ROBERT PEAT : ABA : Case Study : The Good Behaviour Game
[Skip to Navigation | Skip Past Navigation]


The purpose of this study was to reduce the amount of 'talking out of turn' in a second-year, secondary school class. The procedure entailed splitting the class of ten boys into two teams. The teams were required not to engage in the target behaviour above a specified maximum limit in order to "win" a reinforcer. During the Baseline period, the frequency of talking out of turn was recorded for five separate classes; French, Geography, Science, Maths and Religious Education. The intervention phase produced a colossal reduction in the target behaviour. With the return to baseline, the frequency of the target behaviour increased again. These findings support the effectiveness of the good behaviour game at reducing undesirable behaviours in a group setting with older children.


Merrett and Wheldall (1978) surveyed teachers' opinions in the West Midlands borough of the United Kingdom as to what were the most troublesome classroom behaviours. Results indicated that 'talking-out-of-turn' was the most troublesome behaviour and that it accounted for one third of misbehaviour in classrooms. When weighed against problems such as violent behaviour or illiteracy, talking out of turn may not appear to be a serious problem (Axelrod, 1977). However, unsolicited talking in the classroom interferes with the work habits of co-operative students, wastes teacher time, causes aggravation to both pupils and teacher and quiet pupils are often ignored. If disruptive behaviour is allowed to continue without successful intervention it can reach levels where completion of academic assignments are impeded and teaching time is spent reprimanding students.

Many teachers who are not behaviourally trained would advocate a "get tougher in the classroom" strategy to regain respect, control and authority. A study by Van Houten, Nau, Mckenzie-Keating, Sameoto and Colavecchia (1982) exemplifies this approach. They found that when verbal reprimands were delivered with eye-contact and a firm grasp of students' shoulders, a reduction in disruptive behaviour was observed. However the use of such tactics when dealing with an angry six-foot, 16-year- old may be ill-advised. An alternative is to concentrate on group contingencies because of their practicality and the effective manner in which they allow the teacher to gain direct control of the class (Hall, Lund, & Jackson, 1971). Group contingencies also eliminate differential treatment of individuals and are thus both cost and time effective, a view echoed by, Wrobel and Michaelis (1968) and by Litow and Pumroy (1975).

In this paper we examine the effectiveness of The good behaviour game which was pioneered by Barrish, Saunders and Wolf (1969). This is an inter-dependent- group oriented contingency system (Sulzer-Azaroff & Mayer, 1991). In this type of system, receiving reinforcement is contingent upon a specified level of group performance (e.g., the frequency of talking out of turn remaining below ten instances). Interventions based on group contingencies such as the good behaviour game automatically harness the valuable reinforcement of peer attention. Numerous studies have demonstrated that peer attention is a powerful reinforcer for disruptive behaviour (e.g., Northup, Broussard, Jones, George, Vollmer & Herring, 1995). Since it's conception many modified versions of the good behaviour game have been implemented with resounding success (e.g., Fischbein and Wasik (1981) used it in a library setting). Saigh and Umar (1983) demonstrated the game's cross-cultural validity when they used it in an elementary school in The Sudan. Research into the "normal" classroom has, for the most part, however, focused on the primary/junior schools (Barrish, Saunders & Wolf, 1969; Saigh & Umar, 1983; Harris & Sherman, 1973; Merrett & Wheldall, 1978; Fischbein & Wasik, 1981) with relatively few studies concentrating on secondary schools (Mc Namara & Harrop, 1979). In this study, a multiple baseline across settings was used to examine the effectiveness of the good behaviour game with older children in a secondary school. The secondary school differs greatly from the primary school in that there is much less interaction between teachers and pupils as classes are continually moving from classroom to classroom, subject to subject, teacher to teacher.


Participants and Setting

The project was conducted in an all male secondary school in a large town in Northern Ireland. The participants were a bottom-stream second-year class that consisted of 10 boys, all of whom were 14 years of age.

Target behaviour

Two days were spent observing the class prior to the data collection phase in order to select the target behaviours. Initially two target behaviours had been selected but this was reduced to one given the very high frequency of talking out of turn and the physical limitations that this imposed on recording by a single researcher.

The dependent variable selected was the frequency of 'talking out of turn behaviour' in five classes, French, Geography, Science, Mathematics and Religious Education, under the supervision of five different teachers. The following behaviours were classed as 'talking out of turn':

  • Any verbalisations that are made in class that were not requested by the
  • teacher.
  • Any derogatory remarks, jeering or laughter at another pupil's expense.
  • Any shouting at the teacher or complaining about the task allocated.
  • Attempting to initiate conversation with another pupil.
  • Responding to contact from another pupil.
  • Verbally encouraging the misbehaviour of other pupils (for example
  • suggesting ways to annoy others).


The data were recorded three days a week over a seven-week period.

Baseline (Condition A)

All the observations were made by the researcher who sat at the back of the class. During the baseline phase, the participants were not informed as to the nature of the researcher's work.

Intervention (Condition B)

Before introducing of the intervention the teachers were consulted about any worries that they had about being observed in class. A multiple baseline across settings was used. During the intervention phase, the 10 boys were divided into two equal teams of five and the rules of the game were explained. A poster delineated the rules of the game was displayed clearly at the front of the room during the intervention sessions only (see Table 1). The participants were then told that the team with the fewest "X's" would be deemed the "winners", but if both teams kept the "X's" below 10, then both teams would receive their prizes.

Each individual's preferred reinforcer was noted at the beginning of the game. This procedure was followed for all of the single classes when the intervention was used. All of the participants were asked if they fully understood the rules of the game and once all the participants were clear, the researcher made the following announcement:

The game will start now and end when the bell for the end of class sounds.

Prizes were distributed by the researcher at the end of the class period.

After several sessions using the game (in different classes), a new contingency was added for Geography and Science. During double classes, playing the game in the second period was contingent on playing the game successfully in the first period, without receiving any reinforcer. Thus the boys only received one set of reinforcers for two sessions of the game.

Return to Baseline

At the end of the intervention there was a return to Basline conditions. This was unusual because it was unnecessary given the rationale for a multiple baseline design. This decision also carrried with it some ethical concerns given the effectiveness of the procedure. However, the teachers at the school were unskilled in the implementation of the procedure and it was at their request that baseline conditions were resumed.


The participants were consulted as to the types of reinforcers that they would like and their suitability was ratified by the teachers. Initially only a choice of "fun size" chocolate bars were available. However this was insufficient for one participant and consequently a new reinforcer in the form of "football stickers" was made available. Throughout the course of the game, reinforcers varied and each participant was allowed to nominate the reinforcer of his choice.

Inter-observer Reliability

Independent observations were made by a classroom aid. The interobserver reliability was calculated by dividing the smaller obtained frequency by the larger and then divided by 100 to obtain a percentage. During the baselines the percentage of agreement varied averaged 89%. During the intervention phases, the percentage of agreement was 100%.

Results and Discussion

Figure 1 illustrates the frequency of talking out of turn for five different classes throughout the study. During Baseline 1, a very high frequency of the target behaviour was observed across all of the classes. In the first Session the frequency of talking out of turn ranged from 291 in French class, 229 in Geography class, 376 in Science class, 304 in Maths class to 450 in Religious Education class. The trend for relatively high frequencies of talking out of turn across classes remained relatively stable throughout the first baseline period.

During the intervention there was a dramatic reduction in the amount of talking out of turn in each class. The frequencies for both teams were combined to give a total for the sessions. The number of instances ranged from 4 in French class, 7 in Geography class, 5 in Science class, 18 in Maths class. During double classes of Science and Geography, in the first period when no reinforcer was available, the levels of talking out of turn were 7 in Geography and 10 in Science. During the second period, when reinforcement was available, the levels of talking out of turn were 2 in Geography and 9 in Science.

The intervention was not implemented in Religious Education because a decreasing trend was observed. From Sessions 1 through to 5, the levels of talking out of turn dropped from 450 to 130. The lowest frequency was 49 in Session 6 and between Sessions 7 to 10, the levels again increased slightly above 150. During Sessions 9 and 10 a substitute teacher took the class but there was no significant increase or decrease in the levels of talking out of turn. Following the return to baseline, the levels of talking out of turn increased again.

The objective of this study was to examine the feasibility of using the good behaviour game to reduce the frequency of talking out of turn across a variety of classes in a secondary school. Results showed it to be an effective and easily implemented method of reducing the target behaviour. A surprising result was in relation to individual performances. One student in particular spoke out-of-turn more than the others consistently across all classes. Once the intervention was used, his talking was consistently zero! Another reflection on the power of the procedure occurred during double periods. When the opportunity to earn reinforcement in the second class of double periods of Geography and Science was made contingent on "winning" the game in the first class, both teams "won" the game in the first period and were allowed to play the game in the second period, at the end of which they received their reinforcers. In conclusion, we have clear indication that the good behaviour game can be used successfully with older children.


Axelrod, S. (1977). Behaviour modification for the classroom teacher. New York: McGraw-Hill Inc.

Barrish, H.H., Saunders, M. & Wolf, M.M. (1969). Good behaviour game; effects of individual contingencies for group consequences on disruptive behaviour in a classroom. Journal of Applied Behaviour Analysis, 2, 119-124.

Bushell, D. Jr., Wrobel, P.A. & Michaelis, M.L. (1968). Applying group contingencies to classroom study of pre-school children. Journal of Applied Behaviour Analysis, 1, 55-61.

Fishbein, J.E. & Wasik, B.H. (1981). Effects of the good behaviour game on disruptive library behaviour. Journal of Applied Behaviour Analysis, 14, 89-93.

Hall, R.V., Lund, D. & Jackson, D. (1971). Effects of teacher attention on study behaviour. Journal of Applied Behaviour Analysis, 4, 141-149.

Harris, V.W. & Sherman, J.A. (1973). Use and analysis of the 'good behaviour game' to reduce disruptive classroom behaviour. Journal of Applied Behaviour Analysis, 6, 405-417.

Litow, L. & Pumroy, D.K. (1975). A brief review of classroom group-oriented contingencies. Journal of Applied Behaviour Analysis, 8, 405-417.

McNamara, E., & Harrop, A. (1979). Behaviour modification in secondary school : A cautionary tale. Occasional Papers of the Division of East Child Psychology of the B.P.S., 1, 139-150.

Merrett, F & Wheldall, F. (1978). Playing the game : A behavioural approach to classroom management in the junior school. Educational Review, Vol. 30, No. 1, 41-50.

Northup, J., Broussard, C., Jones, K., George, T., Vollmer, T.R. & Herring, M. (1995). The differential effects of teacher and peer attention on disruptive classroom behaviour of three children with a diagnosis of ADHD. Journal of Applied Behaviour Analysis , 28, 227-228.

Saigh, P.A. & Umar, A.M. (1983). The effects of a good behaviour game on the disruptive behaviour of Sudanese elementary school students. Journal of Applied Behaviour Analysis, 16, 339-344.

Sulzer-Azaroff, B. & Mayer, G.R. (1991). Behaviour analysis for lasting change. New York: Holt, Reinhart & Winston Inc.

Van Houten, R., Nau, P.A., Mckenzie-Keating, S.E., Sameoto, D. & Colavecchia, B. (1992). An Analysis of Some Variables Influencing The Effectiveness Of Reprimands. Journal of Applied Behaviour Analysis, 15, 65-83.

Date Posted: 26/06/2007

Associated Downloads

You can right-click upon the filename and 'Save Target As..' to download the file to your computer, or left-click upon the filename to view it now.

All of the downloads listed below will open in a new window:

You are here: home > aba > case studies

Simple Steps

News from peat

Stamppp II Launch for Simple Steps 23.09.2013

Science and the Treatment of Autism: A Multimedia Package for Parents and Professionals

Read more about Stamppp II Launch for Simple Steps »

PEAT to Empower Young People with Autism who are at risk of Educational Exclusion 16.09.2013

PEAT's Big Lottery funded project now up and running.

Read more about PEAT to Empower Young People with Autism who are at risk of Educational Exclusion »

Proactive Inclusion through Preventing Exclusion 10.07.2013

PEAT is collaborating with Queen's University Belfast and the University of Ulster in a new project to support young people with an autistic spectrum disorder disengaged from education through funding from the Big Lottery Fund.

Read more about Proactive Inclusion through Preventing Exclusion »


ask peat

You can ask PEAT a question by visiting our Contact Us page and completing the online form »