Good Behavior Game
(A 1st-2nd grade classroom management strategy for decreasing aggressive/disruptive behavior)
Randomized controlled trials show major reductions in students’ subsequent substance abuse, and behavioral and mental health disorders.
Description of the Intervention
The Good Behavior Game is a 1st – 2nd grade classroom management strategy for decreasing aggressive/disruptive student behavior. In some cases, it is implemented in both 1st and 2nd grade (see study 1 below); in other cases, it is implemented only in 1st grade (see study 2 below).
The teacher initially divides her class into three teams, each with an equal proportion of boys and girls, and aggressive versus well-behaved children. The teacher then clearly describes (i) good student behaviors (e.g. working quietly on assigned tasks), and (ii) disruptive behaviors (e.g. talking out of turn, fighting etc.). Teams receive check marks on a posted chart when one of their members exhibits a disruptive behavior. Teams that receive few check marks are rewarded at the end of each game period, and consistent winners are again rewarded at the end of the week – at first, with tangible rewards (e.g. classroom activities, stickers, erasers), and later during the year with more abstract ones (e.g. gold stars). The Game is played for ten minutes three times a week early in the year, gradually extended in time, and eventually incorporated into the whole day and entire week.
Teachers receive approximately 40 hours of training in the proper implementation of the Game, and supportive mentoring during the school year. In some studies of the Good Behavior Game’s effectiveness, the Game has been supplemented with other classroom strategies (see Study 2 below).
The Good Behavior Game cost approximately $500 per student per year to implement in the studies below (including the cost of training and mentoring the teachers, and monitoring their implementation). It may cost less than this when implemented on a larger scale outside a study setting.
Click here for the Good Behavior Game manual.
EVIDENCE OF EFFECTIVENESS
Study 1
Randomized controlled trial of 14 1st grade classrooms (containing 407 students) in six elementary schools from five Baltimore urban areas ranging from very poor to moderate income, and predominantly African-American to predominantly Caucasian. Classrooms and teachers within each school were randomly assigned to a group that participated in the Good Behavior Game, or a control group that did not.
Students in the Good Behavior Game group received the intervention for two years (1st and 2nd grades), staying with the same class of students both years. Students in the control group also stayed with the same class for both years. All teachers in the Good Behavior Game group received 40 hours of training in the Good Behavior Game, and supportive mentoring during the school year; control-group teachers received comparable amounts of training and support in other, unrelated areas. Researchers conducted classroom observations to monitor whether teachers were properly implementing the intervention.
Effects on the Good Behavior Game group at the age 19-21 follow-up (approximately 14 years after random assignment), versus the control group:
- For male students:
- 50% lower rate of lifetime illicit drug abuse/dependence (19% rate for Good Behavior Game males vs. 38% rate for control group males).
- 59% less likely to smoke 10 or more cigarettes per day (7% vs. 17%). This finding was statistically significant at the .10 level, but not the .05 level.
- Approximately 35% lower rate of lifetime alcohol abuse/dependence (the precise rate for Good Behavior Game males versus controls is not reported).
- Suggestive evidence of a sizeable increase in high school graduation rate, and a sizeable decrease in lifetime Major Depressive Disorder and Anti-Social Personality Disorder. However, these effects did not reach statistical significance (possibly because of the small sample size).
- No effect on rate of lifetime Generalized Anxiety Disorder.
- For female students, no effects were found.
- The study found suggestive evidence that the Good Behavior Game had the largest effects on virtually all outcomes for those males rated highly aggressive by their teachers at the beginning of first grade.
In a second stage of the study, the teachers in the Good Behavior Game group were asked to continue using the Good Behavior Game in their classes during the subsequent school year on a separate sample of incoming students. However, the teachers received little mentoring or retraining in the Good Behavior Game during this subsequent year. In this second sample, the effects on male students were smaller (positive but not reaching statistical significance, and roughly half the size of the effects listed above for the first sample). A possible reason for the smaller effects is the absence of continued mentoring and retraining to ensure teachers’ close adherence to the intervention.
Discussion of Study Quality (click here for a glossary of terms)
- The study had a long-term follow-up with low to moderate attrition: 76% of the original sample was interviewed at the age 19-21 follow-up. There was no difference in attrition between the Good Behavior Game group and control group.
- The study measured outcomes using an intention-to-treat analysis.
- Teachers in the control group received an attention placebo (namely, training and mentoring in areas other than classroom management).
- Substance abuse and mental health outcomes were measured through structured phone interviews using a survey instrument with established validity and reliability (the Composite International Diagnostic Interview – University of Michigan Version (CIDI-UM).
- The interviewers were blind as to whether students were assigned to the Good Behavior Game group or the control group.
- This was a multi-site study, administered in typical public school classrooms by regular teachers, thus providing evidence of the Good Behavior Game’s effectiveness in real-world settings.
- Prior to the intervention there were no significant differences between the Good Behavior Game group and the control group.
- Study Limitations: This was a relatively small study (14 classrooms); thus corroboration of the results in other studies (such as study 2 below) is particularly important. Also, the substance abuse and other outcome data were obtained through self-reports and not corroborated by more objective measures (e.g. saliva tests for tobacco use).
References
(Click on linked authors’ names for their contact information)
Kellam, Sheppard G., C. Hendricks Brown, Jeanne Poduska, Nicholas Ialongo, Hanno Petras, Wei Wang, Peter Toyinbo, Holly C. Wilcox, Carla Ford, and Amy Windham. “Effects of a Universal Classroom Behavior Management Program in First and Second Grades on Young Adult Behavioral, Psychiatric, and Social Outcomes.” Drug and Alcohol Dependence, 95S, 2008, S5-S28.
Kellam, Sheppard G. and James C. Anthony. “Targeting Early Antecedents to Prevent Tobacco Smoking: Findings From an Epidemiologically Based Randomized Field Trial.” American Journal of Public Health. Vol. 88, No. 10, October 1998, pp 1490-1495.
Study 2
Randomized controlled trial of 678 students entering 1st grade in nine urban Baltimore public schools. The students were predominantly African-American (87%) and economically disadvantaged (62% received free or reduced-price school lunch).
The schools’ 27 1st grade classrooms were randomly assigned to (1) a group that received the Good Behavior Game plus curriculum enhancements designed to improve students’ reading, math, and critical thinking skills; (2) a group that received the Family-School Partnership (an intervention designed to improve parent-teacher communication and parents’ teaching/parenting skills); or (3) a control group that received neither intervention. Students and teachers were then randomly assigned to the classrooms.
Both interventions were provided only in 1st grade. Teachers in both intervention groups received 60 hours of training prior to implementation. Significant efforts were made to ensure key elements of the two interventions were properly implemented (e.g., frequent meetings between intervention experts and teachers, monitoring of classrooms, etc.).
Of the two interventions, the Good Behavior Game had the largest effects versus the control group; its effects are summarized below. The Family-School Partnership generally had smaller effects (results not summarized here).
Effects on the Good Behavior Game group at the average age 11 follow-up (end of 6th grade), versus the control group:
- Much less likely to have a lifetime conduct disorder, as diagnosed in interviews with the students and their parents (4% of Good Behavior Game students had a disorder, versus 10% of the controls).
- 35% less likely to have been suspended during the previous school year, based on teacher reports (22% versus 34%).
- 29% less likely to have received mental health services during their lifetime (15% versus 21%).
- 36% less likely to have received mental health services in the previous year (23% versus 36%).
- 43% less likely to be in need of mental health services, based on their teachers’ reports (13% versus 23%).
Effects on the Good Behavior Game group at the average age 13 follow-up, versus the control group:
- 28% less likely to have started smoking (34% of Good Behavior Game students had started smoking versus 47% of control group students).
- Substantially less likely to have tried cocaine, crack or heroin (3% versus 7%).
- No effect on whether students had tried alcohol, marijuana, or inhalants.
Discussion of Study Quality (click here for a glossary of terms)
- The study had a long term follow-up with low to moderate attrition: 74% of the original sample completed the age-11 follow-up, and 84% completed the age 13 follow-up. There was no difference in attrition between the Good Behavior Game group and control group.
- The study measured outcomes using an intention-to-treat analysis.
- The study was administered in typical public school classrooms by regular teachers, providing evidence of the Good Behavior Game’s effectiveness in real-world settings.
- Students completed confidential, self-administered computer surveys on substance use to help ensure that researchers’ biases (e.g. as proponents of the intervention) would not influence their outcome measurements.
- Behavioral outcome data were obtained through fully structured interviews using well-established survey instruments. The data obtained from parents and students were supplemented with data obtained from 6th grade teachers and school officials who were blind as to whether students had been assigned to the Good Behavior Game group or the control group.
- Study Limitations: (1) It is unclear which elements of the intervention are critical to effectiveness (e.g. whether the Good Behavior Game without curriculum enhancements would also be effective); and (2) Substance use outcomes were assessed through student self-reports, and not corroborated by more objective measures (e.g. saliva tests for tobacco use).
References
(Click on linked authors’ names for their contact information)
Furr-Holden, C. Debra M., Nicholas S. Ialongo, James C. Anthony, Hanno Petras, and Sheppard G. Kellam. “Developmentally inspired drug prevention: Middle school outcomes in a school-based randomized prevention trial.” Drug and Alcohol Dependence. Issue 73, 2004, pp. 149-158.
Ialongo, Nick, Jeanne Poduska, Lisa Werthamer, and Sheppard Kellam. “The Distal Impact of Two First-Grade Preventive Interventions on Conduct Problems and Disorder in Early Adolescence.” Journal of Emotional and Behavioral Disorders. Volume 9, Issue 2, Fall 2001, pp. 146-160.
Storr, Carla L., Nicholas S. Ialongo, Sheppard G. Kellam, and James C. Anthony. “A randomized controlled trial of two primary school intervention strategies to prevent early onset tobacco smoking.” Drug and Alcohol Dependence. Issue 66, 2002, pp 51-60.
Other Studies
Two other randomized controlled trials of the Good Behavior Game have been carried out, but do not meet this site’s criteria (e.g., not a sufficiently long follow-up period). However, their results are consistent with those of the studies described above.