Flanagan, Goldiamond & Azrin, 1958), mental patients (Hutchinson & Azrin, 1961), autistic (Ferster & De Myer, 1961), and normal children (Bijou, 1961, Azrin & Lindsley, 1956) has been shown subject to the same controls" (p. 53). His study reports an extension of these principles to the intensive individual treatment of three separate psychotic behaviors in a single patient. The first experiment was with food stealing by the S. Formerly the S was persuaded, coerced or coaxed not to steal or to return stolen food. Treatment consisted of discontinuing these methods and removing the patient from the dining room immediately whenever she approached a table other than her own, or picked up any unauthorized food. This usually resulted in her missing most or all of a meal. Results showed that stealing was eliminated in two weeks. She also lost weight from 230 to 180 pounds. The second experiment with the same S was concerned with the hording of towels which the S had done for nine years. She had from 19 to 29 towels at all times over a six week pre treatment period. Treatment consisted of giving her more towels each time she was in her room. When 625 towels were accumulated she began to remove them herself. Treatment was discontinued and during the next 12 months the mean number of towels found in her room at any one time was 1.5. Ayllon says that this was satiation of a reinforcer and concludes that a reinforcer loses its effect when an ex cessive amount is made available. In the third experiment, the same S wore approximate ly 24 pounds of excess clothing at all times and the aim was to get her to reduce that figure to the norma] amount. This was done by having her weigh herself before a meal. 37 She was always told that she weighed too much and could not eat until she lost weight. With each weigh-in she was required to weigh slightly less than she had at the pre vious meal. Within 12 weeks she was dressed normally. This experiment was an example of a response that is fol lowed by a reinforcement. In this case, the reinforcement was food which was given for the removal of excess cloth ing. This study is also a good example of countercondi tioning because excess weight was made incompatible with eating and was replaced with less weight, in the form of clothing. Studies with Autistic Children. This section will review some of the recent studies using operant condition ing techniques with autistic children. This issue has been a controversial one because several investigators have argued that the failure of autistic children to develop normally is based on their failure to be affected by their social environment, or more specifically, by social stimuli (Ferster & De Myer 1961; Lovaas, Freitag, Kinder, Ruben- stein, Schaeffer, & Simmons, 1964), while others have maintained that autistic has an organic basis (Rimland & Yannet, I960). According to Bijou & Baer (1961), it is generally agreed that a parent's response to his child's behavior is one of the major factors determining the child's develop ment. They believe that a parent responds to his child by presenting and removing stimuli which are contingent upon the child's behavior, i.e. he rewards and punishes the child. Then, as the infant matures, the parent relies mainly on socially reinforcing stimuli. These stimuli have aquired their reinforcing power through association with primary reinforcers. Thus, to affect the child, these stimuli necessarily must acquire meaning for him, and the child who for some reason remains unaffected by the presentation of these stimuli will be deficient or lack social and intellectual development. Ferster & DeMyer (1961) have attempted to analyze how the basic variables determining a child's behavior might operate to produce the particular behavioral deficit of the autistic child within the reinforcement theory framework. They consider the major performance deficits of the autistic child to be in the area of social control, particularly with behaviors which have their principle effects through the mediation of other people. They pos tulate that the autistic child's major mode of social con trol is through behaviors which are disturbing to others. Behavior, they believe, which is normally maintained by its effect on the parent, is likely to be weakened or re moved by extinction or intermittent reinforcement, if parental reinforcement is a function of variables which are not controlled by the child. For example, a child might verbalize a request which would pass by unreinforced because of the parents preoccupation with another person or acti vity. Ferster & DeMyer believe these behaviors are most vulnerable because a child spends the first few years of his life with his parents, and speech and other social behaviors are the most vulnerable aspects of the child’s repertoire. In attempting to account for the strong and inappropriate repertoire of the child, they hypothesize that the autistic child had been established as a condi tioned aversive stimulus for the parents. The child’s behavior may have been incompatible with the adult's nor mal and preferred repertoires. For example, since the mother had to stay home and care for the child, she was unable to work or socialize away from home. If the parents finally did try to escape the annoying behavior of the child, the lack of reinforcement of a previously reinforced response, such as a tantrum, would have been followed by an increase in the intensity of the behavior. If the parent then reinforced the behavior to eliminate the annoyance, both the probability of its occurrence and its resistence to extinction would have been increased. Lovaas, Freitag, Kinder, Rubenstein, Schaeffer, & Simmons (1964) attempted to determine whether autistic children could, in fact, be operantly conditioned to acquire social reinforcers. They studied the establishment and maintenance of social rewards as reinforcers for certain behaviors of two schizophrenic children characterized as autistic. The experimental procedure consisted of two phases. The first phase was concerned with the establish ment of the social reward as discriminative for food. This was necessary because simply pairing a social stimulus with the presentation of food was ineffective because autistic children normally will not attend to social stimuli. The child had to attend to the reinforcement before it could become discriminative for food. Thus, the first phase con sisted of reinforcing the child when he was attentive dur ing the presentation of a social reward. The second phase involved the delivery of the social reinforcement, contin gent upon a bar-pressing response, as a test of any rein forcing properties it had acquired during the first phase. The results showed that the social reward did acquire reinforcing properties for the children. Also, as long as the social stimulus was maintained as discrimina tive for food, it showed no sign of losing its acquired reinforcing properties. Hewett (1965) applied operant techniques to the teaching of speech to autistic children. Both positive and negative reinforcements were administered in a special teaching booth. A booth was constructed with a shutter division between the S and the E. The only source of light in the room is on the teacher's side and when a correct response is given by the child the shutter remains open and hi he can view the teacher. Additional positive reinforce ment is had in the form of candy, a spin on the stool where he is sitting, or perhaps music played along with the show ing of a cartoon movie. At the onset the child is kept in semi-isolation and all meals are given in a contingent man ner in the booth. Social imitation and speech training thus are taught under these conditions. Hewett reports that results have proven the technique as highly successful. (The study was still in progress at the time of the re port,) In other studies with autistic children Salzinger, Salzinger, Pisoni, Eckman, Mathewson, Dutsch, & Zubin (1962) attempted to socialize a four-year-old boy without speech. After a nine month period they report that the boy had increased his vocalizations and was able to speak at least a dozen words. Perster (1961) and Perster & De Meyer (1961) have studied performances of autistic children in an operant behavior environment. They have "shaped" behaviors in relation to specific objects such as vending machines, record players, pinball machines and a trained pigeon, and their study is most valuable for the information obtained as to the effectiveness of different training schedules and reinforcers. Studies of Deviant Behavior in Children and Adoles cents . The next group of studies in this section will be concerned with the use of operant conditioning techniques with children and adolescents having some type of deviant behavior. Williams (1959), in what is now considered a classic study, described the treatment of tantrum-like behavior of a 21 month old boy. After a prolonged illness the child had come to expect special care and "to enforce some of his wishes, especially at bedtime, by unleashing tantrum behavior to control the actions of his parents" (p. 269). It was necessary for one parent to spend one- half to two hours each bedtime waiting for the child to go to sleep before leaving his room if a tantrum was to be pre vented. Treatment consisted of putting the child to bed, and then leaving the room and closing the door. The temper tantrum dropped from 45 minutes the first night to not at all on the fourth night. After recurrence following the visit of an aunt, the same procedure was followed and the tantrums extinguished in the same manner as before. The treatment involved only the removal of the reinforcement (parent attention) to stop or extinguish the behavior. The findings are in accord with those of Ayllon, et. al. (I960; 1962; 1963) and other studies reporting a decrease in be havior whenever the social stimuli in the form of attention by others maintaining it was removed. Baer (1962) studied the effects of withdrawal of reinforcement on three five-year old male subjects. All had long histories of thumb sucking. In the first experiment a single S's thumbsucking was extinguished by turning off 43 a cartoon show whenever his thumb was in his mouth. In the second experiment two boys watched the same cartoon but were separated by a pannel so that they could not see each other. The E would turn off the cartoon or turn it on in accordance with the behavior of one of the two boy's thumb- sucking, i.e. a yoked design. It was found that the boy who was being reinforced for not thumbsucking quickly ceased sucking while the cartoon was on but that the other boy, who was in effect randomly reinforced, showed no im provement. Baer felt that these results indicated that thumbsucking may be interpreted as a learned response which was modifiable by environmental control using stimulus con sequences of the response. These findings are supported by Ullman & Krasner's (1965) general principle that "all be havior modification boils down to procedures utilizing systematic environmental contingencies to alter the subjects response to stimuli" (p. 29, italics mine). Two studies using operant conditioning with delin quent boys also are reported. In the first study Schwitz- gebel & Kolb (1964) used "shaping" techniques and token reinforcement to get boys to come to a laboratory. The boys were from 15 to 21 years of age and all had criminal records and had served terms in reformatories. The shaping of arrival time for appointments was done with reinforcers such as money, cokes, occassional sandwiches, etc., which were awarded when the subject was on time or his arrival 44 was closer to the set time than previously. Reward was on an intermittent schedule. Verbal praise was also used with concrete reinforcement. The boys were then paid to talk into a tape recorder about anything they wished in another part of the experiment. The results of the shaping showed that by 15 appointments, arrival time had become dependable for 90 per cent of the boys. This was considered very im portant by the authors because one of the reasons they found these boys "unemployable" was their inability to con sistently arrive at their jobs on time. In a second study, Schwitzgebel (1965) recruited 35 male adolescent delinquents to participate in a study to determine what effect a particular type of interpersonal relationship based on operant conditioning might have on subsequent behaviors. There were 48 S's, all with police records. Their mean age was 16.2 years and they had a mean of 1.4 years incarceration. The operant conditioning technique used with one group was to positively reinforce statements of concern about other people and to reinforce promptness. Group II was negatively reinforced for hostile statements and posi tively reinforced for socially desirable non-verbal be havior that gave evidence of employability. The control group received no overt reinforcement except pay for the initial interview. The task was to talk into a tape re corder, with E giving appropriate response according to the 45 experimental group of the talker. In the first group par tial reinforcement was also used to shape prompt arrival. Results showed that the group I subjects were significantly more prompt in arrival and had a significantly larger in crease in positive statements than group II. Negative re inforcement of hostile statements of group II was not significantly effective, though a trend showing improve ment in social behavior was indicated. There was improve ment in social behavior but it was not significant. Schwitzgebel concluded that the explicit use of social reinforcers was both effective and acceptable with adolescents. Another interesting finding with bearing upon the present investigation was that the number of reinforce ments required was much smaller than might be expected. Studies using Social Consequences as a Reinforcer. This section will review selected studies which had as their main point of investigation the use of social consequences, whether by an adult, adults, or the peer group, to reinforce adaptive behavior. Patterson & Littman (1965), in reporting the find ings and implications of work done over several years have stated: What is being suggested here is that the social be havior of other people does not function as a general ized discriminative stimulus for modeling nor as a generalized reinforcer. The amount of control exerted over the behavior of the child varies as a function of 46 the agent who provides these stimuli. There is not a general "trait" of responsiveness to social reinforcers. The general trait concept would most likely assume that the child is first conditioned to respond to the social reinforcers or to model some behaviors of the mother; this responsiveness would in turn generalize to respon siveness to other adults and to the peer group. We, on the other hand, believe that conditioned responsiveness to the mother is a necessary, but not sufficient stage. If conditioned responsiveness to the mother does not occur then it is unlikely that other social agents will have control over the child; e.g., the autistic child. If, on the other hand, the child is conditioned to re spond to the mother, this will not necessarily lead to a situation where other social agents have control over the behavior of the child. We hypothesize that the child’s responsiveness to each class of social agents occurs as a function of the quantity and quality of his interaction with members of that class, e.g., a child must be conditioned to respond to the reinforcing and modeling behaviors of the peer group. Once conditioned to respond as a result of his interaction with a few members of this class, the responsiveness tends to generalize to other members of the class. . . . We are assuming that in order for the child to display the wide range of social behaviors which we label as "nor mal child" that he must be conditioned to respond to all major classes of social agents. To the extent that he is selective in his responsiveness, he will be de ficient in those social behaviors normally shaped by the class of agents which he is "selecting out." . . . We are offering then, two general explanations of the production of deviant behavior in children. First, the subculture may provide social reinforcers for deviant behavior and thus directly shape the very behavior to which it objects. Or we may have a four step procedure described in the preceding paragraph: (1) selective non-responsiveness (2) omission in acquisition of certain social behaviors (3) agents avoid/punish the child (4) deviant behavior, (pp. 5-6) It may well be that professional training in the "helping" sciences exacerbates this Christian decency and lowers the threshold; after such training the in dividual can dispense positive reinforcement (love) in the face of repeated and highly aversive child be haviors. (p. 7) Three studies on the use of social reinforcement are reported next. In the first, Hart, Allen, Buell, Harris, 47 Florence & Wolf 0964) extinguished operant crying In two four year old boys. Both boys cried frequently during school period. A baseline of crying frequency during school periods was established for both. It was noted that when either boy cried there was usually an adult around. Both were often seen to glance at the teacher before starting to cry. The operant technique applied consisted of with drawal of all reinforcement during the period of crying. When either boy cried the teacher turned her back and com pletely ignored the child. Both received attention and praise when they were not crying. After extinction the behavior was reinstated and with each child operant crying was quickly reestablished at its previous level. Extinc tion was again applied in the final phase of the experi ment. Both S*s had cried from five to 10 times a morning. With the first S, crying was reduced to from zero to two times a day and with the second, to less than two times a day. The authors concluded that frequent crying may be largely a function of adult attention. In the second study, Harris, Johnston, Kelley & Wolf (1964) used an operant conditioning technique to develop walking behavior in a nursery school child. The S was three years old and had previously walked but would only crawl on the floor when anyone other than her imme diate family was near her or when she was in her nursery school. The purpose was to see whether positive social 48 reinforcement by the teacher could be used to help a three year old use her already well established walking behavior. The child had shown strong withdrawal behavior when first placed in the nursery school. She crawled about the room and crouched in out-of-the-way places. This behavior was also evident at home whenever visitors were there. She was estimated to be off her feet 75 per cent of the time or more in the school. Treatment was similar to the previous study; it consisted of withdrawing all social reinforcement when the child was off her feet and instantly rewarding (or reinforcing) her any time she stood up. Within one week she was on her feet most of the time and within a month her behavior was indistinguishable from that of other children. She was then reversed and within two days was again off her feet continually. Reinforcement for on-feet behavior quickly reinstated standing behavior. In neither study were any unfavorable symptoms observed as a result of the experiments. In the third study (Allen, Hart, Buell, Harris, & Wolf, 1964) similar techniques were used to change the be havior of a four year three month old girl who spent most of her time by herself and interacted with teachers very little. Less than 10 per cent of her time was spent with other children. Treatment consisted of: (1) maximum at tention when she was with other children; (2) no attention when alone; (3) a minimum of attention when she contacted adults unless with another child. In less than two weeks she was spending 80 per cent of her time with other child ren. The authors discuss these results with respect to common guidance principles such as "Encourage him to play with other children" (p. 517). These, they say, imply giving attention when undesirable attention is occurring and hence instead of extinguishing it and increasing the likelihood that the child will play with other children. Such attention will increase the probability of his remain ing isolated. The three studies above are particularly pertinent to the present investigation and to teaching in general be cause they suggest methods by which the regular or special teacher may help students develop adaptive behaviors within the regular course of teaching. In what Cohen (1962) describes as the first operant study of human social behavior, Azrin 8 c Lindsley (1956) determined whether cooperation could be developed in an operant setting. The S’s, 20 children, seven to nine years of age, were formed into 10 two-member teams. During a "game," pairs of children were seated at opposite sides of a table with a stylus and three holes in front of each child. After the experimenter had ascertained that the children knew how to insert their styli into three holes in front of them, he merely told them that from time to time a jelly bean would fall into a cup placed midway between them and then left them alone In the room. During the experi menter’s absence, the children were rewarded with a jelly bean whenever they inserted the styli in corresponding holes within 0.04 seconds of one another. All teams learned to cooperate, even without specific instructions, within the first 10 minutes of play. Leader-follower re lations developed, yet the children in eight of the 10 teams almost immediately began to share the candy they received. Equitable division of the candy occurred in the remaining two teams after some initial verbal dispute. Following the reinforcement period, no more candies were dispensed, and the cooperative behavior then gradually ex tinguished. The authors concluded that Operant conditioning techniques can be used to develop, maintain, and eliminate cooperation between children without the use of specific instructions concerning co operation. The rate of a cooperative response changes in much the same way as a function of single reinforce ments as does an individual response. (p. 102) In a follow-up study Cohen (1962) studied the effects of the free operant method on Justin, a normal 13 year old boy. His social behavior was studied in relation to people with whom he came into contact in his daily life. The ex perimental situation was two separate rooms separated by a plexiglas window. Cooperation on the Lindsley plunger by Justin and the subject in the other booth led to a reward while inappropriate pulls led to punishment— a darkened room and no reward. Trials were run with Justin and his 51 brother, a friend, a stranger, his sister, and his mother. All responses were recorded and evaluated automatically so that the order of plunger-pulling, rewards and punishment could be studied. Analysis of the data showed that Justin had led those people with whom he had had previous non-ex- perimental experience of leadership: his brother and his freind. His mother and sister, who led him in non-experi- mental experiences, exhibited resistance to following him in the experimental situation. In general, the results ob tained in the experimental period were very similar to those observed in actual life. Cohen says of this experiment that: The experimental determination of five different dy namic patterns of cooperative and competitive leader ship between a young man and five other persons clearly demonstrates the sensitivity of this free operant method to important social variables. The close similarity between the experimentally measured patterns and the extra-experimental relationships as determined by questionnaires and interviews demonstrates that these experimental measures have high validity. The method permits a laboratory analysis of a child's social world. (p. 717) Lott & Lott (I960) reported on the "cohesiveness" of a group of 48 children, 24 from grade three and 24 from . grade five. The children were divided into 16 three member teams following a sociometric test to determine desired and undesired classmates. They were then teamed with those children they had not chosen to play with and asked to play a game of taking a rocket ship to three different planets. Half the teams were successful in completing the journey 52 and were rewarded. The other half failed and were not re warded. Following the game the subjects were again given a sociometric test and a significant number-of the success ful groups selected previously non-selected members of their team. The authors report that the findings indicate that positive attitudes toward persons will be developed as a result of receipt of reward in their presence, and that though the results were obtained with children, there was no reason to expect that they would not occur with adults as well. This study, while not conceived as an attempt to demonstrate operant principles, was included in this review because of its relevance to the present study in which the entire class is rewarded for group efforts. In other studies, Spadlin & Girardeau (1966), in their review of experimental research with retarded child ren report that social consequences also have been used in operant studies and instrumental studies with retarded per sons (Barnett, Pryer, & Ellis, 1959; Hollis, 1965a; 1965b; Lock, 1962). Harris, Wolf, & Baer (1964) have demonstrated that the attention of certain adults can serve as an ef fective reinforcer for several types of behavior emitted by nursery school children. Spadlin & Girardeau concluded that: Our casual observations indicate that its effects with severely-retarded children are quite unpredictable. This is not to say that social reinforcers will not work with some severely retarded children. Rather, they will work only with selected children and, even 53 then, their effects may be limited or unpredictable. Social comments have limited value as reinforces due to the fact that they depend so much on the unique history of the child. (p. 267) Studies using Operant Conditioning Techniques in the Classroom. This section will review studies in which operant techniques were applied in the classroom setting. While, as in the previous section, the reinforcement is usually peer group or teacher-administered social approval, the emphasis here will be in its utilization in the classroom setting. In a review of studies on social factors, Patterson & Ebner (1965) reported that: It is the social environment which plays a major role in shaping and maintaining deviant behavior in children. . . . Our data from other observation studies are quite clear in implicating the social environment as an impressive dispenser of social reinforcers for all manner of deviant behaviors. One study, Patterson, Bricker, and Green (1965) showed that the pre-school peer group provided massive reinforcement schedules for aggressive behavior. In another series of studies Patterson (1963) Furniss (1964) the data showed Immed iate social reinforcers were provided by delinquent girls for the majority of occassions in which deviant responses occurred. In the present study, a preliminary look at Ebner1s data showed that 10 to 30 per cent of the "hyperactive" behaviors are reinforced positively by the teacher or the peer group. Zimmerman (1965) has also provided data which show that trained aids in a residential treatment setting reinforce a variety of deviant behaviors in their child patients. Taken to gether, all of these studies indicate that the social environment Is undoubtedly contributing to the main tenance (and perhaps acquisition) of deviant be haviors. It wou-14 n°t involve too much of an extra polation to also assume that any treatment program which has dramatic effects upon the deviant behavior of the child must in some way alter the program of reinforcements provided by the social environment. (pp. 11-12) Zimmerman & Zimmerman (1962) used operant condi tioning techniques with two 11 year old (in special class) boys. The first experiment was with a boy who could not spell and who demanded an excessive amount of the teacher’s time and attention. He was also a behavior problem in the classroom. The operant technique applied was to completely withdraw any attention when maladaptive behavior occurred. This procedure was first done with spelling. When asked to spell a word on the board the boy would spell it many incorrect ways, waiting for the teacher to coach him. The teacher sat at his desk, busy with other work, until the boy had finally got the word correct whereupon he would then give the boy social approval and attention. Within a month spelling errors were reduced to almost zero. To shape other classroom behavior attention in the form of smiling, talking or physical proximity was given only after the emission of desired behavior. Maladaptive be havior had declined significantly at the end of a month of treatment. The second student was a boy who displayed severe temper tantrums and often had to be carried to class by attendants. Treatment was similar to that of Williams (1959), and Hart et al. (1964). The operant procedure was to completely ignore and not reinforce any behavior during 55 a temper tantrum. The teacher sat at his desk and waited until all crying had stopped and then aided the boy. The first day the tantrum lasted approximately three minutes, then a crying spell of five minutes more occurred before the boy was given any attention. The tantrums extinguish ed "within a few weeks." The teacher reported other be havior also improved as the tantrums decreased. Bijou (1965), in the main address to the 43rd Council for Exceptional Children convention reported on the results of studies using operant techniques with retarded children in a special classroom over a three year period. He reported that his results show that with operant con ditioning techniques the retarded child can learn academic subjects far beyond what was formerly thought possible. The teacher in this situation is viewed as one who is at tempting to bring forth strengths and maintain desired be havior. He states that the optimum classroom situation is one that is arranged so that the child develops skills, gains new knowledge, and learns how to learn better. This is best accomplished by giving (a) immediate knowledge of results; and (b) programmed instruction techniques. His teaching philosophy and methods were all programmed to meet these ends (Birnbrauer, Bijou, Wolf, Kidder, & Tague, 1965; 1966). One of the studies in the series discussed by Bijou & Baer that has particular relevance to the present investigation was a study of token reinforcement for be havior by Birnbrauer, Wolf, Kidder, & Tague (1965)* The subjects were mentally retarded special class children eight to 14 years of age and 50 to 70 IQ. The class pro cedure had been to reward performance in an experimental programmed classroom with knowledge of results, verbal approval and tokens. The tokens were always given with verbal comments of approval and consisted of check marks that could be exchanged for candies, toys, or money at the end of the day. To determine that token reinforcement was essential to the relatively high levels of accuracy and rates of studying maintained by the retarded pupils tokens were not dispensed for a period of at least 21 days and then were reinstated. Daily records of items completed, percentage of errors, and disruptive behavior were kept. During the no-token period three general patterns of re sults were obtained: (1) Five of the 15 pupils showed no measurable change in performance. (2) Six pupils increased either markedly in over-all percentage of errors or suffi ciently to reduce progress in the programs. (3) Four pupils showed an increase in percentage of errors, a de cline (or considerable variability) in amount of studying, and an increase in disruptive behavior. Baseline perfor mance was recovered in these 10 pupils when token rein forcement was reinstated. Other support for the above results that sometimes verbal reinforcement alone is not adequate are the results of the study of Levin & Simons (1962a, 1962b). They found that praise alone was actually averslve for some emotion ally disturbed children and did not serve as an accelerat ing consequence. When given only verbal praise for a specific approved behavior, it was found that these child ren emitted drastically fewer responses. The authors stated that using praise as an accelerating consequence is a goal and that teachers must attempt to modify child ren's behavior so that it will come under the control of normal consequences. This effect is what Birnbrauer et al. (1965) were demonstrating— that part of the training of special children must consist of coupling concrete rewards that are known to be reinforcing with verbal rewards that are neutral or even aversive and then arranging the environ ment so that verbal rewards become reinforcing through stimulus generalization. Quay, et al. (1966), addressing themselves this topic said: . . . It therefore appears that in the early stages of retraining the aggressive, unsocialized child, it may be necessary to utilize reinforcers of a fairly con crete nature, such as candy, trinkets, and toys. How ever, the dispensing of more primary rewards ought to be paired delibertly with social reinforcers, such as praise and approving gestures, to facilitate the de velopment of responsivity to more usual reinforcers. (p. 512) He devised an experiment to operantly condition five child ren in the same classroom to pay attention to the teacher. 58 After determining a baseline for maladaptive behavior, a box was placed on each child's desk. The children were told that if they were paying attention to the teacher a light would go on from time to time and they would receive one piece of candy for each light flash at the end of the period or activity. Reinforcement was on a fixed ratio schedule of 1:5. Results after 52 days "suggest that this method is, indeed, having some effect" (p. 513). Whether any other reinforcement such as teacher approval was given was not indicated. In another part of the same experiment a child was taught the alphabet in 12 10 minute sessions. Here the child was shifted from a primary reinforcement— candy, to a symbolic reward— poker chips. The chips were exchanged later for candy. It appears to this writer that the first part of the experiment would have yielded better results had the reinforcement schedule been 100 per cent at the start and then switched to the ratio schedule once the behavior was under the control of the E. The final study to be reviewed was done by Patter son, Jones, Whittier, & Wright (1965). This study is of particular importance because the present investigation is an attempt to extend the operant techniques used there with a single boy to an entire classroom of disturbed children. Patterson et al. hypothesized that there was a curvilinear relationship between the activity levels of a child and the acquisition of social behavior; that up to 59 moderately high levels of activity the child’s behavior will bring about acceptable responses from other people but that extremely high rates of behavior are aversive to other people, hence the reactions from the culture are likely to be punitive. In this situation the child exhibiting hyper active behavior would be likely to be punished even when he is displaying socially acceptable behavior. The two subjects were both 10 years old and both had medically diagnosed brain damage. Both were also de scribed as very hyperactive. One subject was the experi mental S and one the control S. Observations of the be havior of the two children were made in the classroom set ting, from an observation booth adjoining the classroom and provided data on the frequency of occurrence of the following high rates of responses: walking, talking, dis traction, wiggling. Each child was observed for a minimum of ten minutes a day, four days a week. Following several weeks of baseline observation, the conditioning procedure was begun with the experimental subject in the classroom. During each time interval in which one of the high rate re sponses did not occur, S received an auditory stimulus which had been previously paired with the delivery of candy and pennies. The stimulus was dispensed by a radio device through an earphone worn by the subject. At the end of each conditioning trail, S received whatever candy or pen nies he had "earned.” The control subject showed no 60 significant change in frequency of occurrence of the high rate responses during the three month period, while the experimental subject showed a significant decrease in non attending behaviors. This reduction in rate was maintain ed over a four-week extinction period. In their discussion they hypothesize that the im portance of any change in behavior lay in the effect it produced on the reactions of the social culture. They be lieved that in future programs it would be important to arrange a hierarchy of responses which would lead to the child being ignored or punished by the social culture for emitting maladaptive behavior. A provision of just such a social culture was one of the purposes of the present in vestigation. Summary This chapter reviewed two areas of the literature dealing with the application of learning theory to the mod ification of behavior of human subjects. The first area reviewed was the use of behavior therapy based on the principle of reciprocal inhibition or counterconditioning. Treatment based on these techni ques was shown to aim at replacing one behavior with another. Thus, if strong responses which are incompat ible with anxiety reactions can be made to occur in the presence of the anxiety evoking cues, the incompatible 61 response would then become attached to the cues and the anxiety response weakened or extinguished. Studies were reviewed in which this method was used to treat anxiety and other maladaptive behavior in humans. It was shown that the counterconditioning method has been used successfully to reduce the fear of physical objects (Jones, 1924; Lazarus, I960; Bentler, 1962; and Landreth & Read, 1942); in group therapy sessions to re duce anxieties (Lazarus, 1961) and extinguish a fetish (Raymond, 1956). These studies were all examples of the implementa tion of the basic principle of reciprocal inhibition— sub stituting a more acceptable behavior for a maladaptive one. In each case the subjects were placed in pleasant, non threatening situations, where they were gradually exposed to the one which had been anxiety provoking. This pro cedure was repeated until the cue no longer could evoke anxiety in the patient. Haring & Phillips (1962) extended this concept to the school situation, creating an environ ment for emotionally disturbed children that provided carefully planned success in school work in a reassuring atmosphere. They hypothesized that the children then would substitute feelings of liking and approval of the school that would replace fear and distrust of the school environment that, in part, caused their placement in the special program. These authors used several 62 psychologists and other school personnel working in close conjunction with the parents of the experimental class chil dren. They reported significant findings but left unanswer ed the question of whether the individual classroom teacher could effectively reduce the anxiety and dislike or hatred of school manifest in many "emotionally disturbed" children without expensive outside aid. The second major section reviewed the theory of the free operant technique as developed by Skinner (1938: 1953) and Perster (1961) . The free operant technique was shown to be concerned with the systematic reinforcement (reward) of a desired (adaptive) behavior to increase its occur- a rence. The effectiveness of the reinforcement was deter mined by noting whether or not there was an increase in the reinforced behavior. While the operant technique has been used exten sively with subhumans, its use with humans was shown to have been of rather recent origin, with most work being done in the last 25 years. A review of studies in which these techniques was used has shown that it can effectively be applied with psychotic patients to reinstate verbal be havior (Issacs, et al., I960); to shape eating and other behavior (Ayllon, I960; Ayllon; 1963). Operant techni ques were used with autistic children (Ferster, 1961; Lovaas, et al. , 196*1; Hewett, 1965) and shown to have had limited but encouraging success. They have been used 63 successfully to modify the behavior of adolescents and children (Williams, 1959; Baer, 1962; Schwitzgebel & Kolb, 1964; Schwitzgebel, 1965). Social reinforcers, such as verbal praise and the attention of peer groups or adults, have been used to in crease adaptive behavior (Hart, et al., 1964; Harris, et al- 1964; Allen, et al., 1964). Azrin & Lindsley's (1964) study showed that cooperation could be developed between children through the application of operant reinforcement by a third party to the interaction of the subjects. Lott & Lott (I960) showed that children can be conditioned to develop favorable attitudes toward other children with whom they are successful in a common game or activity. In the classroom, operant conditioning has been used to a more limited extent. In one study (Zimmerman & Zimmer man, 1962) a single boy was conditioned to spell properly. Bijou (1965) reported the use of operant methods with child ren in special classrooms for the mentally retarded and stated that these techniques were extremely successful. His method has somewhat limited application to the public school because, as in the Haring & Phillips (1962) study, there were several highly trained personnel with a very small group of children. Another study (Quay, et al., 1966) used individual reinforcers with each child in a classroom to shape attending behaviors and Patterson, et al. (1965) reported the successful conditioning of a single 64 hyperactive child in a special classroom. This review of literature has shown that counter conditioning can be used to replace one response with another. The free operant method is one way in which a specific behavior can be increased or decreased. The oper ant conditioning technique has not been used to condition an entire group of human subjects at the same time with a common reinforcer since, while it has been used with groups, it has been used to modify the behavior of specific indi viduals within the group. The next step would appear to be to extend the principles of operant conditioning and count erconditioning to entire groups to determine whether it is possible to extinguish and replace maladaptive behavior and anxiety with adaptive, non-threatening behavior. CHAPTER III METHODOLOGY The purpose of this chapter is to present a descrip tion of the subjects and material used and to outline re search design and procedures of this investigation. The Subjects The study was done in a lower middle-class neigh borhood elementary school in North Redondo Beach, Califor nia. The school, Franklin Elementary School, had an en rollment of 458 children in grades Kindergarten through 6 at the time of the study. There were 26 children enrolled in three classes for educationally handicapped children. The educationally handicapped children were not mixed with regular students for any playground activities, assemblies, or other school functions. The 16 children included in this study were all the students in two of the three classes for educationally hand icapped children at Franklin School. The classes had to be used as they were found; it was not possible to match stu dents in the experimental class with contrast class students and no inter-calss rearrangement of children could be done for the purposes of this study. 65 66 The experimental class was chosen because it was considered to be composed of the most severely disturbed children of all the 13 classes for educationally handicap ped children in the South Bay Area School District, a co operative special education district of which the Redondo Beach School District is a member. It was the opinion of both the guidance and administrative staffs of the dis tricts that this particular class presented the most serious behavioral and classroom control problems of any education ally handicapped class in the South Bay district. The con trast class was believed to be a "typical" group of educa tionally handicapped children with an excellent, above- average, educational program. The children had been placed into the Educationally Handicapped Program in strict accordance with California State Law as specified in Assembly Bill 464, 1963. This Code defines as educationally handicapped any child: . . . who by reason of marked learning problems, or be havioral problems or a combination thereof, cannot re ceive the reasonable benefit of ordinary education facilities; but who are not also handicapped by mental retardation or obvious physical disabilities for which special programs are already provided elsewhere in the Code. (Education Code Section 6750, p. 336) Administrative Code Section 221 further defines an EH minor as: . . . a minor described in Education Code Section 6750 (above) whose learning problems are associated with a behavioral disorder or a neurological handicap or a com bination thereof, and who exhibits a considerable dis crepancy between ability and achievement. (p. 118.15) 67 The IQ range of the experimental class was 88 to 119; the contrast class was 84 to 134. The age range of the exper imental class was 8.8 to 10.6 years; the contrast class was 7.3 to 10.10 years. Older children were two or more years retarded in learning; younger children had been non-achiev ers before placement. Because of the nature of the study and the small number of students involved, a brief description of each student is included. A summary of class characteristics of both classes is also provided. These data are presented in tabular form in Tables 1 and 2. Subjects in the Experimental Class Subject E-l— was found abandoned by the side of a road in Alabama. He apparently had not been fed from the time of his birth until several hours later when he was found. His medical report states that he was near death at the time of discovery and was suffering from malnutrition. He has had two additional step-fathers since he was adopted by his mother. He was nine and one-half years old at the time of this study. A medical evaluation indicated "cere bral dysrhythmia" and recommended medication which he had not been given. The psychological evaluation reported a "lack of impulse control and hyperactive, distractive be havior." Reports from his former teachers all reported hyperactive behavior. In September, 1964, his IQ was 91 on 68 the WISC. His academic level as measured on the Wide Range Achievement Test (WRAT) was at a low second grade level. He was placed in the EH program in November, 1964. This boy was scheduled to be placed on home teaching and excluded from school because he was usually uncontrollable in the EH class and was making no progress. He was retained four more weeks for purposes of this study. Subject E-2— was a nine and one-half year old boy who had been under medication because of hyperactivity since his entrace into kindergarten. His psychological evaluation states that he "has been and continues to be a chronic disturbance in the classroom and a rough, disrup tive force on the playground- . . . In general, he is hyper active, restless, easily distracted and impulsive." He scored 97 on the Stanford-Binet Intelligence Test (S-BIT) in April, 1963- His scores on the WRAT in February, 1965 placed him at a low first-grade level academically. He was placed in the EH program in April, 1965. Subject E-3— was seven years and eight months old at the time of this study. His psychological evaluation states that he is inattentive and has difficulties following direc tions. There is evidence of hyperactivity and con siderable hostility. He is constantly in difficulty with other children and is a persistent disruptive force in the classroom. [S] has had learning and be havioral problems since Kindergarten. . . . He is Hyper active and explosive in his behavior. There is 69 evidence that he is both visually and auditorily imma ture in his perceptual development. He scored an IQ of 99 on the WISC in December, 1965. His academic rating on the WRAT at that time was upper first- grade level. He was placed in the EH program in January, 1966. Subject E-4— was a boy who had been deserted by his mother when he was three years old. The mother’s wherea bouts are still unknown. The psychological report states that he suffered early severe emotional deprivation. . . . [he has] observed serious perceptual difficulties, both visually and auditorily. . . . He was restless and hyperactive at times in the classroom, and refused to attempt any more than a minimal amount of schoolwork. He was eight and one-half at the time of the study, with an IQ of 104 on the WISC, administered to him in June, 1965. The WRAT given at the same time, rated him at a low first- grade academically. He was placed in the EH program in September, 196 5 - Subject E-5— was a boy eight years and nine months of age at the time of the study. He had a long history of physical illness and was under medical treatment for an anemic condition at the time of the study. He was also re ceiving a tranquilizlng drug to help control his hyperactive behavior. His psychological report states that . . . In view of [S's] inability to adjust to the expectancies of a regular classroom . . . he would be 70 better placed in the educationally handicapped class. His low frustration, hyperactivity, and uninhibited outbursts are indicative of a lack of internalized controls. His mother reported complete inability to control his activity at home. He achieved an IQ of 119 on the S-BIT in January, 1964, and was academically at a high first-grade level on the WRAT in November, 1964. He was placed in the EH program in January, 1965* At the start of this investi gation the school had decided to exclude this boy from the EH class because his behavior was uncontrollable at times. He was retained in school at the request of the experiment er for inclusion in this study. Subject E-6— was eight years and seven months old at the time of the study. He was on medication for possible cerebral dysfunction. His psychological evaluation stated that he was an instrument baby, anoxic at birth, and the total labor was eight hours. Mother reports: "He had dents in his forehead." . . . [in testing situation] tends to lack the necessary inhibitions that are basic to learn ing. He tends to act first and think later. . . . ap pears to need inner controls to function more adequate ly. . . . He is impulsive, highly motor driven, and easily distracted by extraneous stimuli. At the time of this study, the boy appears normal with no evidence of instrument "dents" reported at birth. His regular classroom teacher reported him to be "extremely hyperactive, difficult to control, and has a five minute attention span." His IQ on the S-BIT in December, 1963 was 71 90. A WRAT revealed that in all academic school areas he was at an early first grade level in December, 1963. He was admitted to the EH program in September, 1964. Subject E-7— was a ten year and seven month old boy with a long history of medical problems. Information from the School of Medicine at the University of Maryland in cluded the following: Date: 3-15-62 Age: 6-1/2 Clinical Data: At ages 3 and 5 had "concussion" which resulted in pt's vomiting. He has had two episodes of "blackout spells," seems to faint; fell off seat at school once. No tonic or clonic movements. Hx. of chronic anemia now and thrombocytopenia age 1-1/2. Clinical Impression: Adjustment reaction of childhood, behavior disorder, R/0 Brain damage secondary to brain trauma. EEG Interpretation: The resting record is borderline abnormal with a background which is somewhat slow for the age but which is essentially normal. He has received extensive psychiatric therapy for his temp er tantrums and delusions. His psychological evaluation states that he has "a very short attention span. . . . Easily distracted by outside stimuli. . . . Very ner vous . . . makes letters and numbers from the bottom, writes backward. . . . Extremely hyperactive. . . . Often in trouble." His present EH teacher reports that "He fre quently climbed upon the school roof and had to be forcibly brought down; he ran away from the school grounds several days each week and would not stay in his seat." All of these behaviors were observed several times during the week 72 preceding the start of this study. He obtained an IQ of 103 on a WISC given him in October 1965, and was academic ally on a high first-grade level on a WRAT given at the same time. He was placed in the EH program in January, 1965. This boy was scheduled for exclusion from school be cause he was uncontrollable in the classroom. He was re tained in the classroom for the purpose of this study, after which he was to be excluded. Subject E-8— was a boy, eight and one-half years old who was receiving medication daily for hyperactivity at the time of this study. His psychological report contains the following evaluation: . . . Children try to play and get along with him, but he always does things to disrupt their games and play. He does not conform to any rules or group stan dards . . . A high energy output was also present from an early age. . . . [his mother noted] that he had difficulty with immediate memory and recall, this problem has persisted. . . . [psychologist's impres sion] that he was aggressive, manipulative, and dis played a gradually increasing amount of hyperactivity which was difficult for him to handle. He was tested in January, 1965; he achieved an IQ score of 90 on the WISC and was rated at a first-grade level on the WRAT. He was placed in the EH program in January, 1965. Subject E-9— was a boy eight years and seven months of age. He had had his right ear drum removed by surgery and had a suspected neurological dysfunction, according to his medical report. He had trouble with all academic work and his teacher reported that he would often sit and rock for lengthy periods of up to an hour in his chair. His psychological report states that his "parents are desperate to get him to behave. . . . his behavior is typically hyper kinetic." He was given a WISC In October, 1964; his IQ was 88. A WRAT given at the same time rated him at a high first-grade level academically. He was placed in the EH program in January, 1965. He was withdrawn from school during the second week of the experimental program when his family moved from the district. Summary of the Experimental Class Children. The ex perimental class consisted of nine boys at the start of the experiment. One boy was dropped from the experiment when his family moved from the district: his data are not In cluded in class or inter-class computations. The age range was from 8.8 to 10.6 years; the mean age of the class was 9.5. The IQ range was from 88 to 119; the class mean was 98. One boy was on a low second-grade level academically, all others were at a first-grade level. Six of the eight boys were, or had been, under medication for hyperactive behavior. Three of the eight had been medically diagnosed as having some form of cerebral dysfunction or impairment. All were considered to be hyperactive and all were Cau casian. The range of time since placement in the EH pro gram was from two months to 30 months; the class mean was 13.2 months. A decision had been made, before the 74 experimenter approached the school district for permission to do this study, to exclude three of the boys from the program because of their uncontrollable behavior. They were specifically retained in school for purposes of this study. These data are presented in tabular form in Table 1. Subjects in the Contrast Class Subject C—1— was an eight and one-half year old boy. His psychological report states that [S] is a lonely, hostile boy who is expressing his anger over what he sees as unfair authority, by acting out. . . . Test results suggest an insecure, hostile boy who lacks self-control. His former teacher reported that he had frequent outbursts of temper. His IQ on the S-BIT in June, 1964 was 134. He was rated academically at a high first-grade level on the WRAT given at the same time. In September, 1964, he was placed in the EH program. Subject C-2— was an eight and one-half year old boy who had been referred to the psychologist because of fight ing on the schoolgrounds and inability to profit from re gular classroom instruction. The psychological report states that his former teachers reported he daydreams, rarely ever finishes his work, has sloppy habits. . . . [The psychologist reports that] he was referred in the previous year as the seriousness of his misconduct generally mounted. His mother reported frequent temper tantrums at home. His IQ in July, 1965 was 100 on the S-BIT; his achievement 75 level was mid-first grade level on the WRAT when he was admitted to the EH program in September, 1965. Subject C-3— was a nine year old boy who was describ ed on his psychological evaluation as an impulsive, hyperactive, aggressive youngster, whose behavior is chronically inconstant and unpredictable. He almost never sticks with anything over an ordinary period of time. He seems always to run and never walk. Unable to sit still for a minute. He acts before thinking, seldom considering the consequences of his behavior. He achieved an IQ of 132 on the WISC in April, 1965. His academic rating on the WRAT was on the second grade level at the same time. He had been placed on medication (amphe tamine and a tranquilizer which was not identified) before his placement in the EH program in February of 1965. Subject C-M— was a seven year four month old boy. He was placed in the EH program in February, 1966 because he was making no academic progress in school. His psycho logical evaluation reports that he had been on medication [Desoxyn] for hyperactivity for over a year . . . has continuing nocturnal enuresis, nail biting, and stuttering. . . . The teacher reports [S's] behavior as hyperactive, restless, erratic, per- severative, and cyclical. . . . His relationships with peers is immature and awkward, causing many fights and arguments. Scores on the WRAT rated him on a first-grade level academ ically in January, 1966. The S-BIT, also given at that time, indicated an IQ of 95. 76 Subject C-5— was an eight and one-half year old girl. Her psychological report describes her as unaware, slow, confused, in another world at times . . . will stand in one place for minutes at a time . . . extremely immature. . . . [Medical Diagnosis of] neuro logical impairment. . . . Everything must be slowed to almost a stop for her. She was not considered hyperactive or a classroom problem. She achieved an IQ of 81 in June 1965 on the WISC. She was not given an achievement test but her teacher estimated her academic level to be a very low first-grade or Kinder garten level. She was placed in the EH program in Septem ber of 196*1. Subject C-6— a girl, ten years and ten months old who was first diagnosed as mentally retarded and placed in the Educable Mentally Retarded Program. Her psychological case study summarized her as a girl who has severe learning difficulties which will handicap her in any regular classroom. There is evidence of visual perception limitations and immature fine motor coordination. It is also important to note that [S] has an extensive record of medical disorders for a child her age. She has at this time Diabetes Insipidus for which she is getting treatment. She achieved an IQ of 88 on the WISC in June, 1965. She was academically rated at a high-first-grade level on the basis of her scores on the WRAT given in June, 1965. She was placed in the EH program in January, 1966. Subject C-7— was a boy, ten years and four months of age. His former teacher reported that . . . This child has little interest in any learning situation. He has a very short attention span, doesn’t care to learn his colors, draw, or paint. . . . Does nothing on his own effort, talks out anytime feels like it. Gives" up very easily. Eager to learn at times. Refuses to do things at times; no self-control. Easily distracted. In his psychological interview he ’ ’told bizarre-sounding stories of bloody accidents and hospitalization for broken bones, of a non-existent brother. . . . was nervous, fre quently inattentive." His IQ on the S-BIT in March, 1964 was 98; a WRAT given at the same time placed him at a first-grade level academically. He was admitted to the EH program in March, 1965. Subject C-8— was a little girl six and one-half years old. Information on her Case Study Summary states that she was an adopted child who had, one hour after birth, been brought into a hospital more dead than alive; with the cord wrapped around her neck. She was placed in an oxygen tent and remained there for a long period of time. . . . She falls asleep in class and generally romps the house all night. Several different kinds of medication have been tried to alter her abnormal sleep pattern. None have worked. . . . she is very immature. Her first grade teacher had described her as a "different child . . . constantly demanding attention . . . unable to work by herself." Her S-BIT score in February, 1966 was 93. She was a non-achiever, according to her teacher, and was functioning on a pre-kindergarten level when placed in the EH program in February, 1966. 78 Subject C-9— was a seven year seven month old girl. She had been repeating the first grade at the time of her placement in the EH program in the middle of March, 1966. She had a thorough medical examination and no indications of cerebral dysfunction had been found. Her first grade teacher reported that she was unable to copy her name, was reading in the lowest pre-primer, had poor coordination and would often "freeze" when spoken to. She could not take an intelligence test because of her fear of the testing situ ation. Her psychological evaluation stated that "her self- concept is at such a low ebb that she will not extend her self in any way at school." Her IQ estimation was based upon a Peabody Picture Vocabulary Test which was administer ed to her in another school district in December, 1964; at which time she earned an IQ of 102. Since she was placed in the EH program in the middle of March, two weeks after the start of the experimental program, she will not be included in any of the statistical treatment of the data. Summary of the Contrast Class Children. The con trast class consisted of eight children at the start of the study. Five were boys, and three were girls. One girl entered the class during the study but her data are not in cluded in class and inter-ciass analyses. The age range was 7.3 to 10.10; the class mean was 8.7 years. The IQ range was from 84 to 134; the class mean was 103. One boy 79 was achieving academically on a second-grade level, five children were on a first-grade level and three were on a kindergarten level by teacher estimate. Only one child had a definite medical finding of neurological disorder and five were under some type of medication for hyperactive be havior or had been under medication. The range of time since placement in the EH program was from one month to 20 months; the class mean was 9*7 months. All children were Caucasian. This information is presented in tabular form in Table 2. Description of Rating Instruments Used The Activity Check List The Activity Check List was selected for this study because it provided a method by which each child could be observed and rated on seven different types of bodily and verbal activity. The types of behavior observed in the check list are described below: Movements directed toward the body: e.g., wringing of hands, rubbing eyes, swinging arms, leaning forward in the chair, scratching or stretching. Movements in chair: e.g., shuffling of chair, slid ing back and forth in chair, twisting of body in chair, or leaning clear out of chair so that buttocks no longer rested on it. TABLE 1 SUMMARY OF EXPERIMENTAL CLASS DATA Student E-l E-2 E-3 E-4 E-5 E-6 E-7 E-8 E-9 Av. Sex M M M M M M M M M Chronological Age— years 9-6 CO 1 CO 8-8 10-6 9-9 9-5 10-5 9-7 9-5 9-5 Chronological Age— ^months 114 104 104 126 116 113 125 115 113 114 I. Q. Score 91 97 99 104 119 90 104 90 88 98 Wide Range Achiev. Test Reading Grade 2.6 1.4 1.8 1.2 1.8 1.2 1.9 1.1 1.9 Spelling Grade 2.2 1.0 1.7 1.0 1.9 1.3 - - 1.3 1.7 Arithmetic Grade 2.5 1.5 1.6 1.8 2.2 1.5 2.7 2.0 2.2 Under Medication No Yes Yes No Yes Yes Yes Yes Yes Medical Diagnosis of Neuro logical Dysfunction Yes No No No No Yes Yes No Yes Hyperactive Behavior Yes Yes Yes Yes Yes Yes Yes Yes Yes Months since Placement in EH Program 14 11 2 6 14 30 14 14 14 13.2 CO Race C C C C C C C C C o TABLE 2 SUMMARY OF CONTRAST CLASS DATA Student C-l C-2 C-3 C-4 C-5 C-6 C-l C-8 C-9 Av. Sex M M M M F F M F F Chronological Age— years 8-6 8-6 8-11 7-3 8-7 10-10 10-5 7-6 7-8 8-7 Chronological Age— months 102 102 107 87 103 130 125 90 92 104 I. Q. Score 134 100 132 95 84 88 98 93 102 10 3 Wide Range Achiev. Test Reading Grade 1.8 1.4 o * c v j 1.4 1.4 1.1 1.1 _, — Spelling Grade 1.8 1.5 1.7 Arithmetic Grade 2.5 2.2 3.1 1.8 - - 2.5 2.3 - - - - Under Medication No No Yes Yes No Yes No Yes No Medical Diagnosis of Neuro logical Dysfunction No No No No Yes No No No No Hyperactive Behavior Yes No Yes Yes No No No Yes No Months since Placement in EH Program 20 19 13 2 17 3 12 1 0 9.7 Race C C 1 - - - C C C C C C C CO M 82 Distraction: e.g., looking over toward a noise, to ward someone who has just entered the room (unless the teacher also looks up), out of the window, or off into space. Gross movements of legs and feet: e.g., pumping of leg, wiggling of feet, or crossing of legs, or other shift of their position. Fiddling: Arm and hand movements directed toward objects: activities of the hand(s) that interfered with schoolwork or assigned activities, e.g., stroking the desk, fingering the box of color crayons. Communicative or quasl-communlcative activity inter fering with schoolwork: e.g., talking to self, pointing, laughing, attempting to attract someone's attention, talking to someone without permission. Walking or standing that was not encouraged or sub sequently approved of by the teacher. (Patterson, Jones, Whittier & Wright, 1965) Each of the judges rating the children in a class would independently observe a single child for ten seconds, then check on the Activity Check List (See Appendix D) all of the above activities, if any, he had observed the child V commit. Activity that had teacher approval of was subse quently approved by the teacher was not included. The observer would then go on to the next child and repeat the same procedure. Each child was observed individually over 83 100 times during a three day period before the start of the experimental period and for approximately the same amount of observations following the experiment. The total number of non-teacher-approved activities observed in each child were then divided by the total number of times that that child had been observed. The number thus obtained was the Activity Level Average (ALA). The Activity Check List was developed by G. R. Pat terson, et al. (1965) for use as a method of establishing activity level base rates of hyperactive children who parti cipated in operant conditioning studies at the University of Oregon. To establish the reliability of the instrument, four observers observed two hyperactive children over a one week period. They rated the children on the Check List. The total frequency of non-attending behavior patterns were summed for eight ten-minute intervals for each pair who had been observing the experimental subject. When using total frequency of occurrence on non-attending behavior the cor relation between one pair of observers was 0.90 and the coorelation between the second pair was 0.91. Anderson (1964) found a reliability of 92 per cent, 82 per cent, and 72 per cent with pairs of three observers when observing the same child. Inter-rater agreement for this study was established by requiring the three judges to observe the children in a classroom simultaneously. The judges would all observe one child at time, rate him, then move on to another child. Thus, each child in the class was observed 20 separate times by each of the judges. There was a correlation of 0.84 between judges 1 and 2; 0.82 between 1 and 3 and a correlation of 0.80 between judges 2 and 3. Correlation was established by means of the Spearman Rank Correlation method (Siegel, 1956, pp. 202-205). A second method was to rate the total amount of activity observed in the 160 intervals (eight students observed 20 times each). Judge 1 observed 192 responses; Judge 2 observed l8l and Judge 3 168. Thus, there was good agreement between the judges on the number occurring in a given session. The Burks Behavior Rating Scale The Burks Behavior Rating Scale was used in this study because it was designed as a scale to rate hyperactive and maladaptive behavior. Also, it was one of the basic instruments used for the referral, screening and placement of children in the special classes of the Redondo Beach School District. The classroom teachers involved in the study were required to rate each child in their class on the instrument yearly, and were thoroughly familiar with administering it. The scale consisted of three sections: 1. Vegatatlve-Autonomlc; e.g., hyperactiveness, restlessness, erratic, explosive behaviors. 2. Perceptual-Discriminative; e.g., poor in reading, confusion in spelling and writing. 3. Social-Emotional; e.g., demanding much attention, shallow feeling, cries, etc. There were 28 statements describing behavior rela tive to the above headings. Teachers were to rate each child on each statement according to the following five point scale: 1. You have not noticed this behavior at all. 2. You have noticed the behavior to a slight de gree . 3. You have noticed the behavior to a considerable degree. 4. You have noticed the behavior to a large degree. 5. You have noticed the behavior to a very large degree. Teachers rated the children in their classes both before and following the experimental period. Approximately five minutes was required to rate each child. The scale was developed by Harold P. Burks for use in his doctoral dissertation at the University of Southern California (Burks, 1955). It has since been adopted for use by the Los Angeles County Schools Division of Research and Guidance and is presently used in many school districts for the selection of EH children. 86 The validity of the instrument was determined by having classroom teachers rate the entire student body of an elementary school (grades 1 through 6) on the scale. A total of 528 children (263 boys and 265 girls) comprised the reliability group. An analysis of variance was made on the means of the class groups (given in Table IV; Burks, 1955). The results indicated that means among the teachers differed significantly. Burks discussed these implications (p. 85, 1955) and the several possibilities they might re present. Two months later each teacher was asked to rate again ten pupils in her respective classrooms. When pupils were matched with themselves, there was no signifi cant difference between the teachers' ratings on the first and second sessions. Individual items were tested to see if ratings on them remained constant over the two month period. The application of the sign test was used to obtain an Indication of stability. When 124 ratings on each item were compared to 124 ratings two months later only three items showed a significant difference. No sig nificant difference was found between boys and girls. Validity was checked by comparing ratings on the scale against electroencephalographic tracings of normal and abnormal children. The scale significantly selected the abnormals. Behavior Scale ratings were given normals and neurologically impaired with extensive psychological work-ups and the scale was found to differentiate 87 significantly between them. The test and manual are published by the Los Angeles County Schools Office, Los Angeles, California. The Teacher Observation Form The Teacher Observation Form was developed for use in this study. The three judges were asked to observe and note on the Form all instances of teacher reaction or in teraction with any student(s). Teacher activities that strengthened adaptive or socially acceptable behavior were listed in one column, while activities that strengthened maladaptive behavior or reduced adaptive behavior were listed in the second column. Examples of socially accep table behavior are: responding to a child who was sitting quietly with his hand up; complementing a child for a job well done; praising a child in front of his peers, etc. while examples of the latter category are: replying to a shouted question; talking to a boy who left the room without permission; and not responding to a child who was setting quietly with his hand up. These ratings were made over a three day period both before and after the experimental period. Inter-rater agreement was established by having the three judges simultaneously observe a teacher of a class room for educationally handicapped children as she conducted her class. Two fifteen-minute observation periods were 88 used. The ratings of the Judges were compared. The per centage of agreement between Judge one and Judge two was 92 per cent. Between judge one and three, 86 per cent and between judge two and three it was 88 per cent. The Sound Level Meter The General Radio Type 759 Sound Level Meter was used to record the level of background noise in both the experimental and contrast classrooms. The Sound Level Meter was placed in one corner of the classroom (see Pig. 1 and 2). It was situated in the same place in the post- experimental period as in the pre-experimental period. Sound level readings were taken exactly at five-minute in tervals. Readings were taken over a three-day period be fore and after the experimental period. The particular instrument used in this study was secured from the Physics Department of the University of Southern California. It had been calibrated immediately prior to use in this study. The meter was manufactured by the General Radio Corporation in Boston, Massachusetts. Other Sources of Information The results obtained with the above instruments were treated statistically. Parent Information Information about the behavior of the children in 89 the experimental program in their home and community en vironment was obtained by the following methods: The Parent Questionnaire. This form was prepared for use in this study. It was given one week after the termination of the study. Its purpose was to obtain from the parents of children in the experimental class an evalu ation of the child’s behavior at home and in the community. The parent was requested to discuss in particular any be havioral change noticed from the onset of the experimental period to the time when the questionnaire was completed. The form was specifically designed to evoke observations of any new unfavorable symptoms might have appeared since the start of the study. These comments were summarized and discussed below. Parent-Group Meetings. Parents with children in the EH program at Franklin School met one evening a month. The writer interviewed each parent and attempted to evaluate any change In behavior noted outside the school. Parents also discussed the effects of the program on their respec tive children at the meetings. Impressions and Opinions of Administrative and Guidance Personnel of the Redondo Beach School District' Administrators and guidance personnel who observed the class during the experimental program were asked to 90 convey their impressions by means of a letter to the writer’s university advisor. Copies of all letters re ceived are included in Appendix C. Pre-Experimental Period Procedure Judges The three judges consisted of two doctoral students in Exceptional Children and the writer. All were from the University of Southern California. One judge was a former school psychologist and all had had extensive work in evaluating emotionally disturbed and mentally retarded children and adults. The purpose of the experiments and the rating in struments were explained and demonstrated in detail to the judges before the beginning of the observation period. Classroom Observation The two classes were observed by the three judges over a three-day period. Judges were located at tables at the side, and to the rear of each classroom where they were unobtrusive but could clearly see all children in the room. Judges alternated between classrooms, observing in each classroom for approximately the same amount of time and at approximately identical times on alternate days. The judges observed individual behavior of each child and in dicated their observations on the Activity Check List in the manner described in the Instruments section. Each 91 judge also completed two or more random 15 minute obser vations of the classroom teacher and recorded these on the Teacher Observation Form. One judge recorded the classroom noise level at five minute intervals from the Sound Level Meter which was placed on the floor by the judge's table. The judges rated children and teachers independently, i.e., there was no attempt for two judges to observe the same child or teacher at the same time. The judges attempt ed to spend a similar time in each of the two classrooms over a three day observation period in the two observation periods. Teacher Ratings. Teachers were asked to rate each child on the Burks' Behavior Rating Scale before the start of the experimental period. Instructions for the Experimental Class Teacher Following the gathering of the pre-experimental period data the writer met with the experimental class teacher and discussed the purpose of the study in detail. Principles of operant conditioning were explained and she was advised on their application during the experimental period. She was instructed to reward all adaptive behavior, such as working quietly, raising of hands, sitting at the desk, etc. When adaptive behavior such as this occurred, she was to compliment the child, notice him, smile at him, 92 or otherwise verbally acknowledge her approval. She was further advised to ignore all maladaptive behavior during the experimental period. She was not to answer a shouted question, talk to a boy out of his seat without permission, or pay attention to a boy who was not working. Instructions for the Contrast Class Teacher The contrast class teacher was not given any in structions. She was advised that the writer would be available to her for from one-half to one hour daily and that he would aid her in any way that she desired. Dally Procedure in the Experimental Class Physical Arrangement of the Classroom The observation table was located in the front and to the left side of the room where it could be clearly and easily seen by all subjects. A CRA-LAB Model 172 Electric Interval Timer with an 8" dial was placed on the table; a half-quart jar with several stacks of poker chips beside the jar and a few samples of candy. The observer sat be hind the desk where he could watch all children. Desks of the children were evenly placed in three rows, three desks deep, facing the front of the room. The teacher’s desk was located at the front and to the extreme right hand side of the room. The arrangement of the experimental and 93 contrast rooms is presented graphically in Figures 1 and 2. Selection of the Experimental Time Period Educationally Handicapped children attended school from 8:00 AM to 12:00 noon. The experimental time period chosen was from 10:30 AM to 11:30 AM. This period was selected because no systematic classroom work or individual instruction had been possible during that time interval previous to the experimental period. It was the opinion of the teacher and the school principal that the children were most hyperactive and disruptive during this time and were often unmanageable. First Day Instructions to Children At the start of the first daily experimental session the experimenter explained that he would be working with the class for the next month and that he was doing a special study for his university. He explained that he would be playing a game each day with the class and that they could win candy for being quiet and sitting in their seats during the time that they were playing the game. They were shown the electric timer and it was demonstrated to them several times until everyone understood how it worked. Next the poker chips and half-quart bottle were demonstrated and they were told that each time the timer got all the way to zero without being stopped they would receive one chip. It was FIGURE 1 ARRANGEMENT OF THE EXPERIMENTAL ROOM 94 Front f E-3 f E-2 E-l a \ E-9 f E-S E-S E-5 E-7 Legend a. Windows b. Clothes or storage closets c. Window-level cabinets d. Teacher’s desk e. Observation desk f. Student tables g. Sound level meter h. Experimental period observation table 95 FIGURE 2 ARRANGEMENT OF THE CONTRAST ROOM C-2 C-3 C-6 8-5 c-i c-< Legend a. Windows e. Observation table b. Clothes and storage fi Student tables closets g* Sound level meter ci Window-level cabinets h. Partitions d. Teacher1s desk i# Reading table 96 carefully explained that any unapproved talking, moving about, or class disrupting behavior would cause the timer to be stopped and re-set. The chip would be placed into the bottle and would not be removed until the end of the period. If, at the end of the period there were more than 20 chips in the bottle, the chips would be evenly divided among all children present and could be immediately traded for candy. Several different kinds of candy were available and could be obtained by exchanging one chip for one piece of candy. If fewer than 20 chips were won, no candy was won and the class would receive nothing. Extra or left over chips were kept until Friday of each week and dis tributed separately from the regular Friday tokens. It was carefully emphasized to the children that their teacher had nothing whatever to do with their winning or not winning any candy. She would not make them play, tell them to sit down, be quiet, or to behave in any way during the time the game was being played. Finally, they were told that if at any time they should grow tired of the game and did not want to play, the game would be discontinued immediately. At the conclusion of this introduction they were then asked to vote on having the game in their room. All voted in favor of playing. Following this explanation, the first session was begun. 97 Procedure during the Experimental Hour The experimenter would come Into the classroom short ly before the end of the nutrition period and set up the observer's table in the manner previously described. When this was done he would announce that whenever the class was ready the game would begin. He made no further announce ments and started the clock contingent on appropriate be havior of the class. During the experimental period the observer watched the subjects, placed chips into the bottle as they were won or reset the timer after any maladaptive behavior was ob served. At no time did the experimenter teach the class or serve as a teacher during the experimental hour. He gener ally remained either behind the observer's table or at the rear of the room when a student was operating the timer. On the eighth day of the experiment the job of class observation and operation of the timer and poker chips was delegated to one of the children. Thereafter the position of class observer was highly sought after with boys signing up as far as two weeks in advance. Student-observers usu ally were more severe in their evaluation of infractions than the experimenter would have been and took their re sponsibility most seriously. No reward other than an equal share of the class tokens was given. At the conclusion of each daily period, the reward 98 was immediately distributed to each child. Beginning with the third week the reward was changed from a piece of candy for each chip to one penny in money for each chip. A boy could then buy candy on display in the room, take his money home with him, or save it in a "bank" operated by the experimenter. By the termination of the experimental period all boys had bank accounts and were saving some or all of their winnings. Time Intervals. A twenty-second time interval was used for the first seven days, and was increased to 45 seconds for the 8th, 9th, and 10th days. The time interval was increased to one minute for the 11th, 12th and 13th days. On the 14th and 15th days the interval was extended to five minutes. Thereafter and for the remainder of the experiment a time interval of 10-minutes was used. After the one-minute time interval was reached, the class was always given one chip for each minute. In the 10-minute time interval, for example, the clock had to run uninter rupted for the full 10-minutes for 10 chips to be won. Any infraction occurring before the completion of the interval meant that all chips accumulated in that interval were lost and the timer restarted at "10" again. Once the interval was completed and the timer had reached "0," the chips were immediately placed in the half-quart jar and could not be lost because of subsequent maladaptive behavior in the next 99 time period. Teacher Participation in the Experimental Class. During the experimental period the teacher utilized operant conditioning techniques when working with the class. She actively sought opportunities to praise any student who was doing his assignment, sitting quietly, or exhibiting any other approved behavior. She was quick to respond to any subject who held up his hand while sitting quietly and would never fail to compliment him on his good behavior. She attempted to ignore completely all maladaptive behavior, refusing to answer a verbal call, talk with any subject who was out of his seat without permission. No student could obtain attention or talk with the teacher during the ex perimental period unless he was acting in an approved man ner. No teaching directions were given to the teacher by the experimenter. The teacher was encouraged to attempt any type of classwork that she desired during the experimen tal period. Dally Follow-Up. Each afternoon the experimenter and the teacher discussed the progress of the class. Any instances of teacher-reinforcement of maladaptive behavior that had been observed by the experimenter were discussed. Suggestions of various methods to eliminate or decrease the maladaptive behavior were given. 100 Operant Techniques Used. Application of the free operant technique to pupil behavior was done by two dis tinct procedures: (1) social and verbal reinforcement and reward by the teacher for all adaptive and socially accept able behavior while at the same time completely ignoring all maladaptive behavior; (2) use of the stop-clock and poker chips to reward, or pay for, appropriate adaptive motor and verbal behavior. In the first category the teacher actively sought instances of adaptive operant behavior to reward instantly. Maladaptive behavior, no longer securing the attention of the teacher or not resulting in the realization of any of the goals for which it had been emitted, would tend to decrease. In the second case, the object of paying the entire class only if the behavior of all were acceptable was to remove peer reinforcement for previously approved disruptive and maladaptive behavior. Since the class earned one penny for each chip and a single chip represented one minute of group adaptive be havior, the amount of money that could be won in the hour was rather limited. As the percentage of good behavior in creased, the boys requested a longer period so that they could win more money. Consequently, the experimental per iod was lengthened to run from 10:15 to 11:50 In the third week, and from 10:00 to 11:50 in the last seven days. 101 Dally Procedure In the Contrast Class The experimenter worked with the contrast class children dally from 9:00 to approximately 10:00. Half of the period was spent In the classroom where he served as a teacher-alde giving individual reading instruction to children selected daily by the teacher. The remainder of the time was spent on the playground with the class. The experimenter worked with these children exactly as he worked with the experimental class children, the only dif ferences being that no group operant techniques were used and no critique of teacher reinforcement of student be havior was offereck- In situations where his advice or opinion was sought in respect to any problem or student, he responded with what he considered the best means or method of dealing with, or resolving, the problem. Post-Experimental Period Observations The post-experimental period observations were con ducted by the same three judges who made the pre-experimen tal period observations. The same procedure was followed in rating the classes and pupils as had been used in the pre- experimental period. Method of Data Analysis With respect to the first hypothesis, stated in the null form,— that the use of operant techniques will not 102 reduce the occurrence of maladaptive behavior in a class room for educationally handicapped children— the following statistical computations were used in accordance with the procedures described by Garrett (1962, p. 227) for determin ing the significance of the difference between means in small correlated groups: 1. The Activity Level Average (ALA) for each child was computed by dividing the total number of activities not approved by the teacher observed by the total number of 10 second periods during which the child had been under obser vation. 2. Experimental group subjects' ALA scores were divided into Pre- and Post-Experimental categories. 3. The small group formula to determine the differ ence between correlated means was used. 4. Difference was determined by using a two tailed test of significance with the significance level indicated for seven degrees of freedom in Table D, as shown in Gar rett (1962, p. 449). 5. The same procedure was then followed with the data obtained from the Contrast Classroom Subjects. Analysis of data in terms of the second hypothesis— that the use of operant techniques will reduce the class room noise level in a class of educationally handicapped children— was done in accordance with the procedures 103 described by Guilford (1965* P* 182) for testing the dif ference between uncorrelated means. 1. Sound level readings for the Experimental class were cast into two groups: one listing all pre-experimental readings and the other listing all post-experimental period readings. 2. The t-test of a difference between means was used. 3. The same procedure was followed with data ob tained from the Contrast Classroom. Analysis of the data in terms of the third hypo thesis— that a classroom teacher applying operant techniques will not reinforce student maladaptive behavior— required the following statistical computation in accordance with the procedures described by Garrett (1962, pp. 264-266) for the Chi Square statistic for independent samples: 1. For the experimental class, a 2 x 2 contingency table was used, into which observed instances of teacher reinforcement of adaptive and maladaptive behavior were cast. 2. The Chi Square formula for two independent sam- r■ pies, corrected for cintinuity, was used. 3. Significance was determined for a two-tailed test. The significance level used was the one indicated for one degree of freedom by Tabel E, the Table adapted from Fisher. (Garrett, 1962, p. 450) 10 4 4. Contrast classroom data were then observed and analyzed In the same manner. In terms of the fourth hypothesis— that maladaptive and hyperactive behavior would be reduced in the total school environment— the method of statistical anaylsis used was that described by Garrett (1962, p. 227) for de termining the significance of the difference between means in small correlated groups: 1. The experimental class teacher evaluated all subjects in her class on the Burks* Behavior Rating Scale before the start of the experimental period. Following the experimental period the subjects were again evaluated on the same instrument. 2. Subjects were then cast into one group with their scores dichotomized intopre- and post-experimental list s. 3. The small group formula to determine the sig nificance between correlated means was used. 4. Significance was determined by using a two tailed test with the significance level indicated for seven degrees of freedom in Table D, as shown in Garrett (1962, p. 449). 5. The same procedure was followed with the data obtained from the contrast classroom subjects. Significance Level The writer accepted a five per cent risk of making a 105 type I error. Consequently, this significance level was set as the maximum probability at which the null hypothesis could be rejected. CHAPTER IV EXAMINATION OP THE EXPERIMENTAL RESULTS This chapter presents the findings of this study and evaluates the research hypotheses relating to these findings. These hypotheses, described more fully at the end of chapter two, will be restated in their null form and more briefly. The Hypotheses Hypothesis I This hypothesis in null form predicted that the use of operant techniques would not reduce the occurrence of maladaptive behavior of educationally handicapped children. The hypothesis was tested by matching Activity Level Aver age scores received in the pre-experimental period with those received in the post-experimental observation period and analyzing them by means of the t-test. The raw data are contained in Appendix A. Tables 3, 5, and 6 present the analysis of the data in tabular form. When computed separately, the difference between the pre- and post-treat- ment rating of each judge was significant beyond the .001 level of confidence. When combined into a single compari son of all pre- and post-treatment Activity Level Averages, 106 TABLE 3 COMBINED RATINGS OF THREE JUDGES OF STUDENT’S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Experimental Class Subject No: Pre-Experimental Period Post-Experimental Period Difference ALAj— ALA2 No. Observed Activities:a No. Obser vations ala- l No. Observed Activities3 No. Obser vations15 ala2 E-l 207 115 1.80 115 180 .64 1.16 E-2 220 133 1.65 80 180 .44 1.21 E-3 213 141 1.51 89 180 .49 1.02 E-4 278 140 1.99 118 180 .66 1.33 E-5 292 140 2.08 85 180 .47 1.61 E-6 216 138 1.56 78 180 .43 1.13 E-7 286 134 2.13 83 180 .46 1.67 E-8 275 132 2.08 118 178 .67 l. 41 (t = 17-5; SDn = .216; SEM = .075) Mean Difference---- 1.318*** D *** Significant at .001 level of confidence. a Number of maladaptive behaviors observed during 10-second observation periods, b Number of 10-second time intervals during which the child was observed. TABLE 4 INDIVIDUAL RATINGS OF STUDENT’S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Experimental Class Judge No. 1 Subject No. Pre-Experimental Period Post Experimental Period Difference ALAj— ALA2 No. Observed Activitiesa No. Obser vations^ ala- l No. Observed Activities3 No. Obser vations*3 ALA1 E-l 54 32 1.37 48 60 .80 .57 E-2 110 53 2.08 24 60 .40 1.68 E-3 70 53 1.32 37 60 .62 .70 E-4 86 53 1.62 47 60 • 72 .84 E-5 10 7 53 2.02 27 60 .45 1.57 E-6 96 54 1.76 32 60 .53 1.23 E-7 135 54 2.50 23 60 • 38 2.12 E-8 119 55 2.16 55 60 .91 1.25 (t = 6.58; SDd = .53; SEMr, = •19) Mean Difi?erence --- 1.25*** *** — Significant at .001 level of confidence. a — Number of maladaptive behaviors observed during 10-second observation periods, b — Number of 10-second time intervals during which the child was observed. o CD TABLE 5 INDIVIDUAL RATINGS OF STUDENT'S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Experimental Class { s Judge No. 2 Subject No. Pre-Experimental Period Post-Experimental Period Difference ALAX— ALA2 No. Observed Activities3 No. Obser vations*3 ALAl No. Observed Activities3 No. Obser vations*3 ala2 E-l 67 47 1.43 29 70 .41 1.02 E-2 60 50 1.20 24 70 .34 .86 E-3 57 50 1.14 22 70 • 31 .83 E-4 55 50 1.10 23 70 .33 .77 E-5 66 50 1.32 23 70 • 33 .99 E-6 55 50 1.10 15 70 .21 .89 E-7 97 50 1.94 19 70 .27 1.67 E-8 101 50 2.02 21 70 .30 1.72 (t = 7.56; SDd = .38; SEMd = .13V Mean Difference— — 1.094*** *#« — Significant at .001 level of confidence. a — Number of maladaptive behaviors observed during 10-second observation periods, b — Number of 10-second time intervals during which the child was observed. TABLE 6 INDIVIDUAL RATINGS OF STUDENT’S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Experimental Class Judge No. 3 Subject No. Pre-Experimental Period Post-Experimental Period Difference ALAX— ALA2 No. Observed Activities3 No. Obser vations13 AL'A1 No. Observed Activities3 No. Obser vations13 ala2 E-l 86 36 2.39 38 50 • 76 1.63 E-2 50 30 1.67 32 50 .74 1.03 E-3 86 38 2.26 30 50 .60 1.66 E-4 137 37 3.70 48 50 .96 2.74 E-5 119 37 3.22 35 50 .70 2.52 E-6 65 34 1.91 31 50 .62 1.29 E-7 54 30 1. 80 42 48 .87 .93 E-8 55 27 2.04 41 50 .82 1.22 =r • l l -p » ^ D = *03> SE^^ .22) Mean Difference — 1.63*** «*« — Significant at .001 level of confidence. a — Number of maladaptive behaviors observed during 10-second observation periods, b — Number of 10-second time intervals during which the child was observed. TABLE 7 COMBINED JUDGES’ RATINGS OF STUDENT'S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Contrast Class Subject No. Pre-Experimental Per:Lod Post-Experimental Period Difference ALA^— ALA2 No. Observed Activities13 No. Obser vations0 ALA-l No. Observed Activities13 No. Obser vations0 ala2 C-l 134 96 1. 40 137 90 1.52 - .12 C-2 256 106 2. 42 205 90 2.28 .14 C-3 125 110 1.14 140 90 1.56 - .42 C-4 12 8 110 1.16 88 90 CO o> • .18 C-5 135 95 1.42 97 90 1.07 .35 C-6 166 115 1.44 116 90 1.29 .15 C-7 157 110 1.43 128 90 1.42 .01 C-8 185 110 1.68 199 90 2.21 - .53 (t = 0.171; SDp = .494; SEm^ = .175 Mean Difference — - ,03a a — There was no significant difference between means. b — Number of maladaptive behaviors observed during 10-second observation periods, c — Number of 10-second time intervals during which the child was observed. TABLE 8 INDIVIDUAL RATINGS OF STUDENT’S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Contrast Class Judge No. 1 Subject No. Pre-Experimental Period Post-Experimental Period Difference ALAi— ALA2 No. Observed Activities1 3 No. Obser vations c ALAj No. Observed Activities*3 No. Obser vations c ala2 C-l 48 36 1.33 71 40 1.78 - .45 C-2 142 56 2.54 81 40 2.03 .51 C-3 52 50 1.04 70 40 1.75 - .69 C-4 72 50 1.44 45 40 1.13 .31 C-5 41 35 1.17 41 40 1.03 .14 C-6 79 55 1.44 46 40 1.15 .29 C-7 88 50 1. 76 62 40 1.55 .21 C-8 78 50 1.56 113 40 2. 83 - 1.27 (t = .52; SDp = .726; SEp/jD = .257) Mean Difference — - . 13a a — There was no significant difference between means. b — Number of maladaptive behaviors observed during 10-second observation periods, c — Number of 10-second time intervals during which the child was observed. TABLE 9 INDIVIDUAL RATINGS OF STUDENT'S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Contrast Class Judge No. 2 Subject No. Pre-Experimental Period Post-Experimental Period Difference A L A —ALA2 No. Observed, Activities1 3 No. Obser vations 0 ALA^ No. Observed Activities*3 No. Obser vations0 ala2 C-l 62 40 1.55 40 30 1.33 .22 C-2 61 30 2.03 69 30 2.30 - .27 C-3 49 40 1.23 42 30 1.40 - .17 C-4 37 40 .93 23 30 .77 .16 C-5 63 40 1.67 38 30 1.27 .40 C-6 55 40 1. 38 38 30 1.27 - .09 C-7 39 40 .98 37 30 1.23 - .25 C-8 58 40 1. 45 51 30 1.70 - .25 (t = .12; SDd = .72; SEm d = .25) Mean Difference — - .03a a — There was no significant difference between means. b — Number of maladaptive behaviors observed during 10-second observation periods, c — Number of 10-second time intervals during which the child was observed. TABLE 10 INDIVIDUAL RATINGS OF STUDENT’S CLASSROOM ACTIVITY ON THE ACTIVITY CHECK LIST Contrast Class Judge No. 3 Subject No. Pre-Experimental Period Post-Experimental Period pifference ALAi— ALA2 No. Observed Activities'3 No. Obser vations0 Alax No. Observed Activities'3 No. Obser vations0 ala2 C-l 24 20 1.20 26 20 1.3 - .10 C-2 53 20 2.65 55 20 2.75 - .10 C-3 24 20 1.20 28 20 1.40 - .20 C-4 19 20 .95 26 20 1.30 - . 40 C-5 31 20 1.55 18 20 .90 .65 C-6 32 20 1.50 32 20 1.10 .50 C-7 30 20 1.50 29 20 1.45 . 10 C-8 49 20 2. 45 35 20 1.25 1.20 (t = 1.4; SDD = .435; SEm d = .153) Mean Di.fference — . 21a a — There was no significant difference between means. b — Number of maladaptive behaviors observed during 10-second observation periods, c — Number of 10-second time intervals during which the child was observed. a t of 17*5 was obtained, far in excess of the .001 level of confidence. A similar analysis was made of contrast group data (Tables 7, 8, 9 and 10). Results indicated no significant differences either in the ratings of single judges or the combined ratings. Thus the null hypothesis (HOj) was rejected at the .001 level of confidence. Hypothesis II This hypothesis in null form predicted that operant conditioning techniques would not reduce the background noise level of a class of educationally handicapped chil dren. This hypothesis was tested by comparing the pre- experimental period sound level readings with the post-ex perimental period sound level readings of the same class and analyzing them by means of the t^-test for uncorrelated means. Raw data obtained from each class are presented in Appendix B. Statistical treatment of the data is found in Table 11. Results for the experimental class showed that a mean difference of 17*1 decibels between pre- and post-ex perimental period was significant beyond the .001 level of confidence. Analysis of the readings taken in the control classroom show a mean Increase (not significantly different) in sound level of 2.2 decibels. Thus, the null hypothesis (HO2) is rejected at the 116 FIGURE 3 GRAPHIC ILLUSTRATION OF THE RELATIVE POSITION OF MEAN SOUND LEVEL METER READINGS OF THE EXPERIMENTAL AND CONTRAST ROOMS IN THE PRE- AND POST-EXPERIMENTAL PERIODS EC CC •H A B Legend A Pre-Experimental Period B Post-Experimental Period EC Experimental Class CC Contrast Class TABLE 11 STATISTICAL ANALYSIS OP SOUND LEVEL METER DATA OBTAINED IN PRE AND POST-EXPERIMENTAL PERIODS Pre-Experimental Post Experimental Mean t Period Period Difference value Experimental Class Number of Readings ^11 49 Standard Deviation 5.6 2.2 Mean of Readings 72.8 55.7 17.1 19.6*** Contrast Class Number of Readings 48 46 Standard Deviation 5.2 4.9 Mean of Readings 65.7 67.9 - 2.2 1.6a a — No significant difference between means. «*« — Significant at the .001 level of confidence. 118 .001 level of confidence. Hypothesis III This hypothesis in null form predicted that the use of operant techniques by a classroom teacher would not re sult in less teacher reinforcement of maladaptive student behavior. The hypothesis was tested by comparing the pre- experimental period ratings of teacher reinforcement of student behavior with the post-experimental period ratings. Results were analyzed by means of the Chi Square statistic. Data for the experimental class are presented in Table 12. An inspection of Table 12 shows that differences reported by two judges are significant at the .001 level of confi dence and that the difference reported by the third was significant at the .01 level. When combined into a single table a Chi Square of was obtained— far beyond the .001 level of confidence. Additional data are presented in Appendix C. Table 12 also presents data obtained from the con trol classroom. There were no significant differences, either individually or when combined into a single group. Thus the null hypothesis (HO3) is rejected at the .001 level of confidence. Hypothesis IV This hypothesis in null form stated that there would be no reduction in the occurrence of maladaptive TABLE 12 COMPARISON OF TEACHER REINFORCEMENT OF ADAPTIVE AND MALADAPTIVE SUBJECT BEHAVIOR IN PRE-AND POST-EXPERIMENTAL PERIOD Experimental Class Contrast Class Judge Judge Judge Total Judge Judge Judge Total 1 2 3 1 2 3 Pre-Experimental Period Total Reinforcement of Adaptive Behavior 24 23 14 6l 51 22 26 99 Total Reinforcement of Maladaptive Behavior 44 24 20 88 51 19 22 92 Post-Experimental Period Total Reinforcement of Adaptive Behavior 48 22 12 82 24 29 26 79 Total Reinforcement of Maladaptive Behavior 0 0 0 0 22 26 19 67 Chi Square 50*** 15*** 10** y Z 4*** o • o 0.0a 0.0a O.ia a — No significant difference ** — Significant at the .01 level of confidence *** — Significant at the .001 level of confidence 120 behavior in the over-all school environment as a result of the use of operant techniques. This hypothesis was tested by matching teacher ratings of individual students’ be havior on the Burks' Behavior Rating Scale obtained before the start of the experimental period with ratings obtained after the experimental period. For the experimental class the mean difference between ratings was 33.6, which is sig nificant beyond the .001 level of confidence (t-test for small matched samples shown in Table 13). When analyzed by means of the tj-test for small matched samples a t-value of 8.29 was obtained (see table 13). This was significant beyond the .001 level of confidence. The mean difference between pre- and post- experi mental period in the control class was 10, which is signi ficant between the .05 and .01 level (see table 13). This was significant at the .05 level but not at the .01 level. Thus the null hypothesis (HO^) is rejected at the .001 level of confidence. Supplementary Findings In addition to the major findings, the following supplementary findings were recorded. Table 3 shows that every student in the experimen tal class reduced his Activity Level Average by at least two-thirds of his pre-experimental score. Similar results were obtained from the teacher evaluations on the Burks 121 TABLE 13 TEACHER EVALUATIONS BURKS' BEHAVIOR RATING SCALE Student Pre-Experimental Post-Experimental Difference No: Period Score Period Score (Pre-Post) E-l Experimental 92 Class 73 19 E-2 10 3 73 30 E-3 93 53 40 E-4 76 57 18 E-5 95 65 30 E-6 107 65 42 E-7 116 65 51 E-8 111 73 38 Mean Score: 99.1 65.5 33.6#*# (SD difference = 11.4; SE mean difference = 4.04 ; df = 7 C-l Contrast Class 86 80 6 C-2 90 66 24 C-3 68 54 14 c-4 51 55 4 C-5 68 66 2 C-6 60 43 17 C-7 91 78 13 C-8 75 64 11 Mean Score: 73.6 63.2 10.25* (SD difference = 8.83; SE mean difference = 3.1; df = 7) * — Significant at the .05 level of confidence. ### — Significant at the .001 level of confidence. 122 Scale where there was a drop of at least 18 points for each child. No child In the experimental class showed an In crease In any maladaptive or hyperactive behavior in areas covered by the three rating instruments. One of the more striking findings was the complete elimination of reinforcement of maladaptive student be havior by the experimental classroom teacher. During the post-experimental observation period she was not once ob served acting or responding in a way that would maintain or increase maladaptive behavior. Comparisons of the Two Classes The t-test for difference between uncorrelated means in two samples of equal size (Guilford, 1965, p. 184) was used to determine if there was a significant difference between the two mean Activity Level Averages (ALA's) of the experimental and contrast class. A comparison of the pre-experimental period ALA's showed that the contrast class displayed significantly few er maladaptive behaviors than did the experimental class (t = 3.14, df's - 15, p > .01), whereas a similar analysis of the post-experimental ALA's revealed a significant dif ference in the opposite direction, i.e. the experimental class had fewer maladaptive behaviors than the contrast class (t = 10.34, df's = 15, p > .001). These analyses are summarized in Table 14. 12 3 For the Sound Level Meter means of the two classes, the t-test for testing a difference between uncorrelated means was used (Guilford, 1965, p. 182). A comparison of the pre-experimental period Sound Level Meter readings showed that the contrast class displayed a significantly lower mean sound level than did the experimental class (t = 5.9, df’s = 87, P > .001), whereas a similar analysis of the post-experimental period sound level means revealed a significant difference in the opposite direction, i.e. the experimental class had a significantly lower mean sound level than the contrast class (t = 15.5, df's = 93, p. > .001). These analysis are summarized in Table 15. On the third instrument, the Teacher Observation Form, the pre-experimental period scores, and then the post- experimental period scores, were compared by means of the Chi Square statistic (Garrett, 1962, 264-266). A compari son of pre-experimental period scores revealed that there was no significant difference between the two teachers (X2 = .037, df = 1, p > .80), whereas post-experimental period scores were significantly different (X2 = 76.7, df = 1, p. > .001). The experimental teacher had decreased the ratio of reinforcement of maladaptive behavior to zero while the contrast teacher, who had had no explanation of operant techniques, had a behavior reinforcement ratio very similar to her pre-experimental period ratio. These 124 analyses are summarized in Table 16. For the Burks1 Behavior Rating Scale, the t-test for differences between uncorrelated means in two samples of equal size were used to determine if there was a singi- ficant difference between the two classes (Guilford, 1965* p. 184). A comparison of pre-experimental scores reveal a significant difference (t^ = 6.79, df = 15, p. > .001), i.e. the experimental class had significantly higher scores, indicating more maladjustment. A similar comparison of post-experimental period scores revealed no difference (t = 0.6, df = 15, p > .05). These analyses are summarized in Table 17. 125 TABLE 11* COMPARISON OP THE ACTIVITY LEVEL AVERAGE OF THE EXPERIMENTAL CLASS WITH THE CONTRAST CLASS Experimental Class Contrast Class Pre-Experimental Period Observations Number of Pupils 8 8 Class Activity Level Average 1. 83 1.51 Standard Error of the Mean .09 .15 t = 3. 33** seM1-M2 = .081 Post-Experimental Period Observations Number of Pupils 8 8 Class Activity Level Average .53 1. 42 Standard Error of the Mean .04 .18 t = io.34*** seM2-M2 = .086 ## — Significant at the .01 level of confidence. ### — Significant at the .001 level of confidence. TABLE 15 BETWEEN-CLASS COMPARISONS OP SOUND LEVEL METER DATA IN PRE- AND POST-EXPERIMENTAL PERIODS Experimental Contrast Mean t Class Class Difference Value Pre-Experimental Period Data Number of Readings 41 48 Standard Deviation 5.64 5.17 Mean of Readings 72. 8 66.0 6.8 5.9*** Post-Experimental Period Data Number of Readings 49 46 Standard Deviation 2.23 4.86 Mean of Readings 56 68 12 15.5*** *«» — Significant at the .001 level of confidence. H ro o\ TABLE 16 BETWEEN-TEACHER COMPARISON OP TEACHER REINFORCEMENT OF ADAPTIVE AND MALADAPTIVE PUPIL BEHAVIOR Combined Observations of Three Judges Experimental Class Teacher Contrast Class Teacher Pre-Experimental Period Total Reinforcement of Adaptive Behavior 61 99 Total Reinforcement of Maladaptive Behavior 88 92 Chi Square = .0 37a Post-Experimental Period Total Reinforcement of Adaptive Behavior 82 79 Total Reinforcement of Maladaptive Behavior 0 67 Chi Square = 76.7*** a — No significant difference between groups. ««* — Significant at the .001 level of confidence. M ro TABLE 17 BETWEEN-CLASS COMPARISONS ON BURKS’ BEHAVIOR RATING SCALE Experimental Class Contrast Class Pre-Experimental Period Number of Pupils 8 8 Mean of Class on Burks’ Scale t = 6.79*** 99.1 73.6 Post-Experimental Period Number of Pupils 8 8 Mean of Class on Burks' Scale t = .6oa 65-5 63.2 a — No significant difference between means. ««* — Significant at the .001 level of confidence M ru 03 CHAPTER V COLLATERAL INFORMATION This chapter presents reports of changes in student behavior and classroom activity and academic progress of the experimental class and contrast class observed and not ed by the classroom teacher, parents, other school person nel and the writer. Changes Occurring during the Period of the Study In addition to the reduction in maladaptive be havior reported above, a number of other changes occurred in the school setting while the study was in progress. Schoolwork The teacher estimated that the time spent on "seat- work" increased from less than 30 minutes to more than two and one-half hours per day. Her estimate of the academic progress of the boys is based on the comparison of the school work they were doing before and after the study and is contained in a letter to the writer. (Appendix C.) This letter is summarized and presented in Table 22. All boys made significant improvements in reading skills, in every case more than one month, in four cases more than 129 one year, and in two cases more than two years. All boys improved in writing, with four becomming effective with cursive writing for the first time. One boy was slightly improved in arithmetic, one improved, one much improved, one boy made about a year's progress, another two year's progress and another three years progress. The teacher re ported that all subjects had made more academic improve ment in the 22 day experimental period than they had made in the entire school year up to the start of the study. Two boys, she believed, had demonstrated more progress during the period than they had made in their entire school history. One of these boys E-7, was returned to a regular classroom for one hour per day at the start of the third week of the experimental period. This had no apparent ef fect on the study. Reduction in Self- Centered Thinking As the study progressed the subjects became more concerned with what the class would earn in the experimen tal hour than with what each individual would receive. Dur ing the first week of the study, subjects who were ''mad" at a classmate would often threaten to "mess" up the game for all. This threat was usually made once or twice each day by different boys but was not made by any student after the second week of the study. 131 TABLE 18 SUMMARY OF TEACHER'S EVALUATION OF INDIVIDUAL ACADEMIC PERFORMANCE DURING STUDY From Classroom Teacher's Letter; Appendix E Reading Writing Arithmetic Other E-l about one year** improved slightly improved E-2 Improved improved E-3 1 year Improved* much improved E-4 Improved much improved much improved E-5 more than 2 years** much improved about 1 year** E-6 approx. 1*5 years** much improved much improved E-7 approx. M years** very much improved 3 years** requests homework E-8 1 year* much improved more than 2 years** requests homework — refers to Improvement at end of study. — refers to Improvement at end of school year, or, from time study started to time letter was written. 132 On the playground, the teacher-aide reported an Increase in spontaneous group activities such as a game called kick-ball, which was played like baseball only using a soccer ball that was kicked instead of batted. Some games lasted the entire recess period and there was neither adult encouragement nor participation. Previous to that time playground activities had usually been limited to in dividual play and games involving only two or three. Another indication of a growing group awareness was the increase in the use of such words as "we," "us," "our class," etc. when talking with each other and while talk ing with adults. This change was noticed by all adults who worked with the class and I was apparently a gradual devel opment. Parents reported that their children talked more about "their class," how different it was because they really worked and studied like other children; that their children talked about school at home where before they had seldom emntioned it. Fighting and Aggression Before the start of this study the teacher reported that fighting and aggressive acts (throwing things, verbal threats, abusive language) against other children both in the classroom and on the playground occurred several times each day. This fact was also substantiated by the school principal. Students had been frequently sent home for un controlled fighting or aggressive acts or behavior or were 133 kept In the school office for several hours each day. There was a steady decrease in fights and aggressive behavior during the study and in the final eight days no fights of any kind occurred either on the bus, the playground, or in the schoolroom. Behavior on the School Bus The teacher-aide, who rode regularly to and from school in the bus which transported all educationally handicapped children in the three classes at Franklin School, reported that during the course of the study the bus behavior of all children improved. She reported that the experimental class children "obeyed" much better and were easier to control which in turn seemed to affect the behavior of the other children. The bus driver also re ported that behavior had improved greatly although he em phasized that it "still had a long way to go." Post-Experimental Period Behavior A two-month follow-up, in addition to weekly visits to the experimental class showed that adaptive behavior had continued to improve (see Teacher's Letter, Appendix C). The teacher had observed no evidence of regression to for mer behavioral patterns other than an occassional short "relapse" which she was able to control. She also stated that all subjects had maintained their academic improvement and that one pupil (previously mentioned) had advanced to a 134 4th grade reader from the pre-primer level one he was read ing at the start of the study. He was also spending up wards of two hours a day in regular classes. She felt that a definite class pride and cohesiveness was developing along with a growing concern for other students and the class. The principal reported that the class was as quiet and orderly as any class in the school and that students of the experimental class sat quietly at their desks approxi mately three out of the four hours that they were in school. He reported that before the start of the study he had been called to the room several times each day to quell a dis turbance or remove a child but that in the last two weeks of the study and the following six weeks he had not been called to the experimental class room once because of a behavior problem. Only one child had been sent home during that period and that for a playground fight. Three new students, all of whom had been diagnosed as extremely hyperactive and as severe classroom behavior problems, were placed in the class the Monday following the termination of this study. The teacher noted that none of those students ever disrupted the classroom procedure; they very quickly fell into the routine of the class and the teacher reports a substantial drop in hyperactivity in all three. 135 Home Behavior This section is based upon the results of the par ent questionnaire, parent comments at a post-experimental period regular parent's meeting, and personal communication between parents and the writer. Without exception, every parent whose child had been in the experimental class reported improvement, and generally a very dramatic improvement, in their child's behavior at home. The reduction in hyperactive and mal adaptive behavior found in the classroom also was evident in the home situation. The most common comment from par ents was to the effect that the child was able to sit quietly much longer than previously. Improvements in other behavior was noticed: the child obeyed better, did not cry as much, had fewer temper tantrums and aggressive outbursts. The stepfather of one subject who had been sched uled for exclusion from school prior to the study and who was under consideration for regular class placement at the termination of the study, attended his first parent's meet ing to tell of the boy's improvement at home and his de cision not to sue the school district for his son's condi tion. The mother of subject No. E-7, a boy who had been under intensive psychiatric care and who also had been scheduled for exclusion from school because of his class room behavior and aggression, wrote as follows: 136 He is more patient when he has to wait his turn to speak. Just more pleasant all around. Can occupy him self for much longer periods than ever before. . . . Of course he is proud of his progress and the fact that he can spend some time in a regular classroom. . . . The hardest thing for [S] to do was to wait his turn in anything and not speak out or make noises for atten tion. His improvement has been of tremendous help to him. Also his reading abilities have made such strides that he is now proud to bring home a book or work to do and is also able to help his 1st grade sister with her work. He can sit and play Monopoly with his sisters and not get so impatient. Also is attempting scrabble with his 1st grade sister, but not ready to try it with his older sisters yet. The mother of subject No. E-2, a boy who had had severe behavior problems throughout his school career, wrote the following on the Parent Questionnaire: When he wants your attention he raises his hand in stead of hollering it out or interrupting. . . . Likes school now more than when he started it this year and is trying to learn more about how to count and read than before. Although to sit down and read for any length of time is still out of the question, [S] seems more affectionate and more inquisitive. He asks me about adding and if he hears a word he doesn’t know he asks what it means. Also, he trys to add money. [S] is a bed wetter and I’ve noticed that his bed has been dry more frequent than before the tests. I’m happy to have been able to see the results of these tests [Report by writer to Parent's Meeting]. They all were favorable as far as [S] goes. I think we would have seen more improvement if [S] hadn't got his finger hurt and could have been in the whole time. I want to thank you for the chance to have been a part of it and also, for the improvement its made in [S], Appearance of New Undesirable Symptoms or Behavior One question that gave considerable cause for 137 concern before the start of the study was what would happen when maladaptive behavior was extinguished. For example, would new symptoms occur in other areas or would aggressive acts and hyperactive and maladaptive behavior increase out side the school setting as a consequence of their not being experessed in the school setting in accordance with the medical model of behavioral disorder (Krasner & Ullman, 1965, pp. 2-9)? The teacher, principal, school nurse, and other guidance and school personnel were requested to watch for any new symptoms or new behavior disorders and to immediate ly inform the writer if any were observed. The Parent Questionnaire was worded specifically to elicit any infor mation about any new symptoms or undesirable behavior that might have occurred. Parents also were asked specifically and individually if they had observed any such behavior or symptoms since the start of the study. The writer did not observe any new undesirable symptoms or behavior during the study, nor did any parent, teacher, or other adult report any such behavior. Rather, the opposite pattern emerged as previously reported— as school behavior improved, the behavior of the students im proved in other areas as well. Contrast Class Data The contrast class teacher reported (personal communication) that she believed that her class had been making improvement at a constant rate all year and that such improvement as occurred in her class during the experi mental period was about what she had expected. She did believe that the presence of the experimenter in her class room each day was a definite factor in the reading improve ment of two boys that he worked with frequently. She also reported that the presence of a man (the experimenter) who participated in games during the physical education-recess period was of positive value in encouraging the children to play more as a group. The school bus driver and regular teacher-aide re ported improvement (already discussed above) in bus be havior which they felt was mainly due to the improvement of the boys in the experimental class. In the post-experimental period parent meeting there were no reports of dramatic improvement from contrast class parents similar to those from all experimental class par ents who attended the meeting. CHAPTER VI SUMMARY AND DISCUSSION S ummary This study was concerned with the extinction of maladaptive behavior of students in a public school class room for educationally handicapped children. More speci fically, it sought to determine whether operant condition ing techniques, similar to those used and developed by be havior therapists (Wolpe, 1958; Eysenck, I960; Allyon, et al. 1959; Staats & Staats, 1963; Staats, 196M; Krasner & Ullman, 1965) for use with individual adults and children could be extended to use with a-class-as-a-whole in a class room of children with severe behavioral disorders. Purpose It was hypothesized that, as a result of the use of operant techniques, several measurable changes would occur in classroom and individual behavior. First, it was hypo thesized that there would be a measurable decrease in teacher disapproved, or maladaptive, behavior after the ex perimental period. It was hypothesized that there would be an increase in adaptive behavior and a decrease in mal adaptive behavior as a result of the application of operant 139 140 techniques. The second hypothesis was concerned with the back ground noise or activity level of the room and predicted that a decrease In sound level would occur as a result of the use of operant techniques. If maladaptive behavior could be reduced it appeared that a consequent result would be a quieter and less noisy classroom. The third hypothesis dealt with the use of operant techniques by the teacher. It was predicted that as she became aware of her own part in the reinforcement of mal adaptive behavior she would systematically reward adaptive behavior while ignoring or refusing to reward maladaptive behavior. Lastly, it was hypothesized that there would be a reduction of hyperactive and maladaptive behavior in the total school setting as a result of operant techniques used in this study. It was predicted that the teacher would observe such a decrease which would be reflected in her observations and evaluations. Thus, the major purposes of this study were to determine whether the systematic application of operant techniques in a classroom would result in (1) a reduction of maladaptive behavior in the classroom, (2) a reduction of the noise level in the classroom (3) a reduction in previously teacher-reinforced maladaptive behavior and (4) mi a reduction In hyperactive and maladaptive behavior In the total school setting. Methodology The above purposes were met by the utilization of a minimum experimental period of one hour daily for 22 consecutive school days. During the daily experimental period, students played a "game" where they were rewarded with candy or pennies when no member of the class committed any maladaptive behavior for a specified time interval. If maladaptive behavior occurred, the large stop-clock which was used to time the intervals was stopped and re turned to its starting point and then restarted as soon as the class became quiet or the maladaptive behavior ceased. The class was started on 20-second Intervals which were increased gradually to 10-minute intervals by the end of the third week. The one-hour experimental period also was increased to approximately one hour and 50 minutes by the middle of the third week. The teacher was instructed to reward freely any adaptive behavior and to ignore any mal adaptive behavior. At the conclusion of each daily period, the reward earned was divided equally among the children present. Children could exchange pennies for candy, take the money home with them, or place it in a "bank" operated by the writer. All subjects were saving their money by the termination of the experiment. 142 No instructions were given to the teacher on sub ject matter used during the daily period and she used both individual and class instruction on numerous topics. Data Collection and Analysis Baselines for the occurrence of maladaptive be havior in the class were obtained by having three judges rate the experimental class and a contrast class over a three day period on an Activity Check List. A Sound Level Meter was used to establish the mean sound level of the two classes. Teacher reinforcement of both adaptive and maladaptive behavior was recorded on a Teacher Observation Form. The two teachers were asked to rate the total school behavior of each student in their respective classes on the Burks Behavior Rating Scale. Following the 22-day period the same procedure was followed with the same judges and teachers. Pre- and post-experimental scores were compared. Each class was compared with itself in order to test the null hypotheses since the contrast group was not a true control group. The purpose of the contrast class was to determine what effect the presence of the experimenter would have on the children and whether the contrast class would show similar progress with the presence of the experimenter but without the use of systematic operant techniques. 143 Subjects The sample consisted of two special classes for edu cationally handicapped children located in Franklin Ele mentary School in Redondo Beach, California. All children in both classes had been excluded from regular classes be cause of severe behavior disorders and learning difficul ties. All in the experimental class had been classified as hyperactive and all but one had been excluded from their regular class because of their disrupting behavior in class. The experimental class was composed of eight boys of normal or above intelligence, between the ages of 8.8 and 10.6. They were all Caucasian. The contrast class was also com posed of eight children, five boys and three girls, of normal or above intelligence and ranging in age from 7.3 to 10.10 years. Four of the eight had been described as hyperactive. All were Caucasian. Major Findings Four research hypotheses were developed from the general hypothesis that the use of operant techniques can be extended to an entire class of educationally handicapped children to improve classroom behavior. Hypothesis I— predicted that the use of operant techniques would result in a reduction in the occurrence of maladaptive behavior. This prediction was supported by the data in every analysis that was performed. There was no significant change in occurrence of maladaptive behavior in the control class during the same period. Hypothesis II— predicted that the use of operant techniques would reduce the noise level of the class. This hypothesis was strongly supported by the data. There was no significant difference in the pre- and post-experimental sound levels of the contrast class. Hypothesis III— predicted that the systematic appli cation of operant techniques by the teacher would result in less reinforcement of maladaptive behavior. This hypo thesis was supported by all analyses of the data. The analysis of the data on the contrast class showed no change in the rate or amount of reinforcement of adaptive and mal adaptive behavior. Hypothesis IV— predicted that there would be a re duction in maladaptive behavior in the total school setting as a result of the use of operant techniques in the experi mental period. This hypothesis was strongly supported by data. Data reported by the contrast class teacher were significant at the .05 level but not the .01 level. Supplementary Findings In addition to the findings related to the research hypotheses the following observations were recorded for the experimental class. 145 1. Every student reduced his pre-experimental period Activity Level Average by at least two-thirds in the post-experimental period. 2. Teacher evaluations on the Burks Behavior Rating Scale showed that all subjects had reduced their pre-experi mental period score by at least 25 per cent. 3. There was no increase in maladaptive behavior by any subject on any scale. 4. There was a total absence of teacher reinforce ment of maladaptive behavior during the post-experimental observation period. Additional Findings Additional analyses of the data reported on between- class comparisons of the two classes. The activity levels of the two classes as measured on the Activity Check List showed that the contrast class had significantly less maladaptive behavior than the experimental class before the start of the study. Post-ex perimental period observations indicated that the position had reversed: the experimental class had significantly less maladaptive behavior. Sound Level Meter readings indicated the contrast class significantly quieter in the pre-experimental period but the experimental class was significantly quieter in the post-experimental period. 146 Ratings of the two teachers show no differences in their rate of reinforcement of maladaptive and adaptive be havior in the pre-experimental period. In the post-experi mental period however, the experimental teacher was signi ficantly different— she did not want to reinforce a single instance of maladaptive behavior while being observed. The contrast teacher’s rate of reinforcement was similar to her previous rate in the pre-experimental period. On the Burks Scale the ratings of the teahhers showed that the contrast class had significantly lower ratings be fore the start of the study. In the post-experimental period the two classes both had lower scores but the experi mental class had the greatest reduction in scores and was rated equal to the contrast class. (See Table 17.) Collateral Findings In addition to findings from the above systematic data, other collateral findings based on experimental and post-experimental period follow-up observations of children in the experimental class were reported. 1. Schoolwork: There was a substantial improvement in quantity and quality of schoolwork, academic achievement and personal attitudes toward school of all subjects in the experimental class. 2. There was a decrease in self-centered thinking and an increase in group concern and involvement by every child. 3. There was a marked decrease in fighting and aggressive acts by all children in the experimental class. Behavior on the school bus was substantially improved. 5. The above findings were maintained during the balance of the school year— a period of ten weeks. 6. Behavior at home was improved and maladaptive behavior was reduced. 7. There were no manifestations of any adverse effects or symptom-substitutions in any student involved in this study. Discussion The results of this study have supported the gen eral hypothesis that maladaptive behavior in a classroom for educationally handicapped children can be reduced. The data shows that all four singular hypotheses were supported well beyond the .001 level of confidence while in all but one instance, there was no significant improvement in a contrast class. This study was demonstrated that hyperactive and maladaptive behavior of boys who are either physically nor mal or who had neurological impairment can be altered and reduced through a systematically applied technique of be havior modification. Indeed, the impressions of school personnel and the experimenter of the boys in this study 148 was that there was no obvious difference between normal and neurologlcally impaired boys in their rate of improve ment, nor could any difference be detected on the instru ments used. The experiment sought to define a certain type of maladaptive behavior, observe it and then modify it in the manner described by Ferster & DeMeyer (1961). Patterson, Jones, & Whittier (1965) hypothesized that hyperactive behavior resulted in an aversive effect upon other people; that the child with hyperactive or mal adaptive behavior may well be seldom, or never, rewarded for acceptable behavior when he does produce it. This hypothesis seemed to also apply with the boys in this study: the behavior they emitted was not so much ' ’wrong1 ' as simply excessive or inappropriate for the situation. A child wanting help, for example, might call out to the teacher or run to her desk and interrupt whatever she was doing to seek aid instead of sitting quietly at his desk and holding his hand up until appropriately recognized. Patterson & Anderson (1964) and Perster & DeMeyer (1961) have also pointed out that the teacher may not be a reinforcing agent to many of these children and her praise or approval is not reinforcing and may even be aversive to the child. This situation also existed for some of the boys in the class who were often quite unresponsive to any attempts on the part of the teacher to use verbal methods of control. The three boys who were to be excluded, 1H9 for example, were frequently unresponsive to any method of control other than physical restraint. This experiment sought to establish an optimum setting wherein all children would be paid for emitting adaptive, socially acceptable behavior. They were paid with token rewards— money, with which they could do as they saw fit. Another purpose of the study was to couple the token reward with systematic teacher social and verbal reinforcement given at the same time, and as much as pos sible, during the experimental period, i.e. the teacher was rewarding the same behavior the child was being paid to perform: she made no comment about the experiment and pre tended not to notice it. Her function was to create a situation wherein adaptive behaviors of all boys were con stantly and conspicuously reinforced while at the same time she completely ignored and did not reinforce maladap tive behavior. The Lott & Lott (I960) study suggested that if the pupils in a class experienced success and social approval in the presence of other pupils, positive feelings toward the other class members would increase and negative or aversive attitudes would decrease. Their study also sugges ted that if a situation could be created where the students worked together successfully and all were rewarded, a group "cohessiveness" or "esprit de corps" would develop. This experiment required all boys to cooperate and work together 150 In order to win or earn money and that situation quickly resulted in disapproval of any maladaptive behavior and mutual approval for good behavior. Counterconditioning, responce substitution in effect took place as a child learned new ways to be rewarded in place of no longer ef fective, and unrewarded, maladaptive ways. Reward There were three main sources of reward in the experimental period: (1) from the teacher; (2) from peer approval and support and (3) from token reinforcement given by the experimenter. There was also what might be termed a generalized reward of "achievement" which devel oped as the boys became interested in their academic pro gress. Patterson (1965) has suggested that in many cases the deviant child has a behavior disorder because certain avenues for reinforcement are not open to him. In this study, in many instances neither teacher nor peer approval was reinforcing before the start of the experiment. How ever, during the experiment as maladaptive behavior de creased there was a very obvious change in inter-pupil be havior and class opinion came to exert a definite influence on every child. The coupling of social reinforcement with the token rewards also resulted in an increasing effective ness of verbal reinformecment alone. Contrast Class Because of the nature of the study, support for the hypotheses was derived from pre- and post-experimental com parisons of the same class. However, comparisons made be tween the two classes also strongly support the hypotheses. On the two measures in which both classes were scored by three judges common to both classes in the pre-experimental period the contrast class had a much lower Activity Level Average and a much lower room sound level. The difference was significant beyond the .001 level of confidence. On the between teacher ratings, there was no difference. In the post-experimental period the situation was reversed: the experimental class was far lower on both measures than the contrast class. The difference was significant at the .001 level. The experimental teacher had reduced her reinforcement of maladaptive behavior to zero while the contrast teacher still reinforced maladaptive behavior in the same proportion. On the Burks' Scale, an instrument used by each teacher to rate her own class, the mean of the contrast class was far lower (at the .001 level) than the experimental class before the study period but the means were equal at the end of the study. Both had made signi ficant improvement. Part of the improvement of the con trast room on the Burks' Scale, which reported total school behavior, undoubtedly reflected the behavior change in the experimental room. The children in the three classes rode 152 to school together, were in adjacent classrooms, played together at recess and frequently had Joint activities together and the boys in the experimental class, being older usually set the "tone" of any activity in which all were involved and were generally the leaders in disruptive activity. Matching of Groups Shortly after the idea of this study was formulated it was determined to attempt to eliminate some of the strong criticisms Bettleheim (1963) leveled at the Haring & Phillips (1961) study of emotionally disturbed classes. Bettleheim was particularly critical of the method of selecting and matching the experimental and control groups: Irrespective of the nature of the groupings, one might surmise that whatever method Group I was taught by— being much younger, being in classes about half the size of Group III, having known defeat only half as many years— their academic and behavioral progress would be better than in the other two groups, (p. 333) He was also critical of the method of parent conferences with experimental class parents while none were undertaken with the control class parents. In the first discussions with the director of the South Bay Area School Districts for permission to do this study, the experimenter specifically requested that the study be done with the class containing the most severely maladjusted and hyperactive children of the 13 classes for educationally handicapped children in that district. The 153 class selected contained three boys who had been processed for exclusion from school because they were unmanageable. The contrast class had none. The experimental class had a mean age of 9.5 and a mean of 13.2 months since placement in the EH program; the mean age of the contrast groups was 8.7 and mean length of time since placement, 9*7 months. Thus, the experimental class was older, exhibited more mal adaptive behavior and had been the program much longer; it had had far more time to "know defeat," both in life and in school. Also, at the start of the experiment, it was com posed of nine boys to eight boys and girls in the contrast class. One boy dropped out when his family moved to another district leaving the classes exactly matched on size until the 3rd week when a ninth student (C—9) was placed in the control class: a very quiet withdrawn girl, (see C-9). The last criticism, that Haring & Phillips worked with one parent group only, was met in this study by the experimen ter is appearing before a parent-group meeting held for parents of all three classes once each month and explaining the purpose of the study. All parents were equally in formed and no attempt was made to contact or "work-with" any parent or parents in the study. Following the study, the experimenter met with all parents to discuss the find ings . The specific purpose of the matching of these classes was, in fact, to "stack the cards" against any 151 * chance of showing a significant improvement when the two classes were compared. After the conclusion of the study it was the opinion of some members of the district guidance office, (Personal communications) that the experimental class had become one of the best behaved, if not the best behaved, of all EH classes in the district. Generalization of Behavior Change Observations and impressions of parents and school personnel were also collected in an effort to determine whether any generalization of adaptive behavior had occur red outside the experimental period and outside the class room. There was unanimous agreement among school personnel that a dramatic change in overall behavior had occurred and every parent reported a substantial improvement in behavior; two reported a very dramatic change. Parents re ported many instances of specific improvement at the that took place in other areas as the study progressed. For exmmple, a decrease in enuresis, improved sibling relations, less fights, improved table manners, and less hyperactive behavior in all situations. To summarize, behavior was modified in all children in the experimental class as had been predicted from learn ing theory: concrete reinforcement of specific behavior increased its occurrence while competing maladaptive be havior was decreased. As children became successful and 155 were rewarded in school, expressions of liking and approval became more frequent and anxiety and dislike decreased, as predicted by Wolpe's theory. Finally, in every instance, behavior that was systematically rewarded increased and behavior that was not rewarded and made non-reinforcing to the person emitting it decreased, as predicted by operant conditioning theory. Conclusions The following conclusions relative to the problem under investigation are suggested by the evidence provided within the limits of this investigation, and in relation to previous research findings pertinent to the same pro- blems. 1. The use of operant conditioning techniques can effectively reduce the occurrence of maladaptive behavior in a classroom for educationally handicapped children. 2. Maladaptive behavior of individual children can be defined, observed and specifically extinguished within the framework of normal classroom procedures. 3. As behavior improves in the classroom the improvement will generalize to other school situations such as playground behavior, bus behavior, etc. As maladaptive behavior decreases academic progress increases. 5. As the classroom becomes a source of satisfac tion, success, and reward, dislike, hostility, and anxiety toward the school environment will decrease. 6. A classroom teacher can effectively apply operant conditioning techniques to shape behavior in the course of daily teaching and can be taught how to reduce substantailly the occurrence of maladaptive behavior by (a) ignoring it and/or not rewarding it and (b) strongly reinforcing competing adaptive behaviors systematically. 7. It is unnecessary to use dynamic theories of personality to reduce, or explain, maladaptive behavior. 8. The elimination of maladaptive behavior does not result in the emergence of other unacceptable responses; rather, the opposite appears to take place: a decrease in other types of maladaptive or neurotic behavior. R E F E R E N C E S 157 REFERENCES Allen, Eileen K., Hart, Betty K., Buell, Joan S., Harris, Florence R., & Wolf, Montrose M. Effects of social reinforcement on isolate behavior of a nursery school child. Child Development, 1964, 35., 511-518. Anderson, D. A. An application of a behavior modification technique in the control of hyperactivity of a brain damaged child. Unpublished master's thesis, Univer sity of Oregon, Eugene, 1964. Ayllon, T. Some behavioral problems associated with eat ing in chronic schizophrenic patients. Paper read at the meeting of the American Psychological Association, Chicago, I960. Ayllon, T. Intensive treatment of psychotic behavior by stimulus satiation and food reinforcement. Behavioral Research and Therapy, 1963, 1* 53-61. Ayllon, T., & Haughton, E. Control of the behavior of schizophrenics by food. Journal of the Experimental Analysis of Behavior, 1962, 5., 343-352. Ayllon, T., & Michael, J. The psychiatric nurse as a behavioral engineer. Journal of the Experimental Analysis of Behavior, 1959, 2_, 323-334. Azrin, N. H., & Lindsley, 0. R. The reinforcement of cooperation between children. Journal of Abnormal and Social Psychology, 1956, 5jL, 100-102. Baer, D. M. Laboratory control of thumbsucking by with drawal and re-presentation of reinforcement. Journal of the Experimental Analysis of Behavior, 1962, 5, 525-528. Bandura, A. Psychotherapy as a learning process. Psychological Bulletin, 1961, 5J3, 143-157. Barnett, C. D., Pryer, Margaret W., & Ellis, N. R. Experimental manipulation of verbal behavior in de fectives. Psychological Reports, 1959, 5., 393-396. 158 159 Bentler, P.M. An infant’s phobia treated with reciprocal inhibition therapy. Journal of Child Psychology and Psychiatry, 1962, 185-189". Bettleheim, B. A noncontribution to educational research. Harvard Educational Review, 1963, j£3., 326-335. Bijou, S. W. Discrimination performance as a baseline for individual analysis of young children. Child Development, 1961, 32_, 163-170. Bijou, S. W. Application of operant principles to the teaching of reading, writing, and arithmetic to re tarded children. In: New frontiers in special educa tion, 43rd Annual Council for Ecceptional Children Convention, National Education Association, 1965. Bijou, S. W., & Baer, D. M. Child development: A syste matic and empirical theory. New York: Appleton- Century-Crofts, 1961. Birnbrauer, J. S., Bijou, S. W., Wolf, M. M., Kidder, J. D., & Tague, Cecilia. Programmed instruction in the classroom. In L. P. Ullman and L. Krasner (Eds.), Case studies in behavior modification, New York: Holt, Rinehart, and Winston, 1965. Birnbrauer, J. S., Bijou, S. W., Wolf, M. M., Kidder, J. D., & Tague, Cecilia. A programmed instruction classroom for educable retardates. In S. W. Bijou and D. M. Baer (Eds.), Readings in the experimental analysis of child behavior and development. New York: Appleton- Century-Crofts, 1966. Birnbrauer, J. S., Wolf, M. M., Kidder, J. D., & Tague, Celia A. Classroom behavior of retarded pupils with token reinforcement. Journal of Experimental Child Psychology. 1965, 2, 219-235. Breger, L., & McGaugh, J. L. Learning theory and behavior therapy: A reply to Rachman and Eysenck. Pschologlcal Bulletin, 1966, 65, 170-173. Burks, H. F. A study of the organic basis for behavioral deviations in school children. Unpublished doctoral dissertation: University of Southern California, 1955. California Legislature: Education Code of the State of California, 1965, Department of General Services, Documents section, Sacramento (2 volumes). 160 California Legislature: California Administrative Code, Title 5; Education, 1965, Office of Administrative Procedure, Documents Section, Sacramento. Cohen, D. J. Justin and his peers: An experimental analysis of a child’s social world. Child Development, 1962, 22, 697-717. Ellis, N. R., Barnett, C. D., & Pryer, Margaret W. Operant behaviors in mental defectives: Exploratory studies. Journal of the Experimental Analysis of Behavior, I960, 3, 63-69. English, H. B., & English, Ava C. A comprehensive dictionary of psychological and psychoanalytical terms. New York: David McKay, 1964. Eysenck, H. J. Learning theory and behavior therapy. Journal of Mental Science, 1959, 105, 61— 7^- Eysenck, H. J. (Ed.) Behavior therapy and the neuroses, New York: PergamonT I960. Ferster, C. B. The use of the free operant in the analysis of behavior. Psychological Bulletin, 1953, 50, 263-274. Ferster, C. B. Positive reinforcement and behavioral deficits of autistic children. Child Development, 1961, 32, 437-^56. Ferster, C. B., & DeMyer, M. The development of perfor mances in autistic children in an automatically controlled environment. Journal of Chronic Diseases, 1961, 12, 312-345. Flanagan, B., Goldiamond, I., & Azrln, N. Operant stutter ing behavior through response contiggent consequences. Journal of the Experimental Analysis of Behavior, 1956, 56, 49-56. Furniss, Jean M. Peer reinforcement of behavior in an institution for delinquent girls. Unpublished M.A. Thesis, Oregon State University, 1964. Garrett, H. E. Statistics in psychology and education. New York: David McKay, 1962. Greenspoon, J. Verbal conditioning and clinical psychology. In A. J. Bachrach (Ed.), Experimental foundations of clinical psychology. New York: Basic Books, 1962, pp. 510-513. 161 Grossberg, J. Behavior therapy— A review. Psychological Bulletin, 1964, 62, 73-78. Guilford, J. P. Fundamental statistics In psychology and education. New York: McGraw-Hill, 1965. Haring, N. G., & Phillips, E. L. Educating emotionally disturbed children. New York: McGraw-Hill, 1962. Harris, Florence R., Johnston, Margaret K., Kelley, Susan C., & Wolf, Montrose M. Effects of positive social reinforcement on regressed crawling of a nursery school child. Journal of Educational Psychology, 1964, £5, 35-41. Harris, Florence R., Wolf, M. M., & Baer, D. M. Effects of adult social reinforcement on child behavior. Young Child, 1964, 20, 8-17. Hart, Betty M., Allen, Eileen K., Buell, Joan S., Harris, Florence R., & Wolf, Montrose M. Effects of social reinforcement on operant crying. Journal of Experimental Child Psychology, 1964, 1, 145-153. Hewett, F. M. Teaching speech to autistic children through operant conditioning. Digest of a paper from the 42nd Annual Meeting of the American Orthopsychiatric Asso ciation, held at New York, March, 1965* Hollis, J. H. The effects of social and nonsocial stimuli on the behavior of profoundly retarded children: Part I. American Journal of Mental Deficiency, 1965, 6, 755-771. Ca5 Hollis, J. H. The effects of social and nonsocial stimuli on the behavior of profoundly retarded children: Part II. American Journal of Mental Deficiency, 1965, 6, 1 1 2 - l W - T F 5 Hutchinson, R. R., & Azrin, N. H. Conditioning of mental hospital patients to fixed-ratio schedules of rein forcement. Journal of the Experimental Analysis of Behavior, 1961, 4_, 87-95. Isaacs, W., Thomas, J., & Goldiamond, I. Application of operant conditioning to reinstate verbal behavior in psychotics. Journal of Speech and Hearing Disorders, I960, 25, 8-12. 162 Jones, Mary C. A laboratory study of fear: The case of Peter. In H. J. Eysenck (Ed.), Behavior therapy and the neuroses. New York: Macmillan, i960, pp. 45-51. Keller, P. S., & Schoenfeld, W. N. Principles of psychology, New York: Appleton-Century-Crofts, 1950. Kimble, G. H., Hllgard and Marquis1 conditioning and learning. (2nd ed.)5 New York: Appleton-Century- Crofts, 1961. Kounin, J. S., Priesen, W. V., & Norton, Evangeline A. Managing emotionally disturbed children in regular classrooms. Journal of Educational Psychology, 1966, 57, 1-13. Krasner, L. Studies in the conditioning of verbal behavior. Psychological Bulletin. 1958, 55, 148-170. Krasner, L., & Ullman, L. Research In behavior modifica tion. New York: Holt, Rinehart, & Winston, 1965. Landreth, Catherine, & Read, Katherine H. Education of the young child: A nursing school manual. New York: Wiley, 1942. Lawrence, D. H. Learning. Annual Review of Psychology, 1958, 9, 157-158. Lazarus, A. A. The elimination of children’s phobias by deconditioning. In H. J. Eysenck (Ed.), Behavior therapy and the neuroses. New York: Pergamon, I960, pp. 114-122. Lazarus, A. A. Group therapy of phobic disorders by systematic desensitization. Journal of Abnormal and Social Psychology, 1961, 63, 504-510. Lazarus, A. A., & Abramovitz, A. The use of "emotive imagery" in the treatment of children's phobias. Journal of Mental Science, 1962, 10 8, 191-195. Levin, G., & Simmons, J. Response to praise by emotionally disturbed boys. Psychological Reports, 1962, 11, 10, (a). Levin, G., & Simmons, J. Response to food and praise by emotionally disturbed boys. Psychological Reports, 1962, 11, 539-5^6, (b). 163 Lindsley, 0. R. Operant conditioning methods applied to research in chronic schizophrenia. Psychiatric Research and Reports, 1956, 5., 118-139. Locke, B. J. The effects of examiner, role, and reinforce ment variables on the modification of verbal behavior in institutionalized retardates. Unpublished Ph. D. dissertation, Oklahoma State University, 1962. Lott, Bernice E., & Lott, A. J. The formation of positive attitudes toward group members. Journal of Abnormal and Social Psychology, I960, 61, 297-300. Lovaas, 0. I., Preitag, G., Kinder, M. I., Ruberstein, B. D., Schaeffen, B., & Simmons, J. Q. Establishment of social reinforcers in two schizophrenic children on the basis of food. Paper delivered at the Western Psychological Association, Portland, 1964. Morse, W. C., Cutler, R. L., & Fink, A. H. Public school classes for the emotionally handicapped: A research analysis. The council for Exceptional Children, National Education Association, 1964. Orlando, R., & Bijou, S. W. Single and multiple schedules of reinforcement in developmentally retarded children. Journal of the Experimental Analysis of Behavior. I960, 4, 339-348. Patterson, G. R. Reinforcement of aggression by the peer group. Unpublished paper, University of Oregon, 1963. Patterson, G. R. Responsiveness to social stimuli. In L. Krasner and L. P. Ullman (Eds.), Research in behavior modification. New York: Holt, Rinehart & Winston, 1965. Patterson, G. R., & Anderson, D. Peers as social rein forcers. Child Development. 1964, 35, 951-960. Patterson, G. R., Bricker, W., & Green, M. Peer group reactions as a determinant of aggressive behavior in nursery school children. Paper presented at the American Psychological Association Convention, September 4, 1964. Patterson, G. R., & Ebner, M. J. Applications of learning principles to the treatment of deviant children. Paper presented to the American Psychological Association Convention, Chicago, 1965. 16 4 Patterson, G. R., Jones, R., Whittier, J., & Wright, Mary A. A behavior modification technique for the hyper active child. Behavioral Research and Therapy. 1965, 2, 217-226. Patterson, G. R., & Littman, R. Social reinforcers and social behaviors. Progress Report MH 08009-2, University of Oregon Child Study Laboratories, 1965. Quay, H. C. Some basic considerations in the education of emotionally disturbed children. Exceptional Children, 1963, 30, 27-31. Quay, H. C., Werry, J. S., McQueen, Marjorie, & Sprague, R. L. Remediation of the conduct problem child in the special class setting. Exceptional Children. 1966, 12, 509-515. Rabinovich, R. D. Reading and learning disabilities. Handbook of Psychiatry, Vol I. New York: Basic Books, 1959. Raymond, M. S. Case of fetishism treated by aversion therapy. British Medical Journal, 1956, 2, 854-857. Reger, R. Education and Behavior problem children: Some trends and some problems. Psychology in the School. 1965, 2, No. 2, 119-126. Reyna, L. J. Conditioning therapies, learning theory, and research. In J. Wolpe, A. Salter, & L. J. Reyna, (Eds.), The conditioning therapies. New York: Holt, Rinehart & Winston, 1965. Rimland, B. Infantile autism. New York: Appleton-Century- Crofts, 19PT Rubenstein, H., & Aborn, M. Psycholinguistics. Annual Review of Psychology. I960, 11, 291-322. Salzinger, K. Experimental manipulation of verbal behavior: A review. Journal of Genetic Psychology, 1959, 61, 65-94. Salzinger, S., Salzinger, K., Pisoni, S., Eckman, J., Mathewson, P., Dutsch, M., & Zubin, J. Operant conditioning of continuous speech in young children. Child Development, 1962, 31, 683-695. 165 Schaln, R. J., & Yannet, H. Infantile autism: An analysis of fifty cases and a consideration of certain neuro- physiologic concepts. Journal of Pediatrics, i960, 51, 560-567. Schwitzgebel, R. Short-term operant conditioning of ado lescent offenders on socially relevant variables. Mimeographed material, University of California at Los Angeles, 1965. Schwitzgebel, R., & Kolb, D. A. Inducing behavior change in adolescent delinquents. Behavioral Research and Therapy, 1964, l_, 297-304. Shannon, L. W. The problem of competence to help. Federal Probation, 1961, 25., 32-39. Sidman, M. Operant techniques. In A. J. Bachrack (Ed.), Experimental foundations of clinical psychology. New York: McGraw Hill, 1956. Siegel, S. Nonparametric statistics. New York: McGraw- Hill, 1 9 W T Skinner, B. F. The behavior of organisms. New York: Appleton-Century-Crofts, 1936. Skinner, B. F. Science and human behavior. New York: Macmillan, 1953. Skinner, B. F. The analysis of behavior. In B. F. Skinner (Ed.), Cumulative record. New York: Appleton-Century-Crofts, 1959. Spradlin, J. E., & Girardeau, F. L. The behavior of moderately and severely retarded persons. Bureau of Child Research, University of Kansas, Lawrence, Kansas. Pre-publication chapter for Ellis (Ed.), Mental retardation, in press. Staats, A. W. Human learning. New York: Holt, Rinehart and Winston^ 1964. Staats, A. W., & Staats, Carolyn K. Complex human be havior. New York: Holt, Rinehart and Winston, 1963. Thirman, J. Conditioned-reflex treatment of alcholism. New England Journal of Medicine, 1949, 241, 368-370, 406-4l§.------- ---------------- --- 166 Thompson, G. N., & Bielinski, B. Improvement in psychosis following conditioned reflex treatment in alcholism. Journal of Nervous and Mental Disease, 1953* 117, 537-543. Thorndyke, E. L. Animal Intelligence. New York: Macmillan, 1911. Ullman, L. P., & Krasner, L. Case studies in behavior modification. New York: Holt, Rinehart and Winston, 1 9 ^ Verplanck, W. S. A glossary of some terms used in the objective science of behavior. Psychological Review, 1957, (Supplement). Voegtlen, W. L. The treatment of alcoholism by establish ing a conditioned reflex. American Journal of Medical Science, 1940, 119, 802-810. Wallace, J. A. The treatment of alcoholics by the conditioned reflex method. Journal of the Tennessee Medical Association, 1949, 42, 125-128. Whelan, R. R., & Haring, N. G. Modification and mainten ance of behavior through systematic application of consequences. Exceptional Children. 1966, 3£, 281-290. Wild, J. Education and human society: A realistic view. In N. B. Henry (Ed.), Modern philosophies and education, society for the study of education, 1955, 54, Part I. Williams, C. D. The elimination of tantrum behavior by extinction procedures. Journal of Abnormal and Social Psychology, 1959, 59, 269-270. Wolpe, J. Psychotherapy by reciprocal inhibition. Stanford: Stanford University Press, 1958. Wolpe, J. Isolation of a conditioning procedure as the crucial psychotherapeutic factor: a case study. Journal of Nervous and Mental Disease, 1962, 134, 316-329. Wolpe, J., Salter, A., & Reyna, L. J. The conditioning therapies. New York: Holt, Rinehart & Winston, 1965. Zimmerman, Elaine H., & Zimmerman, J. The alteration of behavior in a special classroom situation. Journal of the Experimental Analysis of Behavior, 1962, 5.,59-60. Zimmerman, J. United States Department of Public Health Application for Research Grant, 1965. A P P E N D I X E S i 168 APPENDIX A ACTIVITY CHECK LIST— RAW DATA 169 TABLE 19 SUMMARY OF RAW SCORE DATA FROM PRE-EXPERIMENTAL PERIOD ACTIVITY CHECK LISTS OF ALL JUDGES Experimental Class Subject No. Movements Directed Toward Body Move ments In Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 15 18 42 21 31 52 27 115 E-2 16 24 48 15 18 44 50 133 E-3 9 2k 45 33 32 40 31 141 E-4 25 35 53 30 41 60 33 140 E-5 16 39 54 4l 36 52 49 140 E-6 18 13 34 21 18 51 52 138 E-7 15 40 42 28 53 54 58 134 E-8 17 36 41 36 50 51 46 132 TOTALS 131 229 359 225 289 40 4 345 10 73 TABLE 20 SUMMARY OF RAW SCORE DATA FROM POST-EXPERIMENTAL PERIOD ACTIVITY CHECK LISTS OF ALL JUDGES Experimental Class Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-1 20 14 36 18 20 4 1 180 E-2 14 19 16 16 5 8 2 180 E-3 16 14 25 27 3 1 4 180 E-4 17 24 18 22 33 6 1 180 E-5 11 22 36 18 5 0 4 180 E-6 12 10 16 25 4 6 7 180 E-7 18 14 23 21 4 3 1 180 E-8 20 21 32 31 14 3 1 178 TOTALS 12 8 138 202 178 88 31 21 1438 TABLE 21 SUMMARY OF RAW SCORE DATA FROM PRE-EXPERIMENTAL PERIOD ACTIVITY CHECK LISTS OF ALL JUDGES Contrast Class Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 12 11 30 16 18 19 60 96 C-2 23 40 40 35 49 28 33 106 C-3 3 15 22 13 10 14 33 110 C-4 7 20 39 22 11 15 14 110 C-5 16 14 39 14 11 28 20 95 C-6 9 10 42 27 15 35 29 115 C-7 12 17 52 13 17 27 20 110 C-8 20 22 35 20 15 28 44 110 TOTALS 102 149 289 160 146 195 253 852 TABLE 22 SUMMARY OF RAW SCORE DATA FROM POST-EXPERIMENTAL PERIOD ACTIVITY CHECK LISTS OF ALL JUDGES Contrast Class Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com municative Activity Walking or Standing Total Obser vation Periods C-l 25 17 36 17 13 11 12 90 C-2 31 40 28 45 29 19 25 90 C-3 7 17 18 14 26 15 18 90 C-4 9 19 16 25 10 5 6 90 C-5 17 16 25 13 10 12 9 90 C-6 19 25 32 30 13 7 6 90 C-7 21 23 24 19 20 13 16 90 c-8 33 40 35 25 17 26 23 90 TOTALS 162 197 214 188 138 108 115 720 TABLE 2 3 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Experimental Class Judge No. 1 Pre-Experimental Period Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communi cat!ve or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 4 6 13 4 4 18 4 82 E-2 13 12 16 9 12 21 18 53 E-3 1 6 16 12 8 12 15 53 E-4 5 7 12 11 15 23 13 53 E-5 11 16 14 13 25 18 53 E-6 6 2 18 6 18 21 25 54 E-7 5 19 24 10 22 25 31 54 E-8 5 11 21 17 22 22 21 55 TOTAL 43 74 136 83 114 167 145 407 TABLE 24 RAW SCORE DATA FROM ACTIVITY CHECK LIST Experimental Class Judge No. 2 Pre-Experlmental Period Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 5 3 14 2 11 18 14 47 E-2 3 5 15 1 4 17 16 50 E-3 5 6 11 3 6 14 13 50 E-4 6 6 13 3 5 11 11 50 E-5 6 10 12 7 2 13 16 50 E-6 4 3 8 5 1 17 17 50 E-7 9 16 10 12 18 19 15 50 E-8 9 15 13 11 16 19 18 50 TOTAL 47 64 96 . 44 63 12 8 120 397 i —1 V J 1 TABLE 25 RAW SCORE DATA FROM ACTIVITY CHECK LIST Experimental Class Judge No. 3 Pre-Experimental Period Subject No. Movements Directed Toward Body Move ments In Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 6 9 15 15 16 9 36 E-2 0 7 17 5 2 16 30 E-3 3 12 18 18 18 14 3 38 E-4 14 22 28 16 21 26 9 37 E-5 6 18 26 20 21 14 15 37 E-6 8 8 8 10 9 13 10 34 E-7 1 5 8 6 13 10 12 30 E-8 3 10 7 8 12 10 7 27 TOTAL 41 91 127 98 112 109 81 269 Post-Experimental Period TABLE 26 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Experimental Class Judge No. 1 Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 9 1 19 4 11 2 1 60 E-2 7 5 4 5 1 2 0 60 E-3 6 8 10 12 0 0 1 60 E-4 3 9 8 9 17 2 0 60 E-5 3 6 9 3 4 0 1 60 E-6 3 7 11 7 0 0 4 60 E-7 2 4 5 8 2 1 0 60 E-8 8 8 20 7 12 2 0 60 TOTALS 4l 48 86 55 47 9 7 480 TABLE 27 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Experimental Class Judge No. 2 Post-Experimental Period Subject No. Movements Directed Toward Body Move ments in Chair Gross Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 2 4 12 6 6 0 0 70 E-2 1 8 8 2 3 2 0 70 E-3 6 2 9 3 1 0 2 70 E-4 8 2 5 1 7 0 0 70 E-5 2 3 14 2 1 0 2 70 E-6 3 0 3 4 2 2 1 70 E-7 5 3 8 2 1 o o 70 E-8 3 4 12 1 1 0 0 70 TOTALS 30 26 71 21 22 4 5 560 TABLE 2 8 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Experimental Class Post-Experimental Period Judge No. 3 Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods E-l 9 9 5 8 3 2 0 50 E-2 6 6 4 9 1 4 2 50 E-3 4 1 4 6 12 2 1 1 50 E-4 6 13 5 12 9 4 1 50 E-5 6 13 13 13 0 0 1 50 E-6 6 3 2 14 2 4 2 50 E-7 11 7 10 11 1 2 1 50 E-8 9 9 0 23 1 1 1 48 TOTALS 57 64 45 102 19 18 9 398 TABLE 29 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Contrast Class Judge No. 1 Pre-Experimental Period Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 4 6 9 7 12 12 12 36 C-2 10 20 22 17 25 13 19 56 C-3 2 3 2 7 4 6 11 50 C-4 9 22 15 6 7 9 50 C-5 3 5 11 4 3 9 6 35 C-6 5 5 19 15 10 12 14 55 C-7 5 9 31 9 9 15 10 50 C-8 10 8 15 9 6 11 18 50 TOTALS 43 65 131 83 75 86 118 382 TABLE 30 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Contrast Class Judge No. 2 Pre-Experimental Period Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 5 1 18 3 3 5 26 40 C-2 7 10 9 11 14 9 10 30 C-3 1 5 14 4 3 5 19 40 C-4 1 7 14 4 3 5 3 40 C-5 8 7 12 n 1 16 12 40 C-6 3 1 17 4 1 17 12 40 C-7 5 5 14 1 2 7 6 40 C-8 9 6 16 7 0 8 12 40 TOTALS 39 ^2 114 4l 27 72 100 310 TABLE 31 RAW SCORE DATA PROM ACTIVITY CHECK LISTS Contrast Class Judge No. 3 Pre-Experlmental Period Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 3 4 3 6 3 2 3 20 C-2 6 10 9 7 10 6 4 20 C-3 0 7 6 2 3 3 3 20 C-4 2 4 3 3 2 3 2 20 C-5 5 2 6 3 7 3 2 20 C-6 1 4 6 8 4 6 3 20 C-7 2 3 7 3 6 5 4 20 C-8 1 8 4 4 9 9 i 14 20 TOTALS 20 42 44 36 44 37 35 160 TABLE 32 RAW SCORE DATA PROM ACTIVITY CHECK LISTS Contrast Class Judge No. 1 Post-Experimental Period Subject No. Movements Directed Toward Body Move ments In Chair Distrac tion Gross Movements of Legs and Feet Piddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 11 7 22 6 10 2 7 40 C-2 12 19 15 22 10 4 11 40 C-3 1 3 7 3 16 6 9 40 C-4 5 9 4 14 5 2 2 40 C-5 10 5 15 5 4 2 5 40 C-6 9 14 15 16 5 2 1 40 C-7 9 10 15 6 15 6 9 40 C-8 22 24 19 7 8 17 16 40 TOTALS 79 91 112 79 73 41 60 320 TABLE 33 RAW SCORE DATA PROM ACTIVITY CHECK LISTS Contrast Class Judge No. 2 Post-Experimental Period Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling [ Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 8 5 11 5 2 5 4 30 C-2 11 11 8 10 12 8 9 30 C-3 3 7 6 7 7 7 5 30 C-4 2 4 7 6 2 0 2 30 C-5 5 9 9 5 5 4 1 30 C-6 4 8 11 6 5 2 2 30 C-7 6 8 6 5 3 5 4 30 C-8 9 12 11 7 5 5 2 30 TOTALS 48 64 69 51 41 36 29 240 Post-Experimental Period TABLE 34 RAW SCORE DATA FROM ACTIVITY CHECK LISTS Contrast Class Judge No. 3 Subject No. Movements Directed Toward Body Move ments in Chair Distrac tion Gross Movements of Legs and Feet Fiddling Communicative or Quasi-Com- municative Activity Walking or Standing Total Obser vation Periods C-l 6 5 3 6 1 4 1 20 C-2 8 10 5 13 7 7 5 20 C-3 3 7 5 4 3 2 4 20 C-4 2 6 5 5 3 3 2 20 C-5 2 2 1 3 1 6 3 20 C-6 6 3 6 8 3 3 3 20 C-7 6 5 3 8 2 2 3 20 C-8 2 4 5 11 4 4 5 20 TOTALS 35 42 33 58 24 31 26 160 APPENDIX B SOUND LEVEL METER RECORDS— RAW DATA 186 187 TABLE 35 SOUND LEVEL METER RECORDS OP EXPERIMENTAL CLASS IN PRE-EXPERIMENTAL PERIOD Raw Score Data Time Reading Time Reading Time Reading 8:30 60 8: 30 68 10:20 70 8:35 65 8: 35 65 10:25 78 8:40 67 8: 40 68 10:30 74 8:45 68 8: 45 66 10:35 68 8:50 62 8: 50 66 10: 40 76 8:55 68 8:55 68 10: 45 78 9:00 78 9:00 75 10:50 75 Recess 10:55 70 9:55 70 9: 45 78 11:00 60 10:00 80 9:50 68 11:05 80 10:05 72 9:55 68 11:10 75 10:10 66 10:00 74 11:15 78 8:15 72 10:05 80 11:20 72 8:20 78 10:10 72 8:25 72 10:15 76 Total Sound Level Readings: 4l Mean of Sound Level 72.8 Standard Deviation 5.6 188 TABLE 36 SOUND LEVEL METER RECORDS OP EXPERIMENTAL CLASS IN POST-EXPERIMENTAL PERIOD Raw Score Data Time Reading TJime Reading Time Reading 8:15 56 10:25 55 11:50 55 8:20 55 10: 30 58 11:55 56 8:25 56 10:35 53 8:05 56 8:30 57 10: 40 55 8:10 54 8:35 56 10:45 56 8:15 56 8:40 • 54 10:50 58 8:20 55 8:45 54 10:55 57 8:25 54 8:50 54 11:00 56 8: 30 56 8:55 56 11:05 57 8:35 55 9:00 62 11:10 54 11:30 55 9:45 58 11:15 55 11:35 52 9:50 62 11:20 53 11:40 54 10:00 58 11:25 55 11:45 53 10:05 60 11: 30 52 11:50 55 10:10 56 11:35 54 11:55 53 10:15 58 11:40 53 10:20 58 11:45 56 Total Number of Sound Level Readings: 49 Mean of Sound Level Readings 55.7 Standard Deviation 2.23 189 TABLE 37 SOUND LEVEL METER READINGS OF CONTROL CLASS IN PRE-EXPERIMENTAL PERIOD Raw Score Data Time Reading Time Reading Time Reading 8: 30 65 10:25 69 9:05 60 8:35 58 9:10 61 9:10 62 8: 40 60 9:15 56 9:15 64 8: 45 62 9:20 62 9:20 62 8:50 66 9:25 61 10:20 68 8:55 68 9:30 80 10:25 66 9:00 66 8:15 62 10: 30 68 9:05 68 8:20 68 10:35 66 9:10 72 8:25 65 10:40 75 9:15 67 8:30 60 10:45 78 9:20 69 8:35 60 10:50 72 9:25- 65 8:40 61 10:55 62 9:30 67 8:45 62 11:00 70 10:10 70 8:50 60 11:05 65 10:15 68 8:55 58 11:10 67 10:20 68 9:00 68 11:15 60 11:20 66 Total No. Readings 48 Mean of Sound Level 65.7 (66) Standard Deviation 5.17 TABLE 38 190 SOUND LEVEL METER RECORS OF CONTRAST CLASS IN POST-EXPERIMENTAL PERIOD Raw Score Data Time Reading Time Reading Time Reading 8:50 62 10: 40 66 8:35 60 8:55 60 10: 45 72 8: 40 70 9:00 65 10:50 71 8: 45 68 9:05 62 10:55 70 8:50 72 9: 10 64 11:00 68 8:55 73 9:15 58 11:05 66 9:00 58 9:20 62 11:10 70 9:05 64 9:25 68 11:15 72 9:10 72 10:05 70 11: 20 68 9:15 71 10:10 66 11:25 72 9:20 68 10:15 78 11: 30 71 9:25 74 10:20 74 11:35 72 10:15 80 10: 25 71 8:20 62 10:20 71 10: 30 62 8:25 64 10:25 68 10:35 67 8:30 65 10: 30 70 10:35 68 Total Number of Sound Level Readings: 46 Mean of Sound Level Readings 68 Standard Deviation 4.86 APPENDIX C TEACHER OBSERVATION FORM— RAW DATA 191 192 TABLE 39 COMPARISON OP TEACHER REINFORCEMENT OF ADAPTIVE AND MALADAPTIVE SUBJECT BEHAVIOR IN PRE- AND POST-EXPERIMENTAL PERIOD Experimental Class Judge No. 1: Pre-Exper. Score Post-Exper. Score Judge No. 2: Pre-Exper. Score Post-Exper. Score Judge No. 3: Pre-Exper. Score Post-Exper. Score Combined Scores of Three Judges: Pre-Exper. Score Post-Exper. Score Adaptive Maladaptive 24 44 68 48 0 48 72 44 Chi Square = 50.4*** 116 Adaptive Maladaptive 23 24 47 22 0 22 45 24 Chi Square = 15.05*** 69 Adaptive Maladaptive 14 20 34 12 0 12 26 20 Chi Square = 10.21** 46 Adaptive Maladaptive 61 88 149 82 0 82 143 88 231 Chi Square = 74. 45*** ** — Significant at the .01 level of confidence «*# — Significant at the .001 level of confidence TABLE 40 193 COMPARISON OP TEACHER REINFORCEMENT OP ADAPTIVE AND MALADAPTIVE SUBJECT BEHAVIOR IN PRE- AND POST-EXPERIMENTAL PERIODS Contrast Class Judge No. 1: Adaptive Maladaptive Pre-Exper. Score 51 51 102 Post-Exper. Score 24 22 46 75 73 148 Chi Square = . 0005a Judge No. 2: Adaptive Maladaptive Pre-Exper. Score 22 19 41 Post-Exper. Score 29 26 55 51 45 96 Chi Square = .00l4a Judge No. 3: Adaptive Maladaptive Pre-Exper. Score 26 22 48 Post-Exper. Score 26 19 45 52 41 93 Combined Scores of Three Judges: Chi Square = . 0000& Adaptive Maladaptive Pre-Exper. Score 99 92 191 Post-Exper. Score 79 67 146 178 159 337 a — No significant difference. TABLE 41 BETWEEN-CLASS COMPARISON OF TEACHER REINFORCEMENT OF ADAPTIVE AND MALADAPTIVE SUBJECT BEHAVIOR Between-Class Comparison of Pre-Experimental Period Scores Adaptive Maladaptive Experimental Teacher 61 88 149 Contrast Teacher 99 92 191 160 180 340 Chi Square = .037a Between-Class Comparison of Post-Experimental Period Scores: Adaptive Maladaptive Experimental Teacher 82 0 82 Contrast Teacher 79 67 146 l6l 67 228 Chi Square = 76.7*** a — No significant difference between groups. #*# — Significant at the .001 level of confidence. APPENDIX D RECORDING FORMS FOR RATING INSTRUMENTS 195 I V 1 0 1 : .... - .......... ........- - . . . . 1 — - ----------------------- : t ................. . . . . . • .* 9 fc * r I ! . ..................... ‘ 7 . . 1 . • » • f ----- 1 t * > ■; --....I 1 h . * % * r I v - ■ ..' .* . . . . . . . , . < ____ ! t • \ • ! ! • li » , v * k ' ■ \ ‘ i ) i . •> • * * * . * a ! J ft » . . . . . . . . . . . . . . i ' * ’ 7 ' . > u ’ T i ] 1 . -i . . . ------- ! . > 1 > * . 1 f 1 * * * i ---- v ■ ' ■ . - *• - i * ! . . A : . .. ----- ! T 1 . i * * \ *- i 1 ____ I . . _ _ . . . . . * t, . i > •• : • . 1 1 ' \ - * “ . - ' ■' : v V r O ... . . ....... ... ■h • • ; s ■ ^ ; ■ •- ' ‘ ' ’ ^ ; - ; U' '* - • * 'v* y t ........ ^ g ' ? . V O J . “ ' S ' ■ » v : V N - . - . **"* ■■ - ------- - -------- -- - r s ~ ........................ " ~ " s v > 2 a ‘ ............ - N ■ ‘ : i .* '• « . i ■ • ; J " ■ \ * " * • | ' x * • /> J s j 1 * .... ' 1 :■ \ i S f J v; • * • ! 1 > ‘ ; * ... »' * > G...MJ y "i 961 196 j 197 Vil o i j ! i i i ' i ! ! • ; j * ‘ ! l i l ! i I I J.'J FO i 1 i [ S 1 t 1 [ l ! : f ! ! i i j . . i . - < ‘ • I • - : . ' m » * * ■ 1 1 ! i I t t ; I r;i 198 * r ' ' ! l I ! Tj ' ! i | CC! ! ; : O i i ; I -J - i ' <L ' r*N ! i ! ■ I .f- ( ' ■ • I V.x I ‘ 198 i i f j ' ! ! | £ > . 'i } i : O i i I .I- ! ! I i I I ! ! I I ! Office of I/Os Angeles County Superintendent of Schools Division of Research and Guidance APPENDIX A HUNKS' BEHAVIOR RATING SCALE Name _________________________________________Age______________ Grade---------- Teacher____________________________School__________________ Date-- Please rate each and every statement by putting an X in the appropriate square after the statement. The squares are numbered from I to 5 and represent the degree to which you have noticed the described behavior. The bases for making a judgment are given below! (1) You have not noticed this behavior at all. (2) You have noticed the behavior to a slight degree. CO You have noticed the behavior to a considerable degree. Ml You have noticed the behavior to a large degree. (5) You have noticed the behavior to a very large degree. „ Rating Scale Vegetative-Autonomic III (2) In) (4) 15) 1. Hyperactive and restless___________________________________________________________ 2. Erratic, flighty, or scattered behavior ___________________ 3. Easily distracted, lacks continuity of effort and perseverance_____________________________________________________________________ 4. Behavior goes in cycles ___________________ 5. Quality of work may vary from day to day ___________________ 6. Daydreaming, alternating with hyperactivity ___________________ 7. Explosive and unpredictable behavior ___________________ 8. Cannot seem to control self (will speak out or jump out of seat)___________________ 9. Poor coordination in large muscle activities (games, ^etc.I__________________________ Perceptual-Discriminative 10. Confusion in spelling and writing_______________________________ ___________________ 11. Inclined to become confused in number processes; gives illogical responses______________________________________________________________ 12. Reading is poor 13. Lacks a variety of responses; repeats himself in many situations ' ____ 'xi iou nave dui nunteu mit> uenaviui m an.. (2) You have noticed the behavior to a slight degree. (3) You have noticed the behavior to a considerable degree. Ml You have noticed the behavior to a large degree. (5) You have noticed the behavior to a very large degree. Rating Scale Vegetative-Autonomic (1) (2) (Hi (4) (5) 1. Hyperactive and restless__________________________________________________________ 2. Erratic, flighty, or scattered behavior __________________ 3. Easily distracted, lacks continuity of effort and perseverance __________________ 4. Behavior goes in cycles __________________ 5. Quality of work may vary from day to day __________________ 6. Daydreaming, alternating with hyperactivity __________________ 7. Explosive and unpredictable behavior __________________ 8. Cannot seem to control self (will speak out or jump out of seatl__________________ 9. Poor coordination in large muscle activities (games, ^etc.I I I I I Perceptual-Discriminative 10. Confusion in spelling and writing __________________ 11. Inclined to become confused in number processes; gives illogical responses __________________ 12. Reading is poor __________________ 13. Lacks a variety of responses; repeats himself in many situations__________________ 14. Upset by changes in routine __________________ 15. Confused in following directions __________________ 16. Confused and apprehensive about rightness of response; indecisive______________________________________________________________________ 17. Classroom comments are often "off the track" or peculiar________ __________________ 18. Difficulty reasoning things out logically with others___________ ___ Social-Emotional 19. Demands much attention____________________________________________________________ 20. Tends to be destructive especially of the work of others________ __________________ 21. Many evidences of stubborn uncooperative behavior_______________ __________________ 22. Often withdraws quickly from group activities, prefers to work by self_________________________________________________________________ 23. Constant difficulty with other children and/or adults (apparently purposeless)______________________________________ __________________ 24. Shallow feeling for others — 25. Cries often and easily __________________ 26. Often more confused by punishment __________________ 27. Seems generally unhappy ______________ __ 28. Often tells bizarre stories 1 1 I HFB)EM *27090 MU • 9/8/58 V f 1 inis is a foilow~up survey of parent opinion of any behavior or.angcG o i ' * other symptoms that you may * • nave noticgg in you** child over the last; two months« Please answer each question •as fully as possible< > lo Have you noticed any changes in your child’s behavior as result of his participation in the s t u d y ? ______ 20 Co yon fee1 that your son has oenefitted from his school experience as a subject in this study? ____ _________ Jo Please list any unfavorable effects of the study on your child’s behavioro &o Kus he developed any rev; behavior*ehnracteristics that you. feel are undesirable? If sq 3 please list. .a your child developed any nsw behavior characteristics .at fe-il are desirable? If so# please list. Please comment,, if you have not already done so, cn th . < > * ~ _ „ „ „ a „ Activity level ______ 200 Vo Has ho dovelopod any new bchavior# charscfceristies that you feel crs undesirable? If so* please lis^ >Uo *.;\c 0:.U Oj l o . W V V U ,r s » lOli child developed any new * ' are desirable? osaavj.or charsciutiua«.-a*»o ii * ^ , T , 0 , f t pisase j. a. owo So Please eo^raont,, if you have not already &onQ so, on tnj iul-U . V - - - i j - - , * a, Acoj.1'avij- .lSYO^ __ i i , bo Relationships with brothers and/or sisters. Attitude: toward sehoo.1 ,u. - do Attitude toward parents Any further information that you can supply about your child5s behavior over the last two inonths will bo greatly appreciated0 2on APPENDIX E LETTERS FROM REDONDO BEACH SCHOOL DISTRICT PERSONNEL 201 SOUTH BAY AREA SCHOOL DISTRICTS COOPERATIVE SPECIAL EDUCATION PROGRAM ADMINISTRATION OFFICE 115 SOUTH FRANCISCA AVENUE REDONDO BEACH. CALIFORNIA FRontier 9-5449 April 22, 1966 Dr. C. E. Meyers Department of Educational Psychology University of Southern California University Park Los Angeles, California Dear Dr. Meyers: At the time of your visit to the educationally handicapped class at Franklin School where Mr. Hotchkiss was doing his experiment, you suggested that a-letter from us regarding our impressions might be appreciated. During the present year I have not given close personal attention to the rather large number of educationally handicapped classes located in Palos Verdes, Redondo Beach, Hermosa Beach and Manhattan Beach School Districts. The Redondo Beach classes have been administered at the local district level. Harry McKee, one of the psychologists for the Redondo Beach City School District, has functioned as a consultant to the program in Redondo Beach. Mr. Hotchkiss carried on his daily work with this class from February 28 to April I?, with one week off from April ii-8 for Easter vacation. Yesterday I conferred with Mr. McKee and William Clemenger, the Franklin School principal, regarding the situation in that class now that Hotchkiss has been absent for more than a week. It must be granted, of course, that impressions of this kind are rather subjective. However, they both report that it is hard to believe that this is the same class which was being taught there before Mr. Hotchkiss cam e im n n t h e s c e n e . T a s lre ri M c K e e u h e t h e r i t . w a s b e c a u s e t h e t e a c h e n w a s 202 jjx • \J • u • ncjr cx o Department of Educational Psychology University of Southern California University Park Los Angeles, California Dear Dr. Meyers: At the time of your visit to the educationally handicapped class at Franklin School where Mr. Hotchkiss was doing his experiment, you suggested that a-letter from us regarding our impressions might be appreciated. During the present year I have not given close personal attention to the rather large number of educationally handicapped classes located in Palos Verdes, Redondo Beach, Hermosa Beach and Manhattan Beach School Districts. The Redondo Beach classes have been administered at the local district level. Harry McKee, one of the psychologists for the Redondo Beach City School District, has functioned as a consultant to the program in Redondo Beach. Mr. Hotchkiss carried on his daily work with this class from February 28 to April 15, with one week off from April U-8 for Easter vacation. Yesterday I conferred with Mr. McKee and William Clemenger, the Franklin School principal, regarding the situation in that class now that Hotchkiss has been absent for more than a week. It must be granted, of course, that impressions of this kind are rather subjective. However, they both report that it is hard to believe that this is the same class which was being taught there before Mr. Hotchkiss came upon the scene. I asked McKee whether it was because the teacher was following through with the techniques employed by Hotchkiss or whether it was a carry-over as a result of Hotchkiss' work. His opinion was that it is largely a carry-over. Mr. Clemenger was of much the same opinion except he felt that perhaps the teacher had modified her practices also as a result of the experi ment. In any case, the class is now quieter and more orderly than in a neighbor ing EH class which earlier in the year was the most orderly of these classes. Any of our personnel who are connected with this class will be only too glad to express their impressions or provide any other information regarding 202 SOUTH BAY AREA SCHOOL DISTRICTS COOPERATIVE SPECIAL EDUCATION PROGRAM ADMINISTRATION OFFICE 115 SOUTH FRANCISCA AVENUE REDONDO BEACH, CALIFORNIA FRontier 9-5449 Dr. C. E. Meyers - 2- April 22, 1966 the progress of this class if we are requested to do so I hope this is of some value. Do not hesitate to let me know if we can be of any further help. Very truly yours, Coordinator GBLjbr rv> o oo R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T FRANKLIN S C H O O L 850 INGLEWOOD AVENUE REDONDO BEACH, CALIFORNIA June 3, 1966 Dr. C. E. Meyers University of Southern California University Park Los Angeles, California Dear Dr. Meyers: Mr. Janes Hotchkiss conducted his Operant Conditioning Program in our school between March 7 and 13. Impartial observers spent three days before and after the experiment making statistical checks on class behavior. Our school is a Kindergarten through sixth grade school, two grades for each level with three Educationally Handicapped classes. Our total enrollment is U58. The enrollment for the E. H. class in March was twenty-six. E.H. children are bussed to us from all over Redondo Beach. These classes meet between 8:00 a.m. and 12:00 p.m. The class chosen for Mr. Hotchkiss' experiment was the older group of boys, ages seven to ten years. This was an exceptionally hyperactive group, four boys were already processed for home teaching when Mr. Hotchkiss came to us. These four were so hyperactive and unmanageable that we almost eliminated them from the program. However, with Mr. Hotchkiss' program these boys were successfully contained and made signi ficant academic progress; their behavior has been markedly improved so that we have not had any severe outburst since the program. One of the most severe cases of the four has been attending a regular sixth grade classroom for a major portion of the day and has made tremendous strides in reading, art, and social studies. Three new E. H. students came into this class on April 18, immediately M 4*4. — - - - - - - - -? - - - - - - - A - - - - - - - - • .... 204 l / V O A i/1 • I'icjrci O i Mr. Janes Hotchkiss conducted his Operant Conditioning Progran in our school between March 7 and 13. Impartial observers spent three days before and after the experiment making statistical checks on class behavior. Our school is a Kindergarten through sixth grade school, two grades for each level with three Educationally Handicapped classes. Our total enrollment is 458. The enrollment for the E. H. class in March was twenty-six. E.H. children are bussed to us from all over Redondo Beach. These classes meet between 8:00 a.m. and 12:00 p.m. The class chosen for Mr. Hotchkiss’ experiment was the older group of boys, ages seven to ten years. This was an exceptionally hyperactive group, four boys were already processed for home teaching when Mr. Hotchkiss came to us. These four were so hyperactive and unmanageable that we almost eliminated them from the program. However, with Mr. Hotchkiss' program these boys were successfully contained and made signi ficant academic progress; their behavior has been markedly improved so that we have not had any severe outburst since the program. One of the most severe cases of the four has been attending a regular sixth grade classroom for a major portion of the day and has made tremendous strides in reading, art, and social studies. Three new E. H. students came into this class on April 18, immediately after the experiment. They were integrated into the class without incident. We usually bring one or two new students into the class at a time to facilitate integration. Improvement was so great that we felt we could enroll these students without difficulty. We did. Mr. Hotchkiss was here day to day, available all day for extra help and to stay with the children a little longer if necessary. He carried on several parent interviews, and sometimes delivered children to their hones late in the afternoon. We never had the same problem with the same child twice. This personal relationship working day by day with these children was a great help and no doubt a significant factor in the success of the experiment. 204 It has been our custom to have monthly meetings of the parents of these three classes of E.H. children. Mr. Hotchkiss spoke before and after the experiment to these parents. In the beginning, they expressed concern about: ••bribery", "will it work, will it carry over at home?" At the final meeting, the specific question of "bribery" was asked and all the parents said it had not come up or been mentioned during the entire month. All parents who had children in the program remarked on the dramatic changes in behavior and attitude at home. There were some noticeable parent changes too?'.' Mrs. Scott, the teacher, learned a great deal in this process. She is now better able to acknowledge, to become aware of and reward good behavior, and less likely to give reinforcement to negative behavior. We are delighted that we were chosen for this experiment and most appreciative of Mr. Hotchkiss' work. Sincerely, James W. Clemenger ■-/ Principal / ro o v _ n JWCsbj R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T FR AN KL IN S C H O O L 850 INGLEWOOD AVENUE REDONDO BEACH. CALIFORNIA June 6 , 1966 Dr. C. E. Meyers Department of Educational Psychology University of Southern California University Park Los Angeles, California Dear Sir: I observed the class Mr. Hotchkiss worked with, before and during his experiment. I was quite surprised at the marked change in the behavior of the boys in Mrs. Scott's class. I was in the classroom one day in October for a half hour. The boys were very difficult to control. The teacher had to spend most of her time and energy keeping the boys quiet and in their seats. In later weeks I observed them in the doorway and hallway when they were suppose to be in their seats working. In March, I was able to spend another half hour in the classroom. The boys' activity was quite different. The boys were all sitting in their seats. I did not see one of them speak without first raising his hand. If one of the boys was not doing his work, the boy attending the timer would restart it on his own initiative. One boy in particular was reluctant to do his work, but he finally gave in the pressure of the group. This experiment in operant conditioning was quite successful for this group of boys. Sincerely yours, Los Angeles, California Dear Sir: I observed the class Mr. Hotchkiss worked with, before and during his experiment. I was quite surprised at the marked change in the behavior of the boys in Mrs. Scott's class. I was in the classroom one day in October for a half hour. The boys were very difficult to control. The teacher had to spend most of her time and energy keeping the boys quiet and in their seats. In later weeks I observed them in the doorway and hallway when they were suppose to be in their seats working. In March, I was able to spend another half hour in the classroom. The boys' activity was quite different. The boys were all sitting in their seats. I did not see one of them speak without first raising his hand. If one of the boys was not doing his work, the boy attending the timer would restart it on his own initiative. One boy in particular was reluctant to do his work, but he finally gave in the pressure of the group. This experiment in operant conditioning was quite successful for this group of boys. Sincerely yours, '7,a , )£*->■ Mrs. Bruce Johnsonbaugh, School Nurse 206 DU 010 BEACH c m SCHOOL DISTRICT 115 SOUTH FRANCISCA AVENUE REDONDO BEACH. CALIFORNIA ADMINISTRATIVE OFFICES FRontier 9-5 4 4 ‘> June 9, 1966 Dr. C. E. Meyers University of Southern California University Park Los Angeles, California Dear Dr. Meyers: I had the opportunity of visiting the EH special class at Franklin School on several occasions during the Operant Conditioning Experiment under the direction of Mr. Hotchkiss. My reaction was one of favorable amazement! The class was one of the most hyperactive ones in the district. Several students were being considered for hone teaching due to their uncontrollable behavior. I was intimately acquainted with three other pupils known to be excessively distractable, hyperkinetic and explosive in their behavior. Yet under the apparent effects of the experimental treatment, students sat at their desks for reasonable periods of tine, raised their hands when asking a question, and eliminated extraneous and disturbing movements. They seemed involved in the experiment, inter ested in the rewards, and responsive to peer pressures to keep quiet or behave appropriately. The teacher was thereby free to teach and respond to individual needs and the students were able to accomplish a reasonable amount of work. The group was so responsive that even students were able to ’ •lead" the class in terms of behavior control. The climate for learning was transformed from minimal to adequate, and the teacher seemed to assimilate some of the pTnprinon+ie v^t-i — University of Southern California University Park Los Angeles, California Dear Dr. Meyers: I had the opportunity of visiting the EH special class at Franklin School on several occasions during the Operant Conditioning Experiment under the direction of Mr. Hotchkiss. Ify reaction was one of favorable amazement! The class was one of the most hyperactive ones in the district. Several students were being considered for home teaching due to their uncontrollable behavior. I was intimately acquainted with three other pupils known to be excessively distractable, hyperkinetic and explosive in their behavior. Yet under the apparent effects of the experimental treatment, students sat at their desks for reasonable periods of time, raised their hands when asking a question, and eliminated extraneous and disturbing movements. They seemed involved in the experiment, inter ested in the rewards, and responsive to peer pressures to keep quiet or behave appropriately. The teacher was thereby free to teach and respond to individual needs and the students were able to accomplish a reasonable amount of work. The group was so responsive that even students were able to "lead" the class in terms of behavior control. The climate for learning was transformed from minimal to adequate, and the teacher seemed to assimilate some of the experiment's behavior shap ing principles as part of her repertoire of discipline techniques. The results of this experimental approach to behavior shaping with hyperkinetic and distractable youngsters were very encouraging. Sincerely, Bernard Feldman District School Psychologist BF:em cc: James Hotchkiss 207 REDONDO BEACH CITY S IJ u STRICT 115 SOUTH FRANCISCA AVENUE REDONDO BEACH. CALIFORNIA ADMINISTRATIVE OFFICES FRontier 9-544' June 7, 1966 Dr. C. E. Meyers University of Southern California University Park Los Angeles, California Dear Dr. Meyers: James Hotchkiss conducted an experiment in "operant con ditioning" in our District from March 3 through April 12, 1966. In accordance with his request, we gave him the most hyperactive group of EH children in one of our elementary schools. Within two weeks the children were quiet, controlled and worked in a concentrated sustained way on the lessons given them. The change in behavior was truly dramatic. We were extremely pleased and encouraged to discover that after Mr. Hotchkiss left,the behavior carried over and, in fact, has persisted to this date. The parents of these children, too, reported a change in behavior and are now at a point where they want to learn more about the techniques involved. Speaking for the Guidance Department of the Itedondo Beach City School District, vie thank James Hotchkiss for a fine 208 u ± , \j, tjm i’ c y t : r s University of Southern California University Park Los Angeles, California Dear Dr. Meyers: James Hotchkiss conducted an experiment in "operant con ditioning" in our District from March 3 through April 12, 1966. In accordance with his request, we gave him the most hyperactive group of EH children in one of our elementary schools. Within two weeks the children were quiet, controlled and worked in a concentrated sustained way on the lessons given them. The change in behavior was truly dramatic. We were extremely pleased and encouraged to discover that after Mr. Hotchkiss left,the behavior carried over and, in fact, has persisted to this date. The parents of these children, too, reported a change in behavior and are now at a point where they want to learn more about the techniques involved. Speaking for the Guidance Department of the Hedondo Beach City School District, we thank James Hotchkiss for a fine bit of work. Sincerely, Harry G. McKee District Psychologist EH Consultant HGMiem cc: James Hotchkiss 208 R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T F R A N K L IN S C H O O L 850 INGLEWOOD AVENUE REDONDO BEACH, CALIFORNIA June 10, 1966 Mr, James M. Hotchkiss 461 West 3Sth Street Los Angeles, California. Dear Jim: This letter is in response to your request for a brief individual evaluation of each boy's academic progress during and after the experiment and of the behavior of the class as a whole since the end of the experiment. I have identified each boy by the number code as you requested so that you may use the letter in your dissertation if you wish. Bov E-l. He was slated for removal from this self-con tained!?. H. class at the beginning of this study. His frequent outbursts, fighting and screaming plus his general unwillingness to work made it impossible for him to remain in class. He was very hyperactive and difficult to control. He would run out of class quite often, climb the fence and g juhp onto the roof. His reading was good. He could vo read about on the third grade level at least. His arithmetic was low second grade level and his writing was illegible and scrawled. His orientation was good. He knew what year it was, what the season was and in general was a bright boy. At the end of the study, this boy was much cAlmer and was able to work far better. He still fought occassionally with his peers but it was and is not as vicious as before. ft « Dear Jim: This letter is in response to your request for a brief individual evaluation of each boyTs academic progress during and after the experiment and of the behavior of the class as a whole since the end of the experiment. I have identified each boy by the number code as you requested so that you may use the letter in your dissertation if you wish. Boy E-l. He was slated for removal from this self-con- tained E.H. class at the beginning of this study. His frequent outbursts, fighting and screaming plus his general unwillingness to work made it impossible for him to remain in class. He was very hyperactive and difficult to control. He would run out of class quite often, climb the fence and juinp onto the roof. His reading was good. He could read about on the third grade level at least. His arithmetic was low second grade level and his writing was illegible and scrawled. His orientation was good. He knew what year it was, what the season was and in general was a bright boy. At the end of the study, this boy was much calmer and was able to work far better. He still fought occassionally with his peers but it was and is not as vicious as before. He has made several friends in the class and the other students seem to accept him more graciously. He is now reading a fourth and fifth grade reader. His arithmetic is still poor. He is in need of much remedial work but he is now willing to try. He no longer is in any danger of being put on a home teaching basis. His step-father was so pleased with his progress during the study that he attended the first parent group meeting he had ever been to to tell me that he no longer was going to sue the school district for his boy’s condition. Bov E-2. At the start of the operant conditioning study, this boy was a non-reader. He continually used a first grade 209 R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T FRANKLIN S C H O O L Page 2 850 INGLEWOOD AVENUE REDONDO BEACH. CALIFORNIA math work-book over and over again. Any attempt to go on to the next task was met with stiff opposition by him. He could not write and barely knew his alphabet. He spent much of his time on the bench outside the room beaause of profanity and violent emotional outbursts. He was constant ly moving and fidgeting about and his parents were unable to control him. There were many fights at home and at school. At the end of the study, he was willing to learn. He would take instructions in math and is now working on addition and subtraction of one and two place numerals• He is diligently attempting to read in the primer "The Little White House." He is not always succeeding but he certainly is trying. He tries to please me by controlling his profanity and when he slips will clap his hands over his mouth and say, wITm sorry." His parents report better behavior at home and say that when he is riding in the car with them, for the first time he will try to read signs. They also say that they are able to play games with him and that his general demeanor is more relaxed. Bov E-3. At the onset of the study, this was the youngest boy in the class. He was about one vear under grade level in reading (primary or primmer level), and about grade level in arithmetic. He was still printing and had many reversal with bfs and d’s. He resisted instruction in cursive writing. He seldom could sit and work for more than a few minutes at a time. While he did not seem to start fights, he was frequently involved in them. He was not quite as bad a behavior problem as the others. He Was VOT~V anf.I_ai*)lnn1 snH v a . f l 41. f 44- U. At the end of the study, he was willing to learn. He would take instructions in math and is now working on addition and subtraction of one and two place numerals. He is diligently attempting to read in the primer "The Little White House," He is not always succeeding but he certainly is trying. He tries to please me by controlling his profanity and when he slips will clap his hands over his mouth and say, "ITm sorry." His parents report better behavior at home and say that when he is riding in the car with them, for the first time he will try to read signs. They also say that they are able to play games with him and that his general demeanor is more relaxed. Boy E-3. At the onset of the study, this was the youngest boy in the class. He was about one year under grade level in reading (primary or primmer level), and about grade level in arithmetic. He was still printing and had many reversal with bfs and d’s. He resisted instruction in cursive writing. He seldom could sit and work for more than a few minutes at a time. While he did not seem to start fights, he was frequently involved in them. He was not quite as bad a behavior problem as the others. He was very anti-school and very^hostile towards it. He presented a severe learning problem and was also a very great concert to his parents. At the end of the study he was reading first grade readers. He developed a good clear and firm handwriting and was able to write the stories from his reader in cursive writing. He made many friends from within the three E.H. classes. His parents are very proud of his achievements and have begun to use point system at home to reward him when he behaves well. Bov E-A. At the onset of the study this boy was a non worker, non-trouble maker and practically a non-entity. He was affectionate toward the teacher and got along fairly well with his peers. He would usually just sit f r R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T FR AN KL IN S C H O O L 850 INGLEWOOD AVENUE REDONDO BEACH, CALIFORNIA Page 3 at his desk, often fiddling and fidgeting with his work but seldom attempting to do anything. When he did do work, he was reading at a pre-primer level and still having difficulty. He was barely adding one place numerals. He would not write cursively and his printing was illegible and large. His knowledge of the world about him and his concept of time and space was limited. At the end of the study, he had begun to write letters. He began to subtract as well as try multiplication! He was proud of the fact that he was learning to read in a higher primer though still wiih great effort. He had progressed from working less than a half-hour a day on school-work to working most of the morning with the rest of the class. His father who Visited him during this study has showed a new interest in him and just a few days ago the boy told the teacher that his father was taking him to live with him this summer and that he would not be back next fall. Bov E-5. He was reading on the primer level when the study began. His arithmetic was on the second grade level. His knowledge of his surroundings was good. He was not writing or printing and fought with his peers constantly. He spent much time planning revenge for real or imagined wrongs done him by his brothers, parents and peers. He appeared quite intelligent but so hyperactive that little could be done to teach him anything. He was sent home quite regularly. When the study was completed, he had saved the most money and had $1.64 in the bank. Being rewarded quickly with money and praise appealed to him. He is now reading on a high second grade level and working in the 3rd GCMP ^ J j - J-J -----*--- - • “ 211 Hu one en a ox me suuay, ne naa Degun to write letters. He began to subtract as well as try multiplication! He was proud of the fact that he was learning to read in a higher primer though still with great effort. He had progressed from working less than a half-hour a day on school-work to working most of the morning with the rest of the class. His father who fisited him during this study has showed a new interest in him and just a few days ago the boy told the teacher that his father was taking him to live with him this summer and that he would not be back next fall. Boy E-5. He was reading on the primer level when the study began. His arithmetic was on the second grade level. His knowledge of his surroundings was good. He was not writing or printing and fought with his peers constantly. He spent much time planning revenge for real or imagined wrongs done him by his brothers, parents and peers. He appeared quite intelligent but so hyperactive that little could be done to teach him anything. He was sent home quite regularly. When the study was completed, he had saved the most money and had $1.64 in the bank. Being rewarded quickly with money and praise appealed to him. He is now reading on a high second grade level and working in the 3rd GCMP text and is doing cursive writing. Classroom control is much better and the child has not been sent home one time since the study ended! Bov E-6. When the study began, he was reading on the early first grade level. His father taught him at home both in arithmetic and in reading. There was much pressure for improvement in these areas. In addition to this he had severe neurological problems verified by a medical specialist and was given medication to help control his hyperactivity. Perseveration was shown by his making innumerable moustaches and stapling goatees onto them. The chains of paper he made for his neighborhood friends were many feet long. If any attempt was made to stop these activities, a screaming fit would start and he would have to R E D O N D O B E A C H C I T Y S C H O O L F R A N K L I N S C H O O L 850 INGLEW OOD AVENUE REDONDO BEACH, CALIFORNIA After the study, he was reading on the high second grade level and his arithmetic text was the 3rd grade level GCMP. His writing had improved from a shaky and wavering scrawl to a fairly legible and relatively well formed cursive style. His parents also have expressed their approval of his progress and say that he ia also much better behaved at home. He has become easier to control in class, and no longer perseverates. For the first time he has joined a Little League ball team and has been successful. Bov E-7. Prior to the onset of the operant conditioning study, this boy was reading at the primer level, could add and subtract one place numerals and was just beginning to write his letters in cursive. He was keenly interested in his surroundings but his awareness of time, distance and space wascvery limited. He was hyperactive, his attention span was short and he was unable to remain seated for longer than a very few minutes at a time. He had no friends and fought constantly. He did not tolerate any physical contact from anyone. Noises upset him to the point where it became impossible for him to function and he would leave the room and/or the school grounds. On many occassions, he climbed onto the roof of the school buildings, hid in the cupboards or trash caas from the teacher or principal. It was fre quently impossible to control him and he was then often sent home. At first his parents were cooperative but as prosures mounted at schod so did pressures at home. There were fights with the three sisters, one ending in an attempt to u n i u i . i j j i - -i - * • * * ^ DISTRICT Page U 212 — — — w .a w **w T C/1 O v v O t A U 1 V A 1 C first time he has joined a Little League ball team and has been successful. Boy E-7. Prior to the onset of the operant conditioning study, this boy was reading at the primer level, could add and subtract one place numerals and was just beginning to write his letters in cursive. He was keenly interested in his surroundings but his awareness of time, distance and space wascvery limited. He was hyperactive, his attention span was short and he was unable to remain seated for longer than a very few minutes at a time. He had no friends and fought constantly. He did not tolerate any physical contact from anyone. Noises upset him to the point where it became impossible for him to function and he would leave the room and/or the school grounds. On many occassions, he climbed onto the roof of the school buildings, hid in the cupboards or trash cams from the teacher or principal. It was fre quently impossible to control him and he was then often sent home. At first his parents were cooperative but as prasures mounted at schod so did pressures at home. There were fights with the three sisters, one ending in an attempt to kill his middle sister. The parents reported that family relationships were becomming impossible. He v/as eager to learn in the three Hfs. Parents stressed this part of his development to him. He was also concerned because his younger sister in the first grade was beginning to catch up with him. Therefore teacher control was effective at times. It was short lived however, because of his inability to sit in his seat. After the operant conditioning study was completed, he was reading on the high second grade level. He was multiDlying and dividing in arithmetic and solving problems involving reading. He jumpred from the Greater Cleveland Mathematics Program book 1 to Book 3. He took papers home R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T F R A N K L I N S C H O O L 850 INGLEW OOD AVENUE REDONDO BEACH. CALIFORNIA P a g e 5 voluntarily for extra work. His most notable success was a combination of improvement in reading and in being integrated into a regular sixth grade classroom. This opened up new vistas of learning for him that he had never experienced before. The children in the regular classroom became his friends and tutors. He now takes part in their social studies program, Spanish lessons and art studies. He now seems much more aware of his surroundings. At the time of this letter he is reading in fourth and fifth grade remedial readers and has an actual reading level of about early fourth grade. He is enthusiastic and can and does set for long periods while he works. He has not been sent home since the end of the study. His parents were pleased and let him, and his teacher, know it. With this kind of praise the boy’s adjustment has been easier. He no longer ever sits in the broom closets, makes animal noises or rung out of the room. Bov E-8. This boy was a non-reading child when the study began. His arithmetic was early first grade level. He was doing manuscript printing and his knowledge of social studies and his general surroundings was very ^ limited. His concept of time and space was immature. He m was very hyperactive and was on medication. ^ At the conclusion of the study, he was reading the first reader in the second grade series— about one year’s progress in one monthl His arithmetic text was the CGMP Book 3. He had been in book 1. He was able to work on program, Spanish lessons and art studies. He now seems much more aware of his surroundings. At the time of this letter he is reading in fourth and fifth grade remedial readers and has an actual reading level of about early fourth grade. He is enthusiastic and can and does set for long periods while he works. He has not been sent home since the end of the study. His parents were pleased and let him, and his teacher, know it. With this kind of praise the boy’s adjustment has been easier. He no longer ever sits in the broom closets, makes animal noises or rund out of the room. Boy E-&. This boy was a non-reading child when the study began. His arithmetic was early first grade level. He was doing manuscript printing and his knowledge of social studies and his general surroundings was very limited. His concept of time and space was immature. He was very hyperactive and was on medication. At the conclusion of the study, he was reading the first reader in the second grade series— about one year’s progress in one month! His arithmetic text was the CGMP Book 3« He had been in book 1. He was able to work on maps of the USA and to participate in discussions of local geography intelligently. He began to write cursively and was able to write in cursive from printed material. Most importantly, he now regularly requests homework which he always brings back completed. The hyperactivity has diminished and there are far fewer arguments with peers. Classroom behavior is much improved and he has not once been sent home for misbehavior 3ince the onset of the study. Shortly after the end of the study a point system was started in the classroom. Children were given up to five points a day for good work and behavior. Small presents were displayed which could be purchased with earned poirte. This system was also started in the other two classrooms and is in use in all three rooms at the R E D O N D O B E A C H C I T Y S C H O O L D I S T R I C T F R A N K L I N S C H O O L 850 INGLEW OOD AVENUE REDONDO BEACH. CALIFORNIA P a g e 6 present time. The boys are very concerned about earning their points and all keep a careful record of what they have won. rn The study has been completed approximately two months. There has been a steady and continued improvement for the class as a whole. Individually each participant of the study has shown growth in one or more areas. Shortly after the study was completed, three new boys were brought into the class and integrated into the daily procedure without incident. The point system was explained to them and they were on their way. There have been no special problems with them, yet they were put out of tTneir regular class because of their didruptive arnd hyperactive behavior. They had severe learning problems but have never exhibited behavior problems. This week the class was asked to assist the teacher in a thorough cleaning of the room. All pitched in worked hard all morning. This would simply not have been possible before. The boys now come in and sit down quietly after putting away their lunches and sweaters. They begin work from their folders or listen to a lecture by the teacher. There are daily class discussions and group activities which also were not possible before. Minni Warnke Scott 214 APPENDIX P INSTRUCTIONS TO EXPERIMENTAL CLASS 215 Instructions Hello. I am from the University of Southern Calif ornia and I am going to be spending every day for the next month with your room and the other two rooms. I will be playing a game with one of the rooms and I would like it to be with your room if you want me to. Before you decide, let me explain how the game will be played. I will set up here in front of the room at this table. This clock, bottle, poker-chips and this bag of candy will be on the table. The poker-chips are worth one penny each and you can win one for the class each time that you all are quiet for a certain time. This clock will show you how long you must be quiet to win a chip. Now watch: I will set it on the 20-second mark and start it (quiet in class while clock runs to 0). If we were playing, and the hand had got all the way to zero without the clock stopping, you would have won one chip, or, one penny. But, if anyone had done anything he wasn’t supposed to do, like getting up without permission, talking, making noise, or bothering another boy, the clock would be stopped and set back on the 20-second mark again and I would not start it until the room was again quiet. Each time you win a chip, I will put it into this jar and then at 11:30 we will stop the game and count the chips. If you have 20 or more chips we will divide them up evenly among each of the people in the room. You may then take your chips and buy some of this candy here on the table, or trade them to me for one penny for each chip and do whatever you want with the money. If you have less than 20 in the bottle, you will not win anything and you lose the chips. No boy can win chips alone. Everyone has to play or nobody can win. Your teacher will not make you play the game or say anything to you about making noise, getting up, or anything else. Whether you play or not is entirely up to you: she has nothing to do with it. Now, if at any time you don't want to play I will stop right then and go to one of the other rooms since I don't want you to play if you don't want to. Let's vote to see if you want to play or not. Everyone who wants the game for this room hold up your hand. 216
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Effects Of Success And Failure On Impulsivity And Distractibility Of Three Types Of Educationally Handicapped Children
PDF
The Relative Efficiency Of Prompting And Confirmation Learning Paradigms
PDF
The Effects Of Two Types Of Experimenter Intervention And Schedules Of Reinforcement On Verbal Operant Conditioning Of Affective Self-References
PDF
Prescriptive Teaching As A Supplement To Behavior Modification In The Remediation Of Learning Disorders
PDF
Confirmation Of Expectancy And Changes In Teachers' Evaluations Of Student Behaviors
PDF
Auditory Perceptual Si Factors As Non-Predictors Of Reading Achievement In An Upper-Class And Upper-Middle-Class Population
PDF
Maternal Child-Rearing Attitudes And Developmental Growth Of Rubella Deaf-Blind Children
PDF
The Conditioning Of A Discriminative Stimulus Measured As An Orienting Reaction In Profoundly Retarded Blind Children
PDF
Testing Attitudes Toward Physical Handicap: Stereotyping And Objectivityin A Partially Integrated Elementary Program
PDF
A Historical Perspective Of Special Education In California. (Volumes I And Ii)
PDF
Acoustic And Associative Variables In The Retention Of Words By Children With Learning Disabilities
PDF
Selected Characteristics Of A Children'S Individual Test Of Creativity
PDF
Effects Of Preschool Enrollment And Parent Participation On Academic Growth
PDF
Classroom Climate And Pupil Characteristics In Special Classes For The Educationally Handicapped
PDF
Effects Of Associative And Category Cuing On Recall Of Items From Exhaustive And Nonexhaustive Lists
PDF
An Experimental Study Of The Degree Of Affective Response Elicited By Several Mediated And Nonmediated Instructional Methods
PDF
Immediate Memory And Its Correlates With School Achievement
PDF
School Behavior And Attitudes Of Retarded And Average High School Boys Asconcomitants Of Intelligence And Socioeconomic Status
PDF
Academic And Non-Academic Achievement Of Freshmen And Seniors At Azusa Pacific College
PDF
The effect of noise on intellectual performance as related to personality and social factors in upper division high school students
Asset Metadata
Creator
Hotchkiss, James Merel
(author)
Core Title
The Modification Of Maladaptive Behavior Of A Class Of Educationally Handicapped Children By Operant Conditioning Techniques
Degree
Doctor of Philosophy
Degree Program
Educational Psychology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
education, educational psychology,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Meyers, Charles Edward (
committee chair
), McIntrye, Robert B. (
committee member
), Slucki, Henry (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c18-123689
Unique identifier
UC11359889
Identifier
6706500.pdf (filename),usctheses-c18-123689 (legacy record id)
Legacy Identifier
6706500.pdf
Dmrecord
123689
Document Type
Dissertation
Rights
Hotchkiss, James Merel
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
education, educational psychology