cooperative behavior of emotionally disturbed boys as a function of contingent application of social...

5
J. Behav. Thu. & Exp. Phychint. Vol 4, pp. 33-37. Pcrgamon Press, 1973. Printed in GreatBritain. COOPERATIVE BEHAVIOR OF EMOTIONALLY DISTURBED BOYS AS A FUNCTION OF CONTINGENT APPLICATION OF SOCIAL APPROVAL JOSEPH KLOTZ and DAVID W. SHANTZ* Oakland University, Rochester, Michigan Summary-Twelve emotionally disturbed boys were randomly assigned to experimental and control groups across 12 activity group sessions. The male activity leader verbally approved all instances of cooperative behavior exhibited during a model building task by experimental subjects across the 12 sessions. An equal amount of approval was given to control subjects in each session, but only for instances of individual achievement. Two-factor analyses of variance indicated (a) a significant 0, < 0.01) increase in cooperation across sessions for experimental but not control subjects, (b) no change in level of non-cooperation across sessions for either group. One week after the 12th activity session dyads of experimental and control Ss partici- pated in a new task (Madsen Board task) to assess generalization of cooperative behavior. T-tests indicated experimental Ss cooperated significantly more @ < 0.01) than control Ss. IT HAS been amply demonstrated that the behavior of exceptional children can be signifi- cantly modified by appropriate application of reinforcement principles (Demeyer and Ferster, 1962). The kinds of behaviors modified have varied from such relatively simple motor responses as wearing glasses and climbing behavior, to such complex behaviors as reading, language skills, and a variety of social inter- action patterns (Allen er al., 1965; Hall and Broden, 1967; Risley and Wolfe, 1967; Staats and Butterfield, 1965; Wolfe, Risley and Mees, 1964). The present study explored further the role contingent reinforcement can play in modifying exceptional children’s social behavior. First, it evaluated the extent to which verbal approval could be used to increase the level of co- operation emotionally disturbed boys displayed during a series of group activity sessions. Second, it examined the extent to which in- creases in cooperative behavior generalized to other tasks and settings. Finally, it assessed the validity of the assumption (Kanfer and Phillips, 1970; Ulrick and Favel, 1970) that reinforcing a desirable response (cooperation) incompatible with an undesirable response (non-cooperation) will result in a decrease in the frequency of the undesirable response. Subjects METHOD All 12 adolescent boys attending the Child- ren’s Orthogenic School served as subjects (SS).~ Mean age and Wechsler Intelligence Scale for Children (Full Scale IQ) for these boys were 14.50 yr (S.D.=15 months) and 95 (S.D.=17) respectively. Ss were randomly assigned to experimental and control groups. Description of treatment conditions The study was conducted during the course of a twice-weekly, 40 min group activity program. In each session the six boys in the experimental group were seated together at one table in the activity room, with the six control boys at another. All Ss were given toy models of their choice to work on during these sessions, with only one tube of glue provided at each table. During the session the group activity leader (the *Requests for reprints should be sent to David W. Shantz, Department of Psychology, Oakland University, Rochester, Michigan 48063. TDiagnostic classification assigned to these boys included brain damage, behavior disorders, learning disabilities, neuroses, and border line psychoses. 33

Upload: joseph-klotz

Post on 17-Oct-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

J. Behav. Thu. & Exp. Phychint. Vol 4, pp. 33-37. Pcrgamon Press, 1973. Printed in Great Britain.

COOPERATIVE BEHAVIOR OF EMOTIONALLY DISTURBED BOYS AS A FUNCTION OF CONTINGENT APPLICATION OF

SOCIAL APPROVAL

JOSEPH KLOTZ and DAVID W. SHANTZ*

Oakland University, Rochester, Michigan

Summary-Twelve emotionally disturbed boys were randomly assigned to experimental and control groups across 12 activity group sessions. The male activity leader verbally approved all instances of cooperative behavior exhibited during a model building task by experimental subjects across the 12 sessions. An equal amount of approval was given to control subjects in each session, but only for instances of individual achievement. Two-factor analyses of variance indicated (a) a significant 0, < 0.01) increase in cooperation across sessions for experimental but not control subjects, (b) no change in level of non-cooperation across sessions for either group. One week after the 12th activity session dyads of experimental and control Ss partici- pated in a new task (Madsen Board task) to assess generalization of cooperative behavior. T-tests indicated experimental Ss cooperated significantly more @ < 0.01) than control Ss.

IT HAS been amply demonstrated that the behavior of exceptional children can be signifi- cantly modified by appropriate application of reinforcement principles (Demeyer and Ferster, 1962). The kinds of behaviors modified have varied from such relatively simple motor responses as wearing glasses and climbing behavior, to such complex behaviors as reading, language skills, and a variety of social inter- action patterns (Allen er al., 1965; Hall and Broden, 1967; Risley and Wolfe, 1967; Staats and Butterfield, 1965; Wolfe, Risley and Mees, 1964).

The present study explored further the role contingent reinforcement can play in modifying exceptional children’s social behavior. First, it evaluated the extent to which verbal approval could be used to increase the level of co- operation emotionally disturbed boys displayed during a series of group activity sessions. Second, it examined the extent to which in- creases in cooperative behavior generalized to other tasks and settings. Finally, it assessed the validity of the assumption (Kanfer and Phillips, 1970; Ulrick and Favel, 1970) that reinforcing a

desirable response (cooperation) incompatible with an undesirable response (non-cooperation) will result in a decrease in the frequency of the undesirable response.

Subjects METHOD

All 12 adolescent boys attending the Child- ren’s Orthogenic School served as subjects (SS).~ Mean age and Wechsler Intelligence Scale for Children (Full Scale IQ) for these boys were 14.50 yr (S.D.=15 months) and 95 (S.D.=17) respectively. Ss were randomly assigned to experimental and control groups.

Description of treatment conditions

The study was conducted during the course of a twice-weekly, 40 min group activity program. In each session the six boys in the experimental group were seated together at one table in the activity room, with the six control boys at another. All Ss were given toy models of their choice to work on during these sessions, with only one tube of glue provided at each table. During the session the group activity leader (the

*Requests for reprints should be sent to David W. Shantz, Department of Psychology, Oakland University, Rochester, Michigan 48063. TDiagnostic classification assigned to these boys included brain damage, behavior disorders, learning disabilities, neuroses, and border line psychoses.

33

senior author) interacted ditTerentially with the number of the sessions through ;I one-\vay two groups of boys as follows: mirror and recorded cooperative and non-

I. E.~perinvv~ttrl group. During each of the cooperative responses for all the boys. Second, sessions the leader smiled and gave verbal the leader was told this check on his objectivity approval (“Good”; “Keep it up”; “That’s the and reliability would be made periodically during way”) to each experimental boy every time he the course of the study, and that he would not interacted cooperatively with a tablemate re- be aware of exactly when the checks were garding use of the tube of glue. A cooperative occuring. The fact that the leader was aware of interchange was defined as one which did not the need for a high level of agreement between involve threat, insult or derogation, either his scoring and that of the uninformed outside verbally or physically. During the course of the observer insured a continuous effort to bc as study the experimental group received verbal objective as possible in recording cooperative approval only for these cooperative responses. and non-cooperative responses.

2. Cotltrol group. Control boys received as much verbal approval in each activity session as Description ?/!I’ the getwralixtion tad-

the experimental group. However, they were 1. ‘The Madsen Board. The extent to which reinforced only for responses reflecting in- increases in cooperative behavior elicited during dividual achievement (working neatly, ac- the activity sessions generalized to another task curately etc.)” Control boys did not receive any and setting was evaluated by means of perfor- approval from the leader for cooperative mance on a modified Madsen board task behavior at any time during the course of the (Madsen, 1967). The Madsen board is a 30 ,: 12 activity sessions. 36 in. plywood board which can be covered by

3. Dependent wriahks rr.s.se.r.red durilg activit)’ a blank sheet of paper on which S makes his .ses.rions. Two response measures were recorded responses. In this study each sheet of papel for each boy in both the experimental and contained three rectangular target zones 4 ;: 12 control groups during the 12 sessions: (a) the in.: one in the center of the board, and one number of cooperative interchanges involving along each of the right and left hand edges of use of the glue tube, and (b) the number of non- the board directly opposite the center zone tar- cooperative interchanges initiated regarding use get. Four strings running through eyelets in of the glue tube. A non-cooperative response was each of the four corners of the board were defined as any interchange over the glue which attached to a wooden spool in which a pencil was involved threat, insult or derogation, of either inserted. Two boys seated at each end of the a verbal or physical nature. board could, by coordinating their manipula-

Because of personnel limitations it was tion of the two strings in their possession, move necessary that the leader unobtrusively record the pencil in such a way as to place pencil the number of cooperative and non-cooperative marks wherever they wished on the surface of responses exhibited by each of the boys during the blank paper. the activity sessions. This procedure, of course, 2. Procedures ad imtructions. One week after raises the issue of the objectivity of the leader’s completion of the activity group sessions dyads recording. In order to enhance objectivity of of Ss were administered the modified Madsen the scorer and evaluate rater-reliability, the board task by the leader. Three dyads of experi- following steps were taken. First, a staff member mental group Ss were formed by randomly familiar with the scoring procedure but LIP pairing the six experimental group Ss. Three I’;imiliar with the nature of lhc study ob\ervcd ;I dyads of control grc’up .Ss were similarly formed.

“l’his control condi(ion wab deGgncd to asbc~ 111~’ extent to which change5 in level of cooperation in the experimental group was a function of contingent approval for cooperative behavior, rather than merely being a by-product of gaining the leader’s approval pe,. .sc.

3.1 JOSEPH KLOTZ and DAVID W. SHANTY

COOPERATlVE BEHAVIOR OF EMOTIONALLY DISTURBED BOYS 35

Each dyad was brought into the testing room To evaluate this prediction a two-factor analysis and seated at opposite ends of the Madsen of variance (Treatment Group X Sessions) was Board. The operation of the Board was ex- conducted on mean cooperative scores (Winer, plained and demonstrated by the leader, and the 1962). Results indicate significant differences in Ss instructed as follows: “You two are a team. level of cooperation as a function of Treatments You have to pull and release these strings and (F = 7.97; # =l/lO; p < 0.05); Sessions (F == move the marker to a side zone and back to the 2.72; df -11/l 10; 1~ < 0.05; and the interaction center zone, putting a mark in each zone. You between treatment and sessions (F = 4.21 ; are to do this as many times as you can within 3 cv =1 l/l 10; p < 0.01). min. 1 will tell you when to start and when your The means associated with the Treatment X time is up.” Sessions interaction are presented in Fig. 1.

Ss were then told that the team making the most number of side-center marking sequences would get a prize. They were also told they 9 .- l -• Experimental group

could talk freely with one another about how - *-A Control group

to go about doing well on the task. No reinforce- 8 -

ments were given for cooperative behavior dur- ing the generalizat!on task. 7-

E - 3. Dependent variable in the Madsen Board ;

task. A cooperative score for each dyad was the number of marks made in the target zones

5 6 L 5

during the 3 min period divided by the total z _ number of the marks made anywhere on the sheet. This scoring procedure controlled for - differences between dyads in speed of res- ponding.

RESULTS

Reliubility of scoring I -

A Pearson-product-moment correlation co- I 1 I I I I I 11 1 1 1 , 2 3 4 5 6 7 8 9 IO II 12

efficient was computed between the leader’s Sessions

scoring of the 12 boys’ cooperative and non- cooperative responses during the session and

FIG. 1. Mean number of cooperative responses function of treatment group and session.

as a

the scoring of the same session by an observer. The correlation between scores was 0.98 for cooperative responses, and 0.96 for non-coopera- Differences between the means associated with tive responses. These results indicate that the leader used the scoring procedures in a reliable fashion.

Eject of treatment oil cooperutioll during activity

group sessions

It was anticipated that the level of cooperative behavior exhibited by Ss in the experimental group would increase significantly across the group sessions and eventually exceed the level of cooperation exhibited by the control group.

this interaction were assessed with the Tukey (A) post mortem test (Winer, 1962). Results indicated that from session 9 onwards Ss in the experi- mental group made significantly more coopera- tive responses than in session I, and significantly more cooperative responses than control Ss (all comparisons 11 ~0.01). Experimental Ss showed an essentially linear increase across sessions in their level of cooperation. Control Ss, in contrast, failed to show any significant change in cooperation.

36 JOSEPH KLOTZ and DAVID W. SHANTZ

Efect of treatment on non-cooperation during activity group sessions

The effect the treatments had on non-co- cooperative behavior across activity sessions was evaluated by a two-factor analysis of variance (Treatment X sessions). The means involved in the analysis are presented in Fig. 2.

F _* Experimental group

A--r Control group

I”““““” I 2 3 4 5 6 7 8 9 IO II 12'

Sessions

FIG. 2. Mean number of non-cooperative responses as a function of treatment group and session.

It can be seen that there were no differences in level of non-cooperation, either between the two groups of Ss, or within either group across sessions.

These data, then, indicate that although social approval significantly increased the level of cooperative behavior exhibited by the experi- mental Ss, this increase was not accompanied by a reciprocal decrease in level of non-co- operation.

Cooperation in the Madsen task

It was anticipated that Experimental Ss would not only behave more cooperatively than Control Ss during model building in the activity sessions, but that they would generalize this behavior to other tasks and settings (i.e. the

Madsen Board). In order to assess the predic- tion that Experimental Ss would cooperate more than Control Ss on the Madsen Task, a t-test for small samples was conducted on the mean cooperative scores of the Experimental and Control dyads (Downie and Heath, 1959). The three Experimental dyads placed an average of 70 per cent of their total marks in the target zones as compared with an average of 53 per cent for the Control dyads. These percentages were significantly different (p < O*Ol), thus confirming the expectation that there would be generalization of cooperative behavior shown during activity sessions.

DISCUSSION

These data clearly demonstrate that a male adult activity leader can, by contingent applica- tion of social approval, significantly increase the frequency of a particular class of cooperative behaviors in a group of boys who exhibit a variety of behavioral problems. Further, there is significant generalization of the effect to a different task setting involving a different class of cooperative responses.

The fact that the level of non-cooperative behavior did not decrease as the level of coopera- tive behavior increased is not consistent with traditional behavior modification expectations. Perhaps a closer examination of the principle of “reinforcing incompatible behavior” would clarify matters. Although cooperation is generally considered to be “incompatible” with non-cooperation, in the sense that both kinds of responses cannot be made simultaneous in a given situation, they obviously can be given successively across a series of interpersonal interactions as was the case in the present study. To the extent that this can happen, one should not expect that reinforcing one class of responses should necessarily result in a diminution of the other class of responses. One might, however, expect a reciprocal reduction in the frequency of class B behavior if the S infers that the ap- proval for class A behavior, together with silence in response to class B behavior, implies disap- proval of class B behavior. Crandel (1963) has

demonstrated that silence under these conditions is in fact interpreted by a random sample of elementary school children as disapproval. Approval of cooperative behavior, then, may have failed to significantly reduce levels of non-cooperation in the present study either because these boys did not make the same inferences normal children would have made in this situation, or because the inferred disapproval was not as negative an outcome for these boys as it would be for normal children, and thus had a weaker inhibiting effect. Finally, it must be recognized that non-cooperation, although not reinforced by E’s approval, may well have been reinforced by other outcomes potentially avail- able in the situation (e.g. peer approval, success- fully obtaining the glue tube from another child etc.). It would appear then that one should ex- pect reinforcement of a desirable response to result in a diminution of the incompatible undesirable response only under those conditions that result in S’s concluding that the undesirable response either will no longer lead to positive outcomes, or will lead to a negative outcome which exceeds the anticipated positive outcome.

The fact that the increased cooperativeness of the experimental Ss generalized to a quite dif- ferent task setting is also somewhat surprising, given the frequency with which improvement in problem children’s behavior fails to generalize. Mischel (1968) has suggested that the generali- zation of behavior can be expected to occur only when the individual anticipates that a res- ponse class will continue to be reinforced in the new setting. Thus the presence during the generalization task of an adult who the experi- mental Ss knew approved of cooperative be- havior may have played a significant role in eliciting the expectation that cooperative res- ponses would continue to be reinforced in the new task setting. To the extent that this is true, it emphasizes the importance in therapy both of generating the expectation that appropriate

behavior patterns will be reinforced in the setting in which one child must operate, and of taking steps to ensure that this indeed happens suf- ficiently frequently to maintain patterns of appropriate behavior.

REFERENCES

ALLEN K. E., HART B. M., BUELL J. S., HARRIS F. R. and WOLF M. M. (1965) Effect of social reinforcement of isolate behavior of a nursery school child, Case Studies it? Behavior ModiJcation (Edited by ULLMAN L. P. and KRASNER L.), pp. 307-312, Holt, Rinehart & Winston, New York.

AZRIN N. H. and LINDSLEY 0. R. (1956) The reinforce- ment of cooperation between children, J. abnorm. sot. Psychol. 52, 10&102.

CRANDALL V. C. (1963) Reinforcement effect of adult reactions and non-reactions on children’s achievement expectations, Child Development 34, 335-354.

DEMEYER M. K. and FERSTER C. B. (1962) Teaching new social behavior to schizophrenic children, J. Am. Acad. Child Psychiat. 1, 4431161.

DOWNIE N. M. and HEATH R. W. (1959) Basic Statistical Methods, Harper & Bras, New York.

HALL R. V. and BRODEN M. (1967) Behavior changes in brain-injured children through social reinforcement, J. Exceptional Child 5, 463479.

KANFER, F. H. and PHILLIPS J. S. (1970) Learning Foundations of Behavior Therapy, pp. 279-280, Wiley, New York.

MADSEN M. C. (1967) Cooperative and competitive motivation in three Mexican subcultures, Psycho!. Rep. 20, 1307-1320.

MISCHEL W. (1968) Personality and Assessment, pp. 176-178, Wiley, New York.

RISLEY T. and WOLF M. (1967) Establishing functional speech in echolalic children, Behav. Res. & Therapy 5, 73-88.

STAAT A. W. and BUTTERFIELD W. H. (1965) Treatment of non-reading in a culturally deprived iuvenile delinquent : ai application of -reinfbrcemeit prin- ciples, Child Development 36, 925-942.

ULRICH R. E. and FAVELL J. E. (1970) Human aggression, Behavior Modification in C&icai Psycholog; (Edited by NEURINGER C. and MICHAEL J. L.). v. 128. Avvleton- Century-Crofts, New York. ” ’ ’ ’

WINER B. J. (1962) Statistical Principles in Experimental Design, McGraw-Hill, New York.

WOLF M.. RISLEY T. and MEES H. (1964) Application of operant conditioning procedures to the behavior problems of an autistic child, Behav. Res. & Therapy 1, 305-312.

COOPERATIVE BEHAVIOR OF EMOTIONALLY DISTURBED BOYS 37

(Received 12 April 1972)