THE DIFFERENTIAL EFFECTIVENESS OF TWO METHODS OF TRAINING INSTITUTIONAL ATTEN'DANT‘S IN THE TECHNIQUE OF BEHAVIOR MODIFICATION Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY HUGH JOHN McBRIDE 1972 w-..- I III III IIIIIIIIII I I III I II III I w L I B R A R Y E Michigan State §I ” University 5 This is to certify that the thesis entitled . THE DIFFERENTIAL EFFECTIVENESS OF TWO METHODS OF TRAINING INSTITUTIONAL ATTENDANTS IN THE TECHNIQUE OF BEHAVIOR MODIFICATION presented by Hugh J. McBride has been accepted towards fulfillment of the requirements for Ph . D . degree in Education llqkmrprolessor Date May 12, 1972 0-7639 F_..k__. y ‘— BINDING BY w I IIIIAB & SONS' « “MEIER! 1E. ABSTRACT THE DIFFERENTIAL EFFECTIVENESS OF TWO METHODS OF TRAINING INSTITUTIONAL ATTENDANTS IN THE TECHNIQUE OF BEHAVIOR MODIFICATION BY Hugh John McBride It is of value for administrators responsible for training institutional attendants in the technique of be- havior modification to know which training method yields the greatest amount of competence. Various methods for training attendants and evalu- ating that training have been developed. Among these are simulation experiences such as role playing or video tape presentations of critical ward situations. Coupled with these, paper and pencil tests have been used for evalu— ation. Most frequently used is the traditional method: lectures, demonstrations, and off the ward practice sessions. This study utilized a method involving the train- ing of the attendant on the ward during his regular assign- ment with children and evaluation of the outcomes of train-- ing in the same setting. The trainers employed behavior Hugh John McBride modification to shape the attendants' behavior in learning the skills. Involved was a comparison of the on-the-ward training (Treatment I) and the traditional lecture method (Treatment II). Forty institutional attendants were assigned by a quasi-random procedure to the two treatment groups, making two groups of twenty. Each treatment used language develop- ment as a training vehicle. Evaluation on a pre-test, post-test basis was done by recording the incidence of particular behaviors related to behavior modification during representative periods of the work day while the attendant worked on the ward. Analysis was done by multivariate analysis of variance on gain scores. It was the general hypothesis that pre-test- post-test gain scores for attendants trained in the on- the-ward method (Treatment I) would be equal to or greater than the gains made by those trained in the traditional method (Treatment II). This hypothesis was supported since Treatment I did no worse than Treatment II. No significant main effect for treatment was found although the means were generally higher for both groups with the trend in favor of Treatment I. The failure of either group to show much improvement weakens any conclusions concerning the relative merits of either method. Hugh John McBride It was concluded that the rating done by the experimenter was sufficiently sensitive in picking up behavior since rater reliability was found to be very high. Experimenter bias was ruled out since there was little significant improvement in either group. The rater's becoming more stringent in his success criteria was ruled out since it can be assumed that the experi- menter, being eager to secure positive results would be more likely to become more lenient in his judgments. Since other researchers report that attendants are capable of learning these techniques it is of interest that the results of this study were not more favorable. It is suggested that these results are related to the fact that this study did not attempt to measure learning alone but also competence in and tendency to use the techniques during the attendants' daily interactions with children on the ward. The following are hypothesized as the factors acting in restraint of the attendants' use of these techniques: 1. The attendants' traditional role as a custodian rather than as a habilitative agent is reinforced by the administration through the emphasis it - places on custodial-housekeeping activities as Opposed to learning activity. Hugh John McBride The need of the institutional staff to maintain control over their changes promotes a continuance of the institutionalization of both residents and staff. The role set of the attendants is one which excludes teaching activities: traditionally attendants are not seen as teachers or paid teachers' salaries. The attendant has a limited opportunity to display his skills because of conflicting demands on his time. THE DIFFERENTIAL EFFECTIVENESS OF TWO METHODS OF TRAINING INSTITUTIONAL ATTENDANTS IN THE TECHNIQUE OF BEHAVIOR MODIFICATION BY Hugh John McBride A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education 1972 CDCopyright by Hugh John McBride 1972 DEDICATION To those who have the courage to change the things they can ii ACKNOWLEDGMENTS The research reported herein reflects a coopera— tive effort on the part of many individuals who contri- buted time and effort towards its completion. Special appreciation is extended to Dr. James E. Keller who directed the study from its conception to com- pletion. Dr. Keller's continual kindnesses and empathy were reflected in the skillful manner in which he incised the manuscript, moulded the writer's learnings, and left him with an intact ego. Grateful appreciation is also extended to Dr. Charles Henley who, as guidance committee chairman, guided and charted the course through the doctoral program. Appreciation is also extended to Dr. Richard Featherstone and Dr. Donald Melcer, committee members who were most supportive. A very special thank you is due Dr. Donald Burke and Ms. Martha Rowland who develOped and administered one of the treatments used in this research. They expended. much personal time and effort as well as a willingness to have their methodology subjected to evaluation of the most critical kind, that being by comparison. iii Gratitude is also due to many staff members of the Coldwater State Home and Training School. Dr. C. Dale Barrett, Superintendent, for his interest in and desire to have the study carried out at Coldwater; to Mr. John O'Brien, Growth and DevelOpment Program Director, who Opened his home, his heart, and the program to me, his many kindnesses will long be remembered; to Mr. David Sluyter, Director of Behavioral Treatment for his help and advice in the development of the evaluation instrument and for the many hours of personal time which he sacrificed for the conduct of the reliability checks. Thanks are also due to Mr. Michael O'Neil, Educational Consultant, for his advice and counsel. And, lastly and most im- portantly, to the attendants who tolerated my weeks of observation and scrutiny and whose dedication and friendliness brightened a most difficult task. Appreciation is also extended to Dr. Andrew Porter and his staff in the Office of Research Consultation, especially Mr. Robert Wilson and Miss Mary Kennedy for their aid in the design of the study and the analysis of the data. My thanks also to Mr. Maxwell Moore, Special Edu- cation Librarian, who ferreted out numerous reference sources . iv Finally my heartfelt thanks to my wife, Marilyn, for her faith and hope in this venture and to our children Martha, Edie, and John, whose exuberance with life never faltered. TABLE OF CONTENTS Chapter Page I. INTRODUCTION AND REVIEW OF RELATED RESEARCH . . . . . . . . . . . 1 Behavior Modification . . . . . . . 3 Training of Attendants in Behavior Modification . . . . . . 4 The Design of the Present Study . . . 8 II. METHODOLOGY . . . . . . . . . . . 12 Selection of Attendants for the Study . 12 Assignment of Attendants to Treatment Groups . . . . . . l4 Characteristics of the Attendants . . . 15 Training Procedures in Treatment I (On—the-Ward) . . . . . . . . 16 Training Procedures in Treatment II (Off-the-Ward) . . . . . . . . 21 The Behavior Analysis Rating Form . . . 27 Description of Rating Form Items . . . 29 Definitions of the Derived Scores . . . 36 Rater Agreement . . . . . . . . 39 Reliability of the Derived Scores and Ratings . . . . . . . . . 4l Observational Procedures . . . . . . 42 Hypotheses . . . . . . . . . 46 Treatment of the Data . . . . . . . 46 III. FINDINGS . . . . . . . . . . . . 49 Distribution of Global and Intent to Teach Ratings . . . . 60 Correlation of the Observation Variables and the Global Rating . . . . . . 61 vi Chapter Page IV. SUMMARY AND CONCLUSIONS . . . . . . . 64 Conclusions and Implications for Current Institutional Practice and Research . . . . . . . . . 67 Implications for Research. . . . . . 70 REFERENCES. . . . . . . . . . . . . . 72 APPENDIX A. Raw Scores on Six Variables. . . . . . 74 B. BARF, Behavior Analysis Rating Form . . . 75 C. Language Observation Form . . . . . . 76 D. Language Scale Assessment Form. . . . . 77 vii 10. 11. 12. LIST OF TABLES Attendants' Mean Age, Education, and Length of Employment by Treatment Group . . . . Correlation Coefficients of Inter Rater Reliability . . . . . . . . . . . Reliability of Scores as Estimated by Pre- Test, Post-Test Correlations . . . . . Number of Cases by Kind, Shift, and Treatment . . . . . . . . . . . Multivariate Anova Table . . . . . . . Pre-Test and Post-Test Measures for Treatment Groups I and II on the Six Variables . . Means, Standard Deviations, and Range for the Sums for Appropriate and Inappropriate Reinforcement . . . . . . . . . . Treatment x Kind Interaction Post Hoc Test on Univariates . . . . . . . . . . Means on Inappropriate Technique for Treatment by Kind Interaction . . . . . . . . Treatment x Kind x Shift Interaction Post Hoc on Univariates . . . . . . . . Variable: Reinforcement Difference . . . Comparison of Post-Test Means for the Combined Groups Using Different Activity Times . . viii Page 16 4O 41 47 51 52 54 55 55 57 58 '60 Table Page 13. Distribution of Pre-Test and Post-Test Ratings for Global Rating and Intent to Teach by Treatment Group . . . . . . 62 14. Raw Scores on Six Variables . . . . . . 74 ix Figure 2. LIST OF FIGURES Page Interaction Model . . . . . . . . . 17 Treatment x Kind Interaction for Variable: Inappropriate Technique. . . . . . . 56 Treatment x Kind x Shift Interaction for the Variable: Reinforcement Difference . 59 CHAPTER I INTRODUCTION AND REVIEW OF RELATED RESEARCH This study compares the effectiveness of two different instructional programs for the training in be- havior modification techniques of attendants in a state residential institution for mentally retarded children. The need for such training has increased as a result of two relatively recent developments: a change in the role of the residential facility and accumulating evidence pointing to the efficacy of behavior modification techniques in the training of moderately and severely re- tarded children. There has been over the past two decades a general shift in the emphasis of treatment programs in residential institutions for the mentally retarded. For many years institutions were built some distance from population centers and their emphasis was on long-term custodial care based on sound medical practice (Tarjan, 1966). This medical model was one in which the role per- ception of the retarded individual was that of a sick person (Wolfensberger, 1969). Characteristic of this model is an administrative hierarchy of physicians and nurses. In such a setting residential care is thought of as nursing care. Programming is regarded in terms of therapy or treatment. Currently in favor is a conception of the role of the institution which has been described by Wolfensberger (1969) as the "developmental model." It is one in which the retarded individual is viewed as a developing person. Even the most severely retarded are seen as capable of growth, development, and learning. Institutions committed to this model are characterized by "facilitating inter— action between the retarded person and his environment, maximum resident-attendant interaction, and the creation of an atmosphere which is similar to that of the non- handicapped community." This new direction has come about as a normal function of social change, and greatly en- couraged through the action of parent groups which are demanding that habilitative services of treatment and training, replace custodial care (Ross, 1966; Tarjan, 1966). The burden of responsibility for training falls most heavily on those staff members with whom the retarded individuals spend the major portion of their time: namely, the ward attendants. Thus, if the new goals of these changing institutions are to be realized it is imperative that these ward attendants add to their custodial skills some competencies in training. Historically, despite the fact that attendants have more contact with the residents than do other institution personnel, they typically have had no relevant specialized training as have nurses, teachers, and social workers (Butterfield, 1969). Most attendants come to their positions with little or no training or experience that would prepare them for this occupation (Parker, 1951). Behavior Modification Training methods usually applied with the mildly retarded child, are ineffective with severely retarded children as they lack the prerequisite language skills and the relatively complex behavior repertoire that is required by these more conventional techniques (Heber, 1961). Moderately and severely retarded children have been successfully trained using behavior modification principles. Currently, research in behavior modification is one of the major areas of concern within mental re- tardation (Gardner & Selinger, 1970). The literature concerned with investigations of the use of operant conditioning techniques with insti- tutionalized retardate is replete with evidence of suc- cess. Most of the studies have dealt with the modifi- cation of self-help and social skills in institutionalized residents. Language acquisition has been accelerated while other behaviors such as head banging have been decreased through this method. In his summary of research on the employment of behavior modification techniques with institutionalized mentally retarded individuals, Ashbaugh (1971) found that of the 57 studies reviewed, 47 showed evidence of success. Training of Attendants in Behavior ModificaEIOn There have been several studies dealing specifi- cally with the training of attendants in the use of be- havior modification techniques. Ashbaugh (1971) compared the effectiveness of training in behavior modification with training in the content of a traditional attendant training program. Regarding the 1atter,Ashbaugh states, ”The specific areas covered . . . included usual attend- ant duties that involve attendant-resident interaction." No mention is made of training the control group in be- havior modification techniques. Effectiveness of the training was measured by the subjects' spontaneous use of behavior modification princi- ples in their suggestions for the handling of six instances of patient behavior displayed in two-minute TV—tape pre- sentations. It is hardly surprising that Ashbaugh found that those attendants trained in the use of behavior modification principles verbalized them more in their suggestions than did those attendants not exposed to behavior modification. It should be emphasized also that the criterion variable was not the behavior that attendants actually demonstrated in handling patients, but their verbal response to the question, "How would you change this resident's behavior?" Thus, the effectiveness of training was measured by what the attendant said should be done in a hypothetical situation. Certainly measure- ments can be more easily obtained in a structured situation such as that used by Ashbaugh. However, data obtained through simulation reflect only what an attendant might do in the comfort of simulation and may or may not reflect his behavior in a real life ward situation. Gardner and Giampa (1971), for example, indicate that this method is susceptible to faking. Additionally only limited behaviors are being observed and therefore comprehensiveness is diminished. Further, since the attendant knows that he is being observed, and is probably aware of the dimensions of the observations, his behavior is not likely to be representative. Mattos (1966), like Ashbaugh, compared attendants trained in behavior modification with those given a tra- ditional training program. However, the dependent vari— able in his study was the degree of "training interaction" as opposed to the "management-interaction" between attend- ant and institution resident. Using time sampling procedures he observed attendant-resident "interaction units" and recorded (1) whether the attendant initiated the interaction or merely responded, (2) whether or not the interaction was aversive, and (3) whether or not it was aimed at training. He counted the amount of inter- action and found no significant difference between the groups in the total amount of interaction, but, as pre- dicted, "training attendants in the use of behavior modi— fication techniques increased the amount of attendant time devoted to resident training." Although Mattos does focus his attention on the attendant as he works on the ward, the criterion of the degree of "training, inter- action" falls short of measuring the important dimension of application of the learned techniques. It is con— ceivable that training in transaction analysis or sensi— tivity training could have brought about the same increase in time devoted to resident training. Gardner (1970) compared two methods of training attendants to use behavior modification principles: a role-playing training procedure and a traditional lecture type of presentation. Attendants were randomly assigned to one of two treatment groups. Those in role playing were exposed to sessions in which operant conditioning techniques were demonstrated by a behavior modification supervisor. Then, attendants working in pairs, each alternately assuming the role of patient or attendant, practiced the pre- viously demonstrated techniques. The lecture group attended sessions covering definitions and causes of mental retardation, shaping procedures and reinforcement. Measurement was done with a paper and pencil Behavior Modification Test and the Training Proficiency Scale, an observational instrument utilized during a role- playing session of a standard and a novel situation. Significant differences were found between the treatment groups indicating that role playing contributed more to training proficiency. Gains in knowledge of be- havior modification were more significant in the lecture group. It could be expected that the group trained by the lecture method would do better on a paper and pencil test in the same fashion as those trained by Ashbaugh were better able to verbalize these techniques during simulation. It should be noted that Gardner, like Ashbaugh and Mattos, did not attempt to demonstrate that the attendants applied what they had learned in real life: that is, in interaction with residents on the ward. In fact, role playing was used both to train attendants and to measure the effects of training. Gardner did attempt, however, to demonstrate that generalization had occurred by changing the types of problems that were presented in the role-playing sessions. Thus this study is open to criticism because of Gardner's use of simulation through role playing as the dependent variable. As noted pre- viously, Gardner himself has criticized simulation for its vulnerability to faking, its "lack of representativeness" and "lack of comprehensiveness." The Design of the Present Study The present study was designed with two primary objectives in mind: (1) to determine whether a method of training attendants on the ward is feasible and effective, and (2) to evaluate the effectiveness of training in terms of the attendant's later ward behavior. Regarding the first objective, training on the ward has the following obvious practical advantages: 1. An attendant trained entirely in the setting in which he works is less dependent on transfer of training for success in his subsequent performance. 2. From the point of view of administrative arrange- ments and service responsibilities there are obvious practical advantages to training on the ward: the attendant need not be replaced while he is attending a class, with the result that staff levels can be maintained and services to the children held at the customary level. The method of on-the-ward training selected was that deve10ped by Burke and Rowland (1971). The principle of primary importance in their method is the reinforcement of the attendants during training for their appropriate utilization of behavior modification techniques. A method using the reinforce- ment of attendants was first employed by Bricker, Morgan, and Grabowski (1968). In their study they used commer- cial trading stamps as general reinforcers to motivate nine institutional attendants working with low-functioning children. The principle dependent variable was the amount of interaction between attendants and children. Other dependent variables were increased use of tangible reinforcers, reduction in punishment, and fading of motor prompts. The Burke and Rowland (1971) method by contrast uses positive reinforcement in the form of verbal rewards for the acquisition of skills involved in behavior modifi- cation techniques. It also requires that the teaching- 1earning and reinforcement occur on the ward while the attendant is working with the children in his charge. This training method was compared with what was considered to be a well-designed program having comparable content but utilizing traditional classroom sessions alternating with practical laboratory, but off-the-ward, experiences. In order to maximize the relevance of the findings of this study for decision making regarding in-service training programs the amount of staff time 10 devoted to the two training procedures was made equal. As a consequence, since each attendant was worked with individually, the amount of time the attendant spent in training by the on-the-ward training group (Treatment I) was considerably less; approximately one-eighth that of the off-the-ward group (Treatment II). The difference in the two methods can be high- lighted by the following comparisons: Treatment I Treatment II 1. Trained on-the-ward l. Trained off-the-ward 2. Trained with children 2. Laboratory experience in the Subject's with "new" children charge 3. Each attendant received 3. Each attendant received approximately 5.0 hours approximately 40 hours of training of training 4. Approximately 120 hours 4. Approximately 120 hours of staff time of staff time In regard to the second objective, the effectiveness of the training programs was evaluated by observations of the behavior of attendants in their interactions with 'children during the performance of their normal duties on the ward. The problems inherent in other methodologies previously described are circumvented by this method of evaluation. In summary, the major issues to which this study is addressed, therefore, are whether an institution can effectively train attendants in behavior modification 11 techniques without the disruptions created by assembling attendants for class meetings, and whether attendants will apply their learning during the routines of later on-the- job performance. CHAPTER II METHODOLOGY This investigation was conducted in the Growth and Development Cottage at the Goldwater State Home and Train- ing School, Coldwater, Michigan. The resident population of this cottage consisted of approximately 178 severely and profoundly retarded, ambulatory boys and girls with minimal mobility problems. The group 1.0. range was from 0 to 30 and the age range from 9 to 21 years. The attendants who were the subjects in this study spent approximately eight hours a day with these children, aiding them in such activities as toileting, eating, dress— ing, free play, as well as structured and unstructured employments. Selection of Attendants for the Study Initially it was planned to randomly select atten- dants from all cottages of the institution. However, as a result of administrative difficulties, as well as logistical considerations in pre-testing, training, and post-testing of attendants located all over a large 12 13 institution, a decision was made to utilize the Growth and Development Cottage attendants who are representative of those found throughout the institution. It was determined that the institution's hiring practices were standard for all units, therefore all at- tendants met the same basic criteria for employment. Ap- portionment of attendants to cottages and wards within them is based on availability of open positions. Union- ization precludes assignment to particular wards as a means of disciplinary action. The employee turnover rate in this cottage is comparable to that of the institution at large. The children, previously described and served by these attendants, approximate (with respect to age, degree of mental handicap, emotional overlay, and ambulation) children found in similar type cottages at other institu- tions. Thus, it was felt that this attendant sample was representative of attendants throughout the institution and that the children served by them typify the population of severely and profoundly retarded children found in similar institutions. It is, therefore, assumed that generalizations drawn from the results of this study are. applicable to personnel in similar state institutions. 14 Assignment of Attendants to Treatment Groups Placement of attendants in treatment groups was done by assignment of wards to treatments rather than by the random assignment of individual attendants. This was necessary as it was felt that, if attendants work in close proximity during the same hours or if they work together for a short time on overlapping shifts, it can be assumed that they will communicate with each other or observe each other in their handling of children. For this reason, it was decided that all of the attendants on a ward, in both sections, and on day or afternoon shift, would be assigned to the same treatment. Thus, the only randomization pos- sible to avoid this contamination was the assignment of wards to treatment groups. One side of the building (two wards) serves younger children, the other side, older. Thus in order to secure comparable ages in the two treat- ment groups it was necessary to pair the two younger wards and the two older wards prior to assignment to a treatment group. One ward on side one of the building was assigned to Treatment I through the toss of a coin. The other ward on that side was then automatically assigned to Treatment II in order to provide a younger ward for each treatment- The same random procedure was followed in assigning the wards on the other side of the building to treatment groups. The coin-tossing was done by persons other than 15 the experimenter, who was kept in ignorance of the treat- ment groups to which the wards were assigned. This was necessary since the experimenter carried out the subse- quent pre-test and post-test ratings of attendant behavior. Characteristics of the Attendants The attendant pOpulation selected for this study consisted of 44 full-time attendants assigned to the day (7:00 a.m. to 3:00 p.m.) and afternoon (3:00 p.m. to 11:00 p.m.) work shifts. Because of the retirement of one employee and 2 cases of illness, only 41 subjects were available during the post-test and one of these was randomly dropped so as to equalize the size of the treat- ment groups for purposes of data analysis. Of the 40 remaining, 22 were female and 18 male. The subjects' mean age was 36.7 years and the average length of employment was 4.4 years. The average educational level of the 40 subjects was 11.4 years. Table 1 summarizes the average age, length of institution employment, and years of edu- cation by treatment groups. The treatment groups were compared with respect to their age, education, and length of service at the institution, using a 3 test for each comparison. None of the E values obtained were statistically significant. 16 TABLE 1 Attendants' Mean Age, Education, and Length of Employment by Treatment Group Variables Treatment I Treatment II n=20 n=20 t Age i = 39.8 i = 33.7 .39 SD = 14.3 SD = 10.5 Education i = 11.7 i = 12.3 .13 SD = .93 SD = 1.26 Length of Employment 2 = 5.44 i = 3.27 .23 SD = 4.7 SD = 3.7 Significant at the .05 level. Training Procedures in Treatment I (On-the-WardTi The basic assumption underlying this method of training is that effective training can best be accom- plished when attendants are involved with a child in their own ward milieu. This treatment, therefore, avoided the use of standardized or simulated experiences as adjuncts to training. The interaction model (Figure 1) depicts the relationship between the trainer (specialist), the attendant, and the resident. 17 SPECIALIST {—9 ATTENDANT (—) RESIDENT Figure 1 Interaction Model The area of behavior selected for the teaching of behavior modification techniques was that of language. Language was chosen as it is employed in all areas of life and language training could be accomplished in any setting and without the need for materials other than those readily available on the wards. The training procedure included involvement of the attendant in an informal assessment of the language behavior of residents in his care and specific instruction in language development training. Attendants on the day shift were assigned to the same group of 9 or 10 residents while afternoon shift attendants were responsible collectively for all resi- dents on their respective wards. Therefore, the initial informal language assessment of each resident was con- ducted routinely as a part of the training procedure with 18 day shift attendants while afternoon shift attendants reviewed the results of the assessments, and added infor- mation when possible, prior to the instructional phase of the training. In general then, within the ward setting, attendants: participated in an informal assessment of the language behavior of residents in their care; were taught, through demonstration by a trainer, techniques to be used in language development training of residents; practiced with residents the techniques demonstrated by the trainer; and had apprOpriate techniques, as they occurred, subsequently reinforced. Ten training sessions of 30 minutes duration were conducted with each attendant. Initially, the two trainers worked as a team, alternating responsibilities to insure standardization of assessment, recording, and training procedures. After each attendant had at least one session with the trainers as a team, it was determined that the trainers had reached agreement on all procedures thus enabling them to work independently while still pro- viding equivalent training opportunities to the attendants. Each attendant on the evening shift, as a result of lower staff to resident ratios, usually was trained with a small group rather than with a single child. The trainers felt that the presence of a small group of chil- dren did not interfere with the training. 19 The trainers walked on the ward, approached an attendant and a child or group of children and explained to the attendant that they were going to provide person- alized in-service training consisting of 10 sessions of 30 minutes duration. The trainer then set a timer for 30 minutes and the assessment began with the trainer asking the attendant, "Does _____ respond to his name?" Regardless of the attendant's answer, a quick informal check was made by having the attendant call the child by name. The result was recorded on the Language Scale Assessment Form (see Appendix D). The assessment continued until the trainer was satisfied that the level of language development at which the child could not perform had been reached. In a rapidly conducted consultation with the attendant, a decision was made concerning the language level at which training would begin. This assessment procedure was usually accomplished in about 15 minutes, allowing the remaining 15 minutes of this first session to be devoted to training of the attendant. If, as in the case of the afternoon shift attend- ants, an assessment of the child or children had already been accomplished, the trainer reviewed the results of the previously conducted assessment and moved directly into the instruction phase of the training. 20 Instruction began with the trainer demonstrating a technique for moving a child from the previously deter- mined level of language functioning. Care was taken to use articles usually found on the ward for both language assessment and training. Included, for example, were the child's own clothing, 'washcloth, a food tray used at mealtime, soap, and the same play articles with which the child was familiar, thus minimizing the need for transfer of learning during training. During the demonstration the trainer reinforced the child's behavior with one of the categories of positive reinforcement included on the Behavior Analysis Rating Form (BARF). Inherent in the demonstration was a description of shaping procedures and their implementation. Poor training strategies such as delayed reinforcement, reinforcement for non-performance or incompatible behavior were described, as well as their contingencies. Where apprOpriate to the situation, the technique of ignoring inapprOpriate behavior, and the contingency of its reinforcement were also described and demonstrated. The necessity for finding the right re- inforcer for the individual child was demonstrated to- gether with the requirement for a change in reinforcer at the point of satiation. The use of punitive measures to bring about a cessation of behavior was discouraged. Subsequently the attendant was given an Opportun- ity to practice the technique just demonstrated. As he 21 demonstrated the appropriate behavior modification tech- nique in teaching language to a resident, the trainer reinforced him for the apprOpriateness of his response. For the most part, attendants worked with differ- ent children in each of their training sessions so as to provide the attendant with a more representative sample of behaviors and responses as well as to demonstrate the applicability of his newly-learned techniques with a variety of children. The settings for training were also changed for the same reason, thus allowing for training in language using behavior modification in the dining area, the ward room, and the bathroom. Even in those instances where all situations were not covered, it was felt that the attendant could easily generalize a language training technique learned during free activity to a dining room situation. Training Procedures in Treatment II (Off-tfie-Ward) Treatment II was more conventional in its struc- ture than Treatment I in that it consisted of formal class sessions, with lecture and discussion, as well as practicum experiences. Attendants on wards assigned to this treatment were relieved of regular duties for a one-week period. The attendants each morning attended class, and each afternoon worked on the ward, applying the content material discussed in the morning session to 22 practical situations with children. The morning sessions included lectures dealing with behavior modification principles, audio-visual presentations, and group discus- sions, including discussion of the previous day's prob- lems in the practicum experience. The content for this program included: Observation and Analysis of Behavior (1 hour) a) specifying behavior b) counting behavior c) utilizing data Management of Behavior Problems (6 hours) a) behavior modification techniques b) restructuring the environment c) teaching appropriate behavior modeling d) watching for blowups or other inappropriate behavior Techniques for Developing Self-Help Skills (6 hours) a) Areas: (1) feeding (2) dressing (3) toileting training b) Methods: (1) appropriate sequencing of needed sub-skills (2) small-step approach 23 (3) proper techniques (repetition, scheduling) (4) utilization of behavior modification 4. Language Development a) What is language? Why is it important? b) How does it develop? c) Why does it not appear to develOp in severely retarded children? d) Fostering language development The first four afternoons of the week the attend- ants worked with small groups of children in rooms not regularly used for ward activity. On Friday afternoon they returned to their regular wards to apply the prin- ciples learned during the week. It should be noted that the behavior modification portion of the treatment was designed and carried out by regular employees of the institution who have had extensive background and experience in the use of the techniques. The language portion was instructed by an experienced language therapist. The following training procedures and practices were directly extracted from the training materials used in this treatment to provide the reader with a better grasp of the treatment methodology. Throughout the week of in-service training for each group of attendants, a series of problems were devised so as to give experience with behavioral observations, data col- lecting, graphing, and a viable exposure to basic concepts of behavior modification under structured, supervised conditions. 24 Below are the problems which were given to each attendant: Monday: Problem I: Problem II: Problem III: Problem IV: Tuesday: Problem I: Observe 1 child for 5 minutes. Choose a behavior. Observe and measure its fre- quency, and duration for 5 minutes. Now do the same for 3 other children. The behavior should be overt, observable, and measurable. Using the same children and behavior in Problem I, hold two or more observation periods on each child, lasting 5 min- utes, with the 5-minute break in be- tween each observation period. Now, determine a base rate for the observed behavior. By use of positive primary and secondary reinforcements, modify the behavior of a resident observed in Problems I and II. The treatment program should last 5 minutes. Now, do the same for the 2 other residents observed in Problems I and II. At the end of the day, using the graph- ing format described in class, and the data collecting during the observation and treatment periods, construct a graph for each child observed to display its base rate behavior and the effects of the treatment program. Using a reversal design, do the follow- ing: 1. Observe a resident for l lO-minute observation period and report the frequency or duration of his behavior. Rest for 5 minutes. 2. Modify the frequency or the dura- tion of the behavior by the pair- ing of primary and secondary reinforcements during a lO-minute treatment period. Rest for 5 minutes. Problem II: Wednesday: Problem I: Problem II: 25 3. Observe and reward the child's behavior again for 10 minutes, but do not reinforce the child for his behavior by either primary or secondary reinforcers. This is the reversal phase. 4. Record the data for all 3 sessions and graph them at the end of the day, using the format described in class. Using a different resident, repeat Problem I as outlined above, but add a post-treatment phase after the reversal phase. The post-treatment phase should last 10 minutes and should be carried out in a manner identical to the treat- ment phase. Record the data for all 4 sessions and graph them at the end of the day, using the method demonstrated in class. Using the Language Observation Form (see Appendix C) given by the Speech Therapist, observe a child and rate his development in each area. Using the basic behavior modification techniques described in class, as well as primary and secondary reinforcers, develop and carry out a miniature language training problem to be used with a Specific child. Include 2 ob- servation periods, a treatment period, a reversal period and a post-treatment period with 10 trials for each child. Problem III: Problem IV: Thursday: Problem I: 26 Using a different child, repeat Problem II. Graph your results using the class format. This problem is designed to illustrate the concept of straining the ratio. 1. Pick a child; observe and record a specific behavior for 10 minutes. 2. Determine a reinforcement sched— ule which you feel will effectively modify the observed behavior; e.g. reinforcement after 3 appropriate reSponses or at the end of every 30 seconds interval, providing the child is behaving apprOpriately. Take a 5 minute break. 3. Initiate a 10 minute treatment schedule utilizing your reinforce- ment schedule. 4. Take a 5 minute break. 5. By at least 5-fold, increase the number of required appropriate responses or elapsed time interval before reinforcement is given. Using this modified reinforcement schedule, initiate another 10 minute treatment phase. 6. Take another 5 minute break. 7. Do a reversal, removing all rein- forcers, both primary and second- ary. - 27 Problem II: Using another child and a different behavior, repeat Problem I. Problem III: Graph your data from Problem I and Problem II, according to the class format. Throughout the week, questions and issues raised by the lectures and practicum were clarified by: 1. Four quizzes and 1 final exam, which were not graded, but were carefully discussed in class. In this manner, each attendant got feedback as to how well he understood certain basic concepts from the accuracy of his answers and any questions he had were answered by the discussion. 2. A discussion period during the first hour of each morning class, with the exception of Monday morning. At this time, the instructor explained why the children acted as they did during the practicum and gave feedbacks as to what the at- tendant did right or wrong during their little experiments. It also gave the instructor a chance to bring the previous day's class lecture down from a theoretical level to the ward level. Additionally, all graphing exercises at the end of each daily practicum were carefully supervised so that data would be displayed in a manner consistent with the pro- fessional literature. This approach was used so that they would learn to diSplay their results in an effective man- ner to interested professionals or para-professionals and also be able to better interpret professional litera- ture dealing with behavior modification. Last, each afternoon before the supervised practium began, each problem for the day was discussed so as to clearly explain the rational and procedure. The Behavior Analysis Rating Form Since this study dealt with the measurement of the utilization of behavior modification techniques by 28 institutional attendants in their day-to-day involvement with their mentally retarded charges, an observational instrument specific to this end needed to be designed. Ashbaugh (1971) utilized an observation check- list for gathering data. This was not suitable in this study since the items of interest to him were not entirely compatible with the Specific skills being taught in the two training methods in this evaluation. Ashbaugh ob- served only for the presence or absence of a behavior. The instrument develOped for this study has provision for the recording of the presence of behaviors, their quality, as well as frequency. A scale described by Gardner et_al.(l970), and developed to assess the proficiency of individuals using behavior modification techniques was also considered. Al- though the authors cite it as having validity and relia- bility, at least in simulated situations (Gardner), its desirability for use in this study was hampered by its vagueness and by the sophisticated level of reinforcement skill being striven for. Again, as was the case with Ashbaugh, consideration was not given to frequency of behavior or its inapprOpriateness, these being factors which this study took into account on its measures. A Behavior Analysis Rating Form (see Appendix B) was, therefore, develOped to record behaviors that re- flected on the success of the attendant in applying 29 behavior modification principles. The items to be ob- served were derived primarily from an examination of both training programs, as well as through field-testing of a prototype instrument at day-care centers serving children similar to those in this study. During these field-tests the experimenter recorded on audio-tape, descriptions of the precise behaviors initiated by the day-care workers as they worked in the classroom milieu with children during an observation period. The tape was subsequently played back and the experimenter attempted to rate the worker on the prototype instrument. During this playback it was observed that its method made for a loss of infor- mation since verbal recording could not be made as fast as action occurred. Room noise from the active classroom also rendered some of the tape unintelligible. However, the recording revealed two variables not previously covered, these being reinforcement for incompatible be- havior, and use of a reinforcer which was too satiating. The definitions of the items, their scope, and limitations were conjointly determined by the experimenter and a professional staff member of the institution who subsequently served as the second rater for reliability and consistency checks during the course of the evaluation. Description of Rating Form Items The Behavior Analysis Rating Form is divided into quadrants. The upper left quadrant contains apprOpriate 30 means of reinforcing behavior. The lower left quadrant contains nine apprOpriate behavior modification tech- niques. The lower right-hand quadrant consists of nine inappropriate strategies in the application of behavior modification techniques. The bottom segment comprises three lines. The first is a time line of 0 to 10 minutes. Its purpose was to enable the experimenter to record, using a code, what it was that the attendant was doing during the 10 minutes observation period. This scale provided background information which, it was felt, could be helpful in explaining the presence or absence of the training behaviors recorded on the remainder of the form. The line marked "Intent" provided for a rating of the strength of the attendants' intent to engage in teach- ing behavior with children during the observation period. The line marked "Beh. Mod." was used to rate the attendants' overall level of competence in the application of behavior modification principles. Following are the definitions of the behaviors categorized for use in the Behavior Analysis Rating Form: Failure to Use Opportunity to Reinforce l. Unused Opportunity: Failure to reinforce an ap- propriate behavior during an interaction sequence. This latter being a time when there is interaction between the attendant and residents and precludes the counting of unreinforced behaviors occurring outside the focus of the interaction. 31 ApprOpriate Reinforcement Administered 2. Appropriate Verbal: Enthusiastic verbal response to an appropriate behavior. Includes such re- sponses as: "good girl," "good boy," "thank you," "very good," etc. It does not include the use of social amenities such as: "hello," "good morning," etc. Appropriate Gestural: Waving or nodding or otherwise signaling approval through some physical manifestation other than touching. Appropriate Physical: Enthusiastic response em- ploying physical contact with the child. In- cludes such responses as patting the child's head or hand or hugging the child. ApprOpriate Tangible: Presentation of food, a toy, a token or anything material to a child as a reward for a particular behavior. Appropriate Physical—Verbal: Response to a child which simultaneously employs the use of physical reward with a verbal response: for example, patting the child's shoulder while saying "that's a good girl." Appropriate Verbal-Tangible: Response to a child which simultaneously employs the use of a verbal reSponse with a tangible reward; for example, say- ing "good boy" while giving the child a raisin. ApprOpriate Physical-Tangible: ReSponse to a child which simultaneously employs the use of a physical reSponse with a tangible reward; for example, patting the child's shoulder while hand- ing him a raisin or an M&M. Inappropriate Reinforcement Administered 9. Reinforcement for Non-Performance: Rewarding the child as described previously but at those times when behavioral criteria are not met; for example, giving the child an M&M for making his bed when, in fact, it is left unmade. 10. 11. 32 Reinforcement Too Delayed: Reinforcing a behavior after another behavior has intervened leads to confusion as the child doesn't know which behavior he's being reinforced for performing. Strength of the reinforcer is reduced as the time between reinforcement and the behavior occurrence in- creases. Reinforcement for Incompatible Behavior: Rein- forcement of a response which is opposite to that which is to be elicited. For example, in training a child to stay seated, reinforcement for standing would be reinforcing an incompatible behavior. Acquired Skills 12. 13. 14. 15. 16. Ignores Behavior: Non-reinforcement either posi- tively or negatively of behaviors outside the target behavior. For example, an attendant might ignore a child flapping his hand in front of his face during toileting. Conscious Change of Reinforcer: Realizing that a reinforcer such as cereal is not eliciting the desired behavior, the attendant switches to some- thing else such as raisins, M&M, etc. Time Out, ApprOpriate: Finding that a child at a certain point in time cannot cope with a parti— cular task and is being disruptive, the attendant removes the child from the task until he is back in control. Also refers to the interruption of the task, such as removing a child's food plate when he becomes disruptive during a mealtime. Time In, ApprOpriate: Restores child to task be- fore the end of the period so that he understands that the removal was not punishment but really a means of his regaining control of himself. Credit would be given if the attendant deliberately let a child remain in a time out state because his disruptiveness persisted. Fading: Gradual reduction in the physical assist- ance given by the attendant. For example, initially it may be necessary to put both arms on a child's shoulders to get him to sit down, sub- sequently one arm, then just a touch. 17. 18. 19. 20. 33 Prompts: Physical assistance necessary to aid a child in learning a task. For example, enclosing the child's hand with that of the attendants while holding a spoon, and gradually lifting hand and spoon to the mouth. Modeling: Physical gesturing to show a child how to perform a task and requiring that he mirror that gesture. Successive Approximations: Reinforcement of the child for the performance of one step in a complex task which approximates the behavior itself. Cues: Use of a verbal prompt; for example, if it is desired that the child respond with "thank you," we might say "thank" hOping to elicit a completed "thank you." Counter-Productive Behaviors 21. 22. 23. 24. 25. 26. Verbal, Punitive Derogatory: Harsh, caustic or shouted verbal responses to the children. Physical Punitive: Slapping, pulling, hitting, or any physical contact seen as aversive in the general pOpulation. InapprOpriate Target: Establishing a target be- havior which is not consistent with established or normally accepted goals for children of a similar functioning level. Non-Recognition of an Inefficient Reinforcer: Attendant fails to realize that the child is not being reinforced by the response given for a behavior, and therefore fails to change the re- inforcer. Reinforcer too Satiating: Attendant gives too many of the same reinforcers in a given period, or gives too much of the reinforcer at one time (for example; a bag of raisins instead of one raisin). Non-Recognition of an Opportunity to Change Behavior: The attendant does something for a child which the child is capable of doing himself. Also includes doing work which apprOpriately could be done by the cottage-working boys or girls (higher functioning residents who work in cottages as an aid to the attendants). 34 27. Non-Recognition of the Difficulty of a Task: Re- quiring a child to do something for which he doesn't have the requisite skills. Would in- clude: requiring a child to pull up his pants when he doesn't understand the concept of up, pull, or possibly even pants. 28. Not Breaking Task Down: In a shaping situation, attendant doesn't break task down into its com- ponent parts. If task has been broken down, the element or components are not sufficiently simple for the child to be successful. 29. Too Many Expectations per Unit of Behavior: Re- quiring that a child perform several unrelated non-sequential tasks during a training sequence in which the child does not have mastery of any of the units and is still in the shaping stage. Following are the definitions for the numerical ratings found on the Behavior Modification Rating Scale (Global Rating) and the Intent to Teach Rating Scale. Behavior Modification Rating ScaIe (Global Rating) The scale was intended to rate the skill or accur- acy with which the attendant applied behavior modification. It was rated on a six-point scale from 0-5 with 5 repre- senting excellent use of behavior modification. The steps on the scale are defined below: 0 (Absent) No visible evidence of any behavior modification skills and/or evidence of inappropriate techniques such as Verbal Punitive Derogatory or any other counterproductive behavior found in. that quadrant of the BARF. 1 (Very Poor) Very meager attempts at the use of behavior modification evidenced by the use of few reinforcements during an observation session as well as the 35 possible use of inappropriate tech- niques such as are found in that quad- rant of the BARF. 2 (Marginal) Some evidence of acquisition of behavior modification skills as shown by the use of reinforcements, but negated by some instances of inappropriate ad- ministration coupled with some incidence of inappropriate techniques. to (Fair) Evidence of the acquisition of behavior modification skills shown in the use of a number of reinforcements with few being inappropriately administered as well as a low incidence of inappro- priate techniques. ab (Good) Evidence of acquisition of behavior modification skills shown in the use of a number of reinforcements with some variety and with few instances of inapprOpriate administration together with the use of apprOpriate techniques with few inapprOpriate techniques being in evidence. UT (Excellent) Evidence of the acquisition of behavior modification skills by the use of num- erous reinforcers with considerable variety and no instances of inappro- priate administration coupled with the use of several appropriate techniques with no inappropriate techniques being in evidence. Intent to Teach Rating Scale This scale was intended to rate the strength of the attendants' intent to engage in teaching behavior with the children during an observation period. It was rated on a six-point scale from 0 to 5 with 5 representing excellent productive involvement in teaching. The steps of the scale are defined below: 36 0 (Absent) No visible evidence of any intention to engage in teaching; for example, interacting with a child by rolling a ball to him with only playful intent. 1 (Very Poor) Minor attempt to engage the child, but no follow-through teaching interaction; for example, attendant shows a child a book and says "see the doggie" but does nothing more. This represents an ineffective try at teaching. 2 (Marginal) Some evidence of teaching intent; for example, showing a child a ball and asking "What is this?" without follow- ing up with continued interactions. The intent seems stronger than for rating 2, but he still fails to main- tain the teaching contact for a suffi- cient length of time. 3 (Fair) Definitely appears to start a teaching sequence, and uses some strategem to influence the child such as modeling, but the attendant is not persistent or varied in his approach. For example, demonstrating the stacking of blocks and having the child do likewise. 4 (Good) Considerable evidence of intention to teach. Shows some variety and per- sistence in his teaching approach. 5 (Excellent) High level of teaching involvement; for example, if using ward materials such as dishes, demonstrates their use, encourages modeling by the child, uses the naming process while promoting the child's verbal response through some form of reinforcement. Definitions of the Derived Scores These items were not used singly in testing the hypotheses of the study. Rather, the frequency of re- sponse to the various items within an item sub-group were summed and combined with other scores in some manner to 37 create new scores that seemed logically to best represent proficiency in the use of behavior modification princi- ples. These derived scores are the dependent variables in this study. Descriptions of these derived scores are given below. 1. Reinfgrcement Difference (Reinf. Diff.) This variable represents the sum of the subjects inappropriate uses of reinforcement subtracted from the sum of the times S is recorded as having used an appropri— ate reinforcer. Thus, in terms of the original items, it is the sum of items 2-8, minus the sum of items 9-11 (see Appendix B). In this measure, it is assumed that S's under- standing of behavior modification principles will be high- lighted, since it penalizes promiscuous and mindless re- inforcement. 2. Reinforcement/Opportunities Ratio (Rein/Oppor Ratio) The score represents the total number of times S is recorded as having used an appropriate reinforcer divided by the number of times 8 is judged to have been presented with an "opportunity" to provide an appropriate reinforcement. An "opportunity" is derived by adding the total number of times S is recorded as having used an appro- priate reinforcer to the total number of unused 38 Opportunities. Thus it represents the proportion of op- portunities to reinforce a child on which the attendant capitalized. If, for example, S reinforced a child five times, but missed five opportunities, he responded ef- fectively 50% of the time. 3. Appropriate Technique (Approp3 TeEhII This variable is defined by the number of behavior modification techniques of which S gives evidence in his behavior; it is a simple count of the number of different techniques used, and not of the frequency of their use. It is assumed that by demonstrating a technique at least once (e.g., "fading," or "successive approximation) S shows a grasp of the principle. Thus, this measure indi- cates demonstration of the presence or absence of a skill, or an understanding, and not the repetitiveness of its use. It measures the breadth of his understanding and not the intensity of his application of one or more single tech- niques. 4. Inapprgpriate Technique TInapprop. Tech.) This score represents the number of categories of behavior displayed by S that represent violations of be— havior modification principles or indicate their inappro- priate application. The rationale for the scoring of this measure is the same as that for variable 3 above. 39 5. Global Rating This measure is a rating by the experimenter as to the attendant's competence in the application of behavior modification principles. 6. Teaching Intention TTeach. Intent.) This measure is a rating by the experimenter as to the strength of the attendant's intent to engage in teaching behavior with children during the obser- vation period. Rater Agreement The experimenter carried out all of the Observa- tions that provided the data for this study. However, in order to determine the reliability of these observations and of the ratings, a measure of rater agreement was ob- tained. A clinical psychologist well—versed in behavior modification, and the experimenter simultaneously observed 10 attendants during the pre-test period and the same 10 attendants during the post-test period. Prior to the pre- test observations, the two raters had discussed the defi- nitions of the items, pre-tested the form on attendants not in the study, and revised the definitions until agree- ment had been reached. The inter-rater reliability for the six variables and for the pre-test and post-test Observations on the 10 subjects were determined by the use of Pearson product 40 moment correlations. These correlation coefficients are presented in Table 2. TABLE 2 Correlation Coefficients of Inter Rater Reliability n=10 Variable r r2 l. Reinf. Diff. Pre .95 90 Post .97 94 2. Rein/Oppor. Ratio Pre .99 98 Post .98 96 3. Approp. Tech. Pre .96 92 Post .90 81 4. InapprOp. Tech. Pre 1.00 100 Post 1.00 100 5. Global Rating Pre 1.00 100 Post .99 98 6. Teach. Intent Pre 1.00 100 Post 1.00 100 These high reliabilities confirm the raters' judgement following their experience that they had reason- ably unambiguously defined the behaviors to be recorded. The high coefficients also are in part attributable to the fact that there were a large number of zero scores, a fact which, of course, does not detract from the level of rater agreement. 41 Reliability of tpngerived Scores and Ratipgs To determine whether the derived scores and the ratings, the dependent variables, were reliable measures of attendant behavior, it would have been desirable to have utilized a test-retest procedure. However, this was not feasible given the time pressure under which the ex- perimenter operated. As an admittedly poor alternative, it was decided to use the pre-test and post-test scores and ratings in a test-retest design. This procedure would be valid if it could be assumed that all attendants would gain equally as a result of training. This assumption, of course, is not likely to be tenable for the entire group. Nevertheless, the pre-test, post-test correlations were determined as the only estimate of reliability avail- able. They are presented in Table 3. TABLE 3 Reliability of Scores as Estimated by Pre-Test, Post-Test Correlations (n=40) Variable r Rein. Diff. .71 Rein/Oppor. Ratio .46 Approp. Tech. .43 InapprOp. Tech. -0.11 Global Rating .57 Teach. Intent .44 42 These coefficients are hardly satisfactory as re- liabilities, but it is of importance that all but one show significant correlation. It can be assumed that they would be appreciably higher if the differential effects of train- ing could be partialled out. In any case, the original decision to measure behavior in the non-structured, natural setting of the ward made it unlikely that every attendant's capacity for the application of behavior modification prin- ciples would be reliably measured. Instead, it was assumed that, if training was more successful for one group, the probability of observing the improved performance would be greater for this group as a whole than for the other group. The fact that rater agreement was excellent argues against the conclusion that the rater failed to perceive behavior accurately. It is more likely that low reliabil- ity would be produced by uncontrolled variation in the type of child the attendant happened to be working with during the observation period, ward conditions as affected by ill- ness of the children, or absenteeism, etc. Observational Procedures As stated previously, the effectiveness of the training programs was evaluated by observing the behaviOr of attendants in their interactions with children during the performance of their normal duties on the ward. Each 43 attendant was observed during those periods of the day when interaction with children was assumed to be greatest. Observations were made on each attendant during each of three intervals, pre-test and post-test. These intervals were eating (E), toileting (T) and free activity time (A). Although there is a scheduled activity time on the day shift which would have provided a period of poten— tially higher interaction it was determined as unacceptable since a comparable observation time was not available dur- ing the afternoon shift. This method then, provided that each attendant be observed for a total of 30 minutes on both the pre-test and post-test. Prior to the pre-test period, the director of the Growth and Development Cottage scheduled three staff meet- ings to provide the experimenter with an opportunity to informally meet with the attendants. During this meeting, it was explained that during the week of November 8, the eXperimenter would be visiting on the wards recording in- formation. It was stressed that the purpose of the ob- servations was to determine the effectiveness of the forthcoming in—service training and that observations would need to be taken prior and subsequent to the training in order to evaluate the training program. It was also pointed out that the experimenter was not in any way as- sociated with the institution and that his only affilia- tion was with Michigan State University. It was emphasized 44 that the information obtained on individual attendants was the property of E and would not be shared with the institution's administration. The use of the stop watch was explained as being a means of keeping track of the observation time, as well as for recording time which the attendant was unable to devote to interaction with the children. The pre-test took place during the week prior to the first session of Treatment II, the classroom-oriented method. This treatment covered three consecutive weeks, one week for each of three groups of attendants. Treat- ment I was begun one week after the beginning of Treatment II and ran several weeks after the conclusion of Treatment II because of the individual nature of instruction in- volved. Concurrent training was possible because attend- ants from the two treatments were drawn from different wards. Post-testing was begun one week after the comple- tion of Treatment I causing a delay of up to five weeks between training and post-testing for some attendants trained in Treatment II. This lag was necessary since the experimenter was kept blind as to which wards were assigned to a particular treatment group and could not,' therefore, be on the wards during any part of the training. The experimenter stationed himself unobtrusively in a ward location which gave him auditory and visual access to the attendant as he related to a child. 45 Responses made by the attendant which were covered on the Behavior Analysis Rating Form were recorded by means of a stroke on the line adjacent to the response. At the end of the lO—minute Observation period, as deter- mined by a stOp watch, the experimenter recorded ratings of the attendant on the Scales which evaluated "Intent to Teach" and niques." "Utilization of Behavior Modification Tech- By observing during different periods of the day a more exhaustive view of the attendants' interaction was available, thus mitigating the criticisms of Gardner and Giampa (1971) regarding "lack of comprehensiveness" and "lack of representativeness" found in simulation type studies. Observation Times 7:45 8:30 11:00 11:45 12:30 2:00 4:00 4:45 6:00 6:30 A.M. A.M. A.M. A.M. P.M. P.M. P.M. P.M. P.M. P.M. 8:30 9:30 11:45 12:30 1:30 4:00 4:45 6:00 6:30 7:30 A.M. A.M. A.M. P.M. P.M. P.M. P.M. P.M. P.M. P.M. Eating Toileting Free Activity Eating Toileting Free Activity Toileting Eating Toileting Free Activity 46 Hypotheses It was the general hypothesis of this study that subjects trained in behavior modification principles by the personal interaction training method (Treatment I) would show equal or greater competence in their applica- tion of behavior modification techniques while working with children during their participation in ward routines than would subjects trained in behavior modification by a more conventional method (Treatment II). This general hypothesis was tested by a comparison of the two groups on gain scores derived from measures obtained from the Behavior Analysis Rating Form (BARF). For each variable the following hypotheses were formed: Variable Predictions for Treatment Group Mean Gain Scores 1. Reinf. Diff. Treatment I 3 Treatment II 2. Rein/Oppor Ratio Treatment I 3 Treatment II 3. ApprOp. Tech. Treatment I 1 Treatment II 4. InapprOp. Tech. Treatment I 3 Treatment II 5. Global Rating Treatment I 1 Treatment II 6. Teach. Intent Treatment I a Treatment II Treatment of the Data Comparison‘pf Pre-Test, Post- Test Gain Scores The data were analyzed by means of multivariate analysis of variance (MANOVA) of pre-test and post-test 47 gain scores for five of the six variables previously described. Variable 5 (Global Rating) was not included since it was a subjective judgment and not a score based directly on the incidence of behaviors. Its treatment is discussed below. Table 4 describes the design. TABLE 4 Number of Cases by Kind, Shift, and Treatment (n-40) Kind Shift Treatment I Treatment II Day 6 6 Younger Afternoon 4 4 Day 6 6 Older Afternoon 4 4 The individual attendant was the unit of analysis. "Shift" refers to the time of day when the attendant was working while "kind" indicates whether the children worked with were older or younger. Shift and kind are blocking variables that were introduced to add precision to the analysis. No predictions were made as to their inter- action with treatment. Relatippship of Scores and Global Rating Multiple and partial correlations were calculated for the relationship of variables 1, 2, 3, and 4 and the Global Rating. The Global Rating was originally included 48 in the study on the assumption that it might reflect behavioral changes not captured by the separate scores. This analysis permits an evaluation of the extent to which the scores and the global impressions are related. Subsequent to the MANOVA, a post hoc univariate test was done in those instances where the multivariate test revealed significance. CHAPTER III FINDINGS The general hypothesis of this study was that sub- jects trained in behavior modification principles by the personal interaction method (Treatment I) would show equal or greater competence in their application of behavior modification techniques while working with children during their participation in ward routines than would subjects trained in behavior modification by a more conventional classroom procedure (Treatment II). Specific hypotheses on each of the dependent variables were: 1. The group mean gain score on the Reinforcement Difference variable for Treatment I will be equal to or greater than that for Treatment II. 2. The group mean gain score on the Reinforcement/ Opportunities Ratio variable for Treatment I will be equal to or greater than that for Treatment II. 49 50 3. The group mean gain score on the Appropriate Reinforcer variable for Treatment I will be equal to or greater than that for Treatment II. 4. The group mean gain score on the Inappropriate Technique variable for Treatment I will be equal to or less than that for Treatment II. 5. The group mean gain score on the Global Rating variable for Treatment I will be equal to or greater than that for Treatment II. 6. The group mean gain score on the Intent to Teach variable for Treatment I will be equal to or greater than that for Treatment II. The significance of the group differences in mean gains was determined through a multivariate analysis of variance. It will be recalled that five of the six variables have been included. The results of this analy- sis are presented in Table 5. It is apparent in Table 5 that there is no signi- ficant treatment effect. Therefore, it can be concluded that the mean gains for the two groups on the five vari- ables are not significantly different. The hypotheses for these five variables are confirmed in the sense that the Treatment I mean gains are "equal to" those for Treatment II. The hypotheses were couched in these 51 TABLE 5 Multivariate Anova Table (Includes Variables l, 2, 3, 4, and 6) Sources of Variance df Multiple f P less than Treatment Effect 5 .2296 .9464 Kind Effect 5 1.4835 .2268 Treatment x Kind Interaction 5 2.6120 .0464* Shift Effect 5 1.4378 .2417 Treatment x Shift Interaction 5 1.0307 .4190 Kind x Shift Interaction 5 1.3502 .2728 Treatment x Kind x Shift Interaction 5 3.0700 .0248* *Significant at the .05 level. conservative terms since the Treatment I group received only one-eighth as many hours of instruction as the Treatment II group. It is obvious, at the same time, that all of the hypotheses would be confirmed if no instruction had occurred at all. Therefore, it is of concern to note whether there was equal improvement or equal failure to improve. To determine whether each of the groups had shown significant amounts of gain, in terms of significant increase in mean scores, a univariate analysis of vari- ance was performed to evaluate the gain for each of the six variables. These pre-test and post-test means, and the significance of the difference between means are pre- sented in Table 6. 52 .Hm>ma mo. man an unmoHMHcmAm. m5.H mo.H Hb.N HN.N Gm mm. va. mm. mm. on. own. 0N. mN.N mo.N M “COUCH SOMOB «H.H av. om.N VB.H Gm mv. ¢mv. 0N. mm. ma. mmm. mh. mm.H OH.H m mafiumm HMQOHU N0.H oo.H mH.H mm. Om mo. I mmm. mo. mN.H 0N.H hon. ma. I mH.H om.H m .flomfi .OHQQMGH hN.H mm. HN.H no.H 0m ma. I 0mm. om. mm. m5. Nbb. OH. I oo.H OH.H m .3008 .Ohmmd mm. mm. ma. mm. Om mo. mmm. mo. He. mm. NON. 50. mm. we. M Owfimm .Homm0\.mcwwm mn.m om.m Hv.m mo.m Gm mH.m mvm. mo.N 0N.m mH.¢ emoo. mN.v mm.NH ow.m m .MMHQ .MGHOM camu .mflm came umom mum .vwm Gama umom mum Hmuoa mOHQMHHm> HH ucmEumouB H useEummna mmaomwum> xwm on» so HH can H mmsouo useEumouB How monsmmmz umoanumom one umoaumum m mqmfia 53 As indicated in Table 6, only one post-test mean was significantly higher than its pre-test mean: Rein- forcement Difference for Treatment I, which was significant at the .009 level. The post-test means for Appropriate Technique were lower for both treatment groups, the pre- test mean for Inappropriate Technique for Treatment II was higher, which indicates poorer performance following treat- ment. The remainder of changes were in the direction of improvement, but the differences are negligible. There- fore, on the basis of these data, it must be concluded that with the one exception, this study failed to detect appreciable improvements in the attendants' use of behavior modification on the ward following either method of in- struction. It should be noted that the Reinforcement Differ- ence variable is the difference of two components, the total number of times an attendant is recorded as having used an appropriate reinforcer and the number of times a reinforcer was administered inappropriately. Table 7 shows means, standard deviations, and ranges for the sums of these two items. These data make it possible to determine whether the gain was brought about by increased use of reinforce- ment or a reduction in its inappropriate administration. For example, in Treatment II there was an increased use of reinforcement but this is reduced in the composite measure (Variable 1) because of the increase in inappropriate re-r inforcement. 54 TABLE 7 Means, Standard Deviations, and Range for the Sums for Appropriate and Inappropriate Reinforcement Treatment I Treatment II Item i SD Range i SD Range Total number of Pre 10.30 8.69 0-22 5.05 6.53 0-29 times S is recorded having used an appro- Gain 5.00 2.02 priate reinforcer Post 15.30 8.74 0-35 7.7 6.96 0-24 Total number of Pre 1.7 3.02 0-6 1.0 1.71 0-6 times S is re- corded having Post 1.0 1.296 0-8 1.25 1.77 0-5 used a reinforcer Gain - .7 .25 inappropriately It is apparent that the improvements on the Rein- forcement Difference variable were primarily due to in- creased use of appropriate reinforcers. The multivariate analysis previously presented in Table 5 revealed interactions other than treatment which yield significance. These will now be examined. The first of these was treatment by kind effect. Table 8 shows the univariate test of treatment by kind which indicates the measure responsible for significance. The variable Inappropriate Technique yields signi- ficance. Table 9 shows the means on this variable for the four sub-groups. This interaction is graphically portrayed in Figure 2. 55 TABLE 8 Treatment x Kind Interaction Post Hoc Test on Univariates Measures MS F P less than Reinf. Diff. 16.9 .310 .5816 Reinf./Oppor. Ratio .018 .185 .669 Approp. Tech. 1.60 .986 .328 InapprOp. Tech. 16.90 9.284 .005* \ *Significant at the .05 level. TABLE 9 Means on Inappropriate Technique for Treatment by Kind Interaction Treatment I Treatment II Younger .50 -.60 Kind Older -.70 +.70 56 Younger Older T1 T2 Figure 2 Treatment x Kind Interaction for Variable: Inappropriate Technique 57 As this figure shows, attendants working with older children, and trained by Treatment I used fewer Inappro— priate Techniques than when working with younger children, while Treatment II attendants working with younger children used fewer Inappropriate Techniques than when working with older children. No predictions were made as to this con- trast and the writer can offer no reasonable hypothesis for this finding. The other interaction showing significance is that of Treatment x Kind x Shift. Table 10 shows the univariate test results revealing the measure generating this signi- ficance. The variable Reinforcement Difference yields significance. Table 11 shows the mean gain by cells for this variable. TABLE 10 Treatment x Kind x Shift Interaction Post Hoc on Univariates Measures MS F P less than Reinf. Diff. 445.537 8.1707 .0075* Reinf./Oppor. Ratio .0107 0.1093 .7431 ApprOp. Tech. 1.8375 1.133 .2952' InapprOp. Tech. 4.5375 2.493 .1243 *Significant at the .05 level. 58 TABLE 11 Variable: Reinforcement Difference Treatment I Treatment II Shift Age i i R i i i n n Pre Post Gain Pre Post Gain Young 6 10.83 14.50 3.67 6 9.00 4.00 -5.00 Day Old 6 9.00 11.50 2.50 6 2.33 9.66 7.33 Young 4 3.00 3.50 0.50 4 2.75 7.50 4.75 Afternoon Old 4 10.25 21.75 11.50 4 1.00 3.00 2.00 Totals 20 8.60 12.85 4.25 20 4.15 6.20 2.05 There is a mean increase in the number of reinforce- ments given in a correct manner by attendants in both treat- ment groups. This increase is significantly greater for Treatment I. The table also shows that attendants trained by Treatment I and working on the day shift do better than those working with older children. Attendants trained by the same treatment but working on the afternoon shift tend to do considerably better with older children. With attend- ants trained by Treatment II an Opposite effect is observed. Figure 3 displays this interaction. No predictions were made as to the occurrence of this contrast and, again, no hypothesis is offered for the finding. The use of prime activity time as an additional observation period was discussed previously. In order to determine whether this time provided an Opportunity for in- creased interaction by attendants from both treatment Score on Reinforcement Difference 12 +11 -11 12 59 Day Shift Afternoon Shift T1 T2 T1 T2 Younger Older Younger Older I Figure 3 Treatment x Kind x Shift Interaction for the Variable: Reinforcement Difference 60 groups, a comparison of the post-test mean measures used in the other analysis was made with post-test means using only the "prime" activity time as one of the three Obser- vation periods. Table 12 shows these comparisons. TABLE 12 Comparison of Post-Test Means for the Combined Groups Using Different Activity Intervals SE 5? Variables ETA ETAl Reinf. Diff. 9.53 11.47 Reinf./Oppor. Ratio .48 .50 Appro. Tech. .78 1.38 Inappro. Tech. 1.20 1.07 Global Rating 1.10 1.65 Teach. Intent 2.00 1.50 With the exception of the variable, Intent to Teach, the means which include only prime activity time as the activity segment of the observation period are higher than those that do not uniformly include prime activity time. An exception to this is on the variable Inappropri- ate Technique where a lower score is indicative of Success. Accepting these differences as reliable, this means that the ward situation in which the attendant was observed played a part in determining whether or not his learning would be applied. Distribution_of Global and Intent to Teach Ratings A finer analysis of the results for the two rating scales, Global Rating and Intent to Teach, provides 61 another estimate of the change in attendant behavior. The means for these two scales do not provide meaningful in- formation for this purpose since the scores utilized in the other analyses are sums of the ratings for the three per- iods rather than averages. While this approach was appro- priate for the statistical Operations, the resulting means do not correspond to the original rating scale values. For a clearer picture, a distribution of the actual ratings is provided in Table 13. The striking fact to be observed in this table is that the modal rating is consistently zero. Appreciable use of behavior modification techniques, and appreciable evidence of intent to teach are evident for only a very few individuals. The scores on all of the measures for each indi- vidual in the study are presented in Appendix A. Correlation of the Observation Variables and the Global RatIfig To what extent did the experimenter's global im- pression of the attendants' facility in the use of behavior modification techniques correSpond to the more objective measures of this behavior? This question was answered by a determination of the multiple and partial correlations for these relationships. The partial correlations for each of the variables and the Global Rating are as follows: Reinforcement Difference .30; Reinforcement/Opportunities 62 comes on ucoucH mcflumm Hanoao N H H m N H N H m N O H H H m H H H H m N H N H H N H H H H N H N N H N m v m v o v N H H H H m m v H N H VH ON 0H 5H mH 0H HH MH MH m 0H VH mH ON 0H mH 0H 0H mH mH mH VH 0H NH o 4 e m < e m a .H. m a e m 4 e m .a a m a .H. m a e m Hon 0.8 How 98 umom mum umom mum 958mm HH UCQEMQHB H UCGEMOHB HH “cmsmmHB H HGflSMwHB qsouo ucosumoue No comma ou ucmucH can mcflumm Hmnoao Mom mscfiumm umoenumom use umoeuoum mo cowusowuumwo MH mqmdfi 63 Ratio -.16; ApprOpriate Reinforcement .37; and Inappro- priate Technique -.48. The multiple R was determined to be .71. It is apparent that there is a significant and appreciable relationship between the Global Rating and the objective measures, although there is also a sizable amount of variance in the Global Rating unaccounted for. In spite of the latter fact, the Global Rating did not prove to have greater power than the other variable to discriminate between the treatment groups, as demonstrated earlier. CHAPTER IV SUMMARY AND CONCLUSIONS Because of the change in emphasis of treatment pro- grams in facilities for the mentally retarded from custo- dial to habilitative, new methods of training the insti- tutional attendants charged with the care of these children need to be developed. One mode of treatment with the severely mentally retarded which has demonstrated results in bringing about needed behavioral change is that of behavior modification. Administrators whose responsibility it is to train these attendants have need to know what method of training will bring about the greatest amount of competence in the use of these behavior modification techniques. Gardner (1970) in training attendants in behavior modification has used an attendant simulating the role of a retarded individual while another attendant is directed by a trainer in the use of behavior modification technique. He evaluated the acquisition of behavior modification skills through the use of another simulation experience 64 65 plus a paper and pencil test. However, the most frequently used format is that which utilizes a classroom or tradi- tional program of lectures, demonstrations, and paper and pencil assignments. This method has been evaluated by the use of another form of simulation: video tape presenta- tions of patient behavior for which the attendant is re- quired to supply a treatment strategy. This study utilized still another method involving the training of the attendant on the ward while he pursues his regular assignment with children. The trainers in this method employ behavior modification to shape the at- tendant's learning of the skills. Two important considerations evolve from this dis- cussion: Do the methods of evaluation used in the past effectively measure how competent the attendant will be in the use of these techniques on the ward during his inter- actions with children? Which method will best provide him with behavior modification skills? A study was undertaken which sought to determine which of two methods provided greater application of learning when on-the—ward evaluation was used as the cri- terion. An experimental training program, Treatment I, was compared with a more traditional lecture-discussion- 1aboratory approach. The experimental method involved On- the-ward training. 66 Two groups of 20 institutional attendants were trained with one or the other method, each using language development as a training vehicle. The effectiveness of training was evaluated by the recording of the incidence of behaviors on an observational recording and rating form during three periods when attend- ants were performing their ward duties with children. Measurement was taken both before and after training. It was hypothesized that attendants trained in the personal interaction method would show equal or greater competence in the use of behavior modification than those trained with the traditional method. The hypotheses were tested by the use of multi- variate analysis of variance on the pre-test and post-test gain scores. The correlation between the experimenter's Global Rating of the attendants' facility in the use of behavior modification techniques and the more objective measures on the instrument was also determined. Results of the multivariate analysis revealed no significant main effect for treatment. The obtained means were generally higher for the post-test than for the pre- test scores, although the gains were small. Thus, the hypotheses of the study are supported, in that the attend- ants trained on the ward did no worse than those trained by the traditional method. In fact, the trend, although not significant, was for the Treatment I group to show 67 greater gains. However, the failure of either group to Show much improvement in their use of behavior modification weakens any conclusions concerning the relative merits of the two training methods. Conclusions and Implications for Current Institutional Practice and Researcfi The finding that there was little improvement for either treatment group as evidenced by both the low numerical scores as well as the preponderance of ratings of zero on the Intent to Teach and Behavior Global Rating scales was unexpected. Based on these results, one might be prone to criticize the rating as not being sensitive in picking up the behaviors which did occur. It was the experimenter's impression that the instrument did discrimi- nate over a wide range of behaviors. The fact that rater agreement is very high substantiates that feeling. The presence of experimenter bias seems to have played a mini- mal role since there was no significant improvement shown by either group as a result of treatment. One possible explanation for the failure to find improvement would be that the experimenter, who did all of the rating, may have become more stringent in his criteria for success. This seems an unlikely possibility, since it is more logical to assume that the experimenter, eager to secure positive results, would be inclined to become more lenient in his criteria and subjective judgments. 68 The question arises then, since other researchers have found that attendants are capable of learning behavior modification, why the results of training in this study are not obvious. The most obvious answer is that this study did not attempt to measure learning only, but also competence in, and the tendency to use, the learned technique during everyday interaction with children on the ward. If it is assumed, on the basis of previous studies, that the attendants were able to learn, and did learn be- havior modification skills, then the results of this study indicate that the learned skills were not applied. These assumptions were supported also by the trainers, all of whom were impressed by the success of their methods of instruction as reflected in the feedback they received from the attendants. What then might be the forces which militate against or act in constraint of the attendants' utili- zation of these newly acquired techniques? The traditional role of the attendant has been that of custodian or caretaker. Therefore, in the institution's reward system he has been positively or negatively rein- forced not for what a child has learned, or the degree to which his behavior has been changed, but rather for how well the ward and the child are kept clean and safe from hazard. In other institutions it has been found that even after successful training programs, attendants' behavior 69 does not change as the newly-acquired skills are not re- inforced by the administration. Keith (1971) and Goffman (1961) describe the in- stitutionalization of staff in residential facilities. One factor which promotes the continuance of this process is the power which the attendant has over the resident and his desire to maintain that controlling position. Bateman and Dunham (1949) cite an excerpt from an early community study of a mental hOSpital which substantiates the findings of Goffman. "The chief aim of the attendant culture is to bring about the control of the patients." If the dependency needs of the resident are reinforced by the succorance needs of the attendants, he will have great difficulty assuming the role of teacher or habilitator especially when it means a loss of his own power. Another reason behind the attendants' not assuming this role is related to his role expectation: He is not a teacher, his colleagues do not look on him as a teacher and he is not paid as a teacher. The institution's culture serves to maintain this view. Other factors militating against the attendants' displaying their skills might be lack Of opportunity due to conflicting demands on their time and effort. This is borne out by the fact that the post-test scores containing prime activity time were higher on five of the six vari— ables for both treatment groups indicating that time free 70 of encumbrances such as housekeeping details could be used effectively with the children. It is reasonable to hypo- thesize, therefore, that observation times more conducive to training interactions would have yielded greater oppor- tunities for attendants to apply their new learning. The scheduled activity time in both the morning and early afternoon is a high interaction time. As the reader will recall, it was not used since a comparable observation period was not available on the afternoon shift. Implications for Research One obvious implication for research is that since the observation times chosen for the study were seriously impinged upon by activities other than changing the child's behavior, the observation time in future studies should be changed so as to provide an Opportunity for the attendant to display his skills. Admittedly, the afternoon shift would then be excluded since it is highly probable that the situation during the late hours of the day is similar in most institutions. The staff ratio is much lower, the children are exhausted by 6:00 P.M., and lastly, no child in this society is expected to be in a structured learning situation for 12 hours per day. The children in institu- tions, like children everywhere, need time out. In order to bring about changes in the attendants' ward behavior, the environment in which he works and the philOSOphy which underlies his relationship to the children 71 must be based on one in which learning is the fulcrum of ward activities. It is somewhat irrelevant to carry out elaborate studies to evaluate alternate methods of train- ing attendants to teach and modify behavior, if the op- portunity to apply what is learned is continually subverted by the environment. Gardner (1971) in his in-service programs assumes that the skills of the attendant are developed in propor- tion to the "accurate and relevant feedback" which they generate. If a reinforcement system following training is not incorporated in the managerial function, training efforts may be largely wasted. They are wasted also if the attendants' duties, patient load, and personal attitudes are not conducive to an environment other than custodial. There are no pat solutions for breaking the chain of institutionalization which binds many attendants. Wolfensberger (1971) describes factors which maintain the institutionalization of any facility. Among these are low expectation levels, large groupings of individuals, reduced autonomy of the residents and high regimentation of schedules, rules and practices. With a reduction in these factors thereby bringing about a change in the total environment, it may be possible for a teaching environment to emerge. REFERENCES REFERENCES Ashbaugh, L. L. An evaluation of an attendant training program based on principles of behavior modification. Unpublished doctoral dissertation, The Pennsylvania State University, 1971. Bateman, J., & Dunham, H. The state mental hospital as a specialized community experience. American Journal of Psychiatry, 1948-49, 105, 46. Bricker, W. A., Morgan, D., & Grabowski, J. Token rein- forcement of attendants who work with low-functioning children. In H. C. Haywood (Ed.), Abstracts of Peabody studies in mental retardatioanl965-l968,l968, 3(2). Burke, D., & Rowland, M. An in-service technique to teach ward attendants how to give language development train- ing to institutionalized retardates. Paper presented at the Annual Meeting, Michigan Speech and Hearing Association, Lansing, Michigan, October 1971. Butterfield, E. C. Basic facts about public residential facilities for the mentally retarded. In R. Kugel and W. Wolfensberger (Eds.), New patterns of residential services for the mentally retarded. Washington: U.S. Government Printing Office, 1969. Gardner, James M. Differential effectiveness of two methods for teaching behavior modification techniques to institutional attendants. Paper presented at the 93rd annual meeting of the American Association on Mental Deficiency, Washington, D.C., 1970. Gardner, J. M., & Selinger, 5. Trends in learning research with the mentally retarded. American Journal of Mental Deficiency, 1970 (in press). Gardner, J. M., Brust, Donna J., & Watson, Luke S. A scale to measure skill in applying behavior modification techniques to the mentally retarded. American Journal of Mental Deficiency, 1970, 24(5), 633-636. 72 73 Gardner, J. M. Innovations in the delivery of psychological services in an institution. American Psychologist, l97l,55(2), 211-214. Gardner, J. M., & Giampa, F. The Attendant Behavior Checklist measuring on the ward behavior of insti- tutional attendants. Amerigan Journal of Mental Deficiency, 1971, 15(5), 617-622. Goffman, E. Asylums. New York: Doubleday, 1961. Heber, R. F. A manual on terminology and classification in mental retardation. (2nd ed.) Monograph Supple- ment to the AmefIEan Journal of Mental Deficiency. Willimantic Connecticut, American Association on Mental Deficiency, 1961. Keith, K. D. Analysis of institutional staff behavior. Mental Retardation, 1972, 55(1), 44—45. Mattos, R. L. An investigation of the effects of attendant training in the use of behavior modification techniques on attendants' interaction with institution- alized mentally retarded children. Unpublished doctoral dissertation, The University of Oregon, 1966. Parker, G. O. Attendant-nurses for the mentally deficient: Some evidence. American Journal of Mental Deficiency, 1951, 55, 326-336. Roos, P. Changing roles of the residential institution. Mental Retardation, 1966, 4(2), 4-6. Tarjan, G. The role of residential care--past, present, and future. Mental Retardation, 1966, 5(6), 4-8. Wolfensberger, W. The origin and nature of our insti- tutional models. In R. Kugel and W. Wolfensberger (Eds.), New patterns of residential services for the mentally retarded. Washington: U.S. Government PrInting Office, 1969. Wolfensberger, W. Will there always be an institution. I: The impact of epidemiological trends. Mental Retardation, 1971, 5(5), 14-20. APPENDICES APPENDIX A RAW SCORES ON SIX VARIABLES 74 TABLE 14 Raw Scores on Six Variables l 2 3 4 5 6 Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post ESTA ESTA ESTA ESTA ESTA ESTA ESTA ESTA ESTA ESTA ESTA ESTA Treatment II Attendant 1 12 3 0 0 l 0 3 l 0 0 2 0 2 8 15 6 2 l l 4 4 0 l 2 2 3 O 1 O 1 1 O 2 2 O O 0 0 4 8 5 0 O l l 2 O 0 O O O 5 29 3 O 3 2 O l 2 1 O 3 O 6 6 3 3 0 O 0 1 l O 0 0 O 7 6 19 l 4 2 3 2 l 2 0 2 2 8 O 3 0 O 1 O 1 O 0 O O 0 9 4 7 l 2 l O l l O O O 0 lo 3 ll 0 4 l 0 1 0 O O 0 l 11 S 21 O 5 0 1 1 1 O l O 5 12 3 5 0 O 0 0 1 l O 0 1 O 13 4 24 4 4 3 S l l 0 5 3 6 l4 3 7 0 O 0 0 0 3 0 O 0 2 15 6 9 3 0 0 0 0 1 O 0 0 O 16 0 1 O 0 0 o O 1 O O O O 17 4 8 O O 0 O 1 2 O 0 0 l 18 O 2 O O 0 0 O 1 O O 0 O 19 0 7 O O 0 0 1 O O O O O 20 0 0 O O l 0 l O O 0 l 0 Treatment I Attendant 21 15 22 0 O O 1 2 1 O 4 7 5 22 4 8 1 4 2 2 1 1 0 l 1 1 23 7 9 0 0 l 1 1 O 0 0 O 0 24 22 28 12 6 3 3 1 2 5 6 5 7 25 19 8 3 O 2 O 1 2 O O 3 1 26 14 2S 0 3 2 2 2 1 1 4 1 6 27 0 6 0 0 0 1 2 l 3 O l 0 28 O 9 0 8 1 0 0 3 1 0 2 l 29 1 3 0 O 0 l 2 1 0 0 4 o 30 14 8 3 4 1 1 0 5 6 O S o 31 10 18 0 1 2 0 2 1 O O 1 O 32 0 l3 0 6 O 1 1 O O O 0 l 33 19 22 2 4 o 0 1 1 2 2 O 6 34 11 7 O O 0 2 0 1 O O 0 0 35 18 13 2 0 3 0 3 1 l 1 6 1 36 0 9 0 2 0 0 0 1 0 O O 0 37 7 18 O 2 2 0 1 O O l O l 38 8 21 O 3 2 0 2 O 0 3 l 4 39 2O 35 S 3 l 3 2 0 l 7 2 7 4O 17 24 6 3 O 4 2 l 2 8 2 5 LEGEND: 1. Total nunber of times 5 is recorded as having used an appropriate reinforcer. 2. Total number of times 5 is recorded as having used an inappropriate reinforcer. 3. Number of behavior modification techniques of which 5 gives evidence in his behavior. 4. Number of categories of behavior displayed by S representing violations of behavior modification principles or their inappropriate use. 5. Average rating for S on his competence in the use of behavior modification principles. 6. Average for each 5 on the strength of his intention to engage in teaching. APPENDIX B BARF BEHAVIOR ANALYSIS RATING FORM 75 APPENDIX B BARF BEHAVIOR ANALYSIS RATING FORM Observation Number 1 2 3 4 Date 1. Unused opportunity Aézfvity E_TFA Al Observer Name Ward Appropriate T 12. Ignores Behavior 2. Verbal 13. Conscious Change of Reinforcer 3. Gestural 14. Time Out Appropriate 4. Physical 15. Time In Appropriate 5. Tangible Reward 16. Fading l7. Prompt 6. Physical-Verbal 18. Modeling 7. Verbal—Tangible 19. Successive Approximations 8. Physical-Tangible 20. Cues 9. For Non—Performance 21. Verbal Punitive, Derogatory 10. Too Delayed 22. Physical Punitive 11. For Incompatible 23. Inappropriate Target 24. Non-Recognition of Inefficient Reinforcer 25. Reinforcer too Satiating 26. Non-Recognition of Opportunity to Change Behavior 27. Non-Recognition Difficult of a Task 28. Not Breaking Task Down 29. Too Many Expectations/Unit B E O ’5 10 Incont o I 2 3 4 5 Beh. Mod. 0 I’ 2 3 4 5 Intent: Intent to Teach Rating Scale Beh. Mod.: Behavior Modification Rating Scale (Global Rating) APPENDIX C LANGUAGE OBSERVATION FORM 7 6 APPENDIX c LANGUAGE OBSERVATION FORM This scale provides criteria for the assessment of the functioning level of communication and a sequence for teaching Needs Much Help Needs Some Help Satisfactory: Part of Routine I. Listens, and reacts with large muscles: patty-cake, bye-bye, rocks a doll to music, rolls balls to music, marching, running, tapping, hand clapping. Listens, and reacts with large muscle activity on verbal command stop, wait, look, sit down, come here, don't touch Listens, shakes head "yes” or ”no," responds to own name (not verbal) Listens, identifies source of sound Listens, identifies source of sound and locates it Listens, and responds by indicating: parts of body, own possessions, boys and girls Listens, and responds to simple directions "show me” (common object) or "put your finger on5_J give me ___J the ball is in the box" Listens, to familiar animal or mechanical sounds and vocalizes in repetition Listens, and mimics words to name common Objects (not pictures) 10. Names objects without opportunity to mimic ll. Rhythmic responses to percussions or music 12. Says words appropriately and spon- taneously (uses words) 13. Qualifies nouns (little box, red ball) 14. Uses overt verbs 15. Combines words to convey idea or need; frames a good question APPENDIX D LANGUAGE SCALE ASSESSMENT FORM 77 mpuoz "cowumu cOHmmoumxm IHHeoo> pooz Inwue>euoa Hw>ma Mono: pcooom .m Hmusumoo HHflflflHw IHHmoo> GOHmmoumxm commouocH pooz .NwH>Huom Ho>oq umuwm .m uouoz commonocH moxwa moxHHmHo moxHq moxHHmHo moxaq moxHHmH muses umocooom HMOHmmnm umanB I nemesm oncommom ucmpcouua HmcnoucH I msHsEHum . m0H>¢mmm 92¢: SOHumNHHmoo> commouocH >uw>Huo¢ uouoz commouocH lawsuouev xmnouaesav Aamsma>o .m mcHHmom mchouqu mconoq nowhno I msHsEHum coHueuHHmoo> commonocH oncomnmm unaccouum huH>Huo¢ mono: commenocH Aflmsuomav Hwnouaosav Aamsma>v .a open usez mcHHomm mchoumHA mconoq .e useocouud I msHosHpm sumo Osmz momHouoxm pcoEQOHo>oo omcmm .H .m open osmz moH>cmmm moaozIHmOmzmm .N "Ho>oq moHchua sumo mamz m.ucmoomuu¢ one: osmz m.ucoonom ZMOh BZMmeMmmd MH¢Um mU¢DOZ£H 78 mmcommom unaccouum mmcomwom unaccouud mmcommom unmpcouud AEHUMQHO> Uuoommv mmmdmmm UHmweommmBm A Aouxnm mpuo3 memos OHomHHNm ImIzV mpuo3 meadow .WHQMHHNm 039 oco Monuo HH< OHooHHmm oco Hosopr, 14 ‘ moosom mo cOHuouHEH comm Doc noouo some on coHumHou puma moon .kuumm Noon cocoa cH mound Noon N m>oz Hm9mH> soccer 1 ‘I wound seem Suez .m powwoo nonuo ou ooHDMHOH evenno Downoo L SH pomnoo o>oz o>oz mnmcoHumum cocoa! . muomfino nude .a muGOEQ/OE H0902 MO GOHUMUHEH .m MOH>¢mmm UHmoum Amocooomv th VH MH NH HH OH m m h m m v m N H V ‘ Aucmacmuum mo monocou newest saddens uca>eo coeucmuua .v mOH>€mmm UZHQZHBB4 mmcommmm ucmocouud 79 an: cons 30m ems: ouooz omcommom endoscope Hmnnm>IImsHs2aum m memos cowumsuomcH m AGOHumuHHeoo> o>Hm ou memos mchD .a % mcommom uouoz coHumsnomcH I. Hensummw O>Hw ou mousUmom mchs .0 m mcHocmumuopcD gamma!“ mopMOHocH umnu mouos mcwocmumuopcs .m m mcHoceumuoocD coHuoEnomcH o>Hm,ueou ® mOOMOHch mousumoo mcwpcmumuoocs .a u“ A a cOHumsnomcH mcHoH>oum .m cH mchH>oum .m mOH>¢mBzH 3 .w monoz HOQMH a "coHuMNHHmoo> op mpuo3 mcha .0 % mcommom nouoz HoomH on M. Hmuzumou mousumom ochD .U a UOHoomq HooMH H mH pens were memos w mODMOHch mcwpcmumuoocs .m ® OOHoomq HoomH u. mH umn3 page mousumow m mODMOHocH mcaocmumnooco .m ucm>m noonno ucmocouud muco>m .muOOhoo .OHmoomIImsHsEHum mcHHooeq .5 mOH>€mmm BUdB ICHIGAN STnTE UNIV. LIBRARIES IIIIIIIIIII III III III III III III II III IIIIIIIIIIII IIIIII 31293102083007