Good Behavior Game

Categories

Categories associated with best practice:

  • Illegal Drugs Icon 3
  • Children (ages 6-12) icon
  • Alcohol Abuse Icon 1
  • Best Practices
  • Education and literacy
  • Elementary School Icon 1
  • Healthy child development
  • Mental Health Icon 2
  • Personal health practices and coping skills
  • Preventing Violence Icon 1
  • Tobacco Use Icon 1

Determinants of Health: Personal health practices and coping skills, Education, literacy and life-long learning, Social environments, Healthy child development

Overview

The Good Behavior Game (GBG) is a classroom-based behaviour management strategy for elementary school that teachers use along with a school’s standard instructional curricula. GBG uses a classroom-wide game format using teams and rewards to socialize children to the role of student. GBG is used to reduce aggressive, disruptive classroom behaviour, which is a risk factor for adolescent and adult illicit drug abuse, alcohol abuse, cigarette smoking, antisocial personality disorder (ASPD), and violent and criminal behaviour.

In GBG classrooms, the teacher assigns all children to teams, balanced with regard to gender; aggressive, disruptive behaviour; and shy, socially isolated behaviour. Basic classroom rules of student behaviour are posted and reviewed. When GBG is played, each team is rewarded if team members commit a total of four or fewer infractions of the classroom rules during game periods. During the first weeks of the intervention, GBG is played three times a week for 10 minutes each time during periods of the day when the classroom environment is less structured and the students are working independently of the teacher. Game periods are increased in length and frequency at regular intervals; by mid-year the game may be played every day. Initially, the teacher announces the start of a game period and gives rewards at the conclusion of the game. Later, the teacher defers rewards until the end of the school day or week. Over time, GBG is played at different times of the day, during different activities, and in different locations, so the game evolves from being highly predictable in timing and occurrence with immediate reinforcement to being unpredictable, with delayed reinforcement so that children learn that good behaviour is expected at all times and in all places.

GBG was tested in first- and second-grade classrooms in Baltimore beginning in the 1985–1986 school year. Follow-up at ages 19–21 found significantly lower rates of drug and alcohol use disorders, regular smoking, antisocial personality disorder, delinquency and incarceration for violent crimes, suicide ideation, and use of school-based services among students who had played the GBG. Several replications with shorter follow-up periods have provided similar short-term results. GBG has been found to be most effective for males with higher levels of aggressive, disruptive behaviour.

Primary Source Document

Kellam SG, Mackenzie ACL, Hendricks Brown C, Poduska JM, Wang W, Hanno P, Wilcox HC, The Good Behavior Game and the future of prevention and treatment Addiction Science & Clinical Practice: 73-84.

Contact information of developer(s) and/or implementer(s)

Sheppard G. Kellam, M.D., Retired

Johns Hopkins Bloomberg School of Public Health

615 North Wolfe Street

Baltimore, MD 21205

(410) 614-0680

Intervention Focus

Intervention Goal / ObjectiveLevel(s) TargetedEquity Focus
To prevent adolescent and adult illicit drug abuse, alcohol abuse, cigarette smoking, antisocial personality disorder (ASPD), and violent and criminal behaviour
  • Individual level
People living in conditions of disadvantage are explicitly stated to be a target population of the intervention.
To socialize children to the role of student
  • Individual level
People living in conditions of disadvantage are not explicitly stated to be a target population of the intervention.
To reduce aggressive, disruptive classroom behaviour
  • Individual level
  • Interpersonal level
People living in conditions of disadvantage are not explicitly stated to be a target population of the intervention.

Health Issue(s) that is/are addressed by the Intervention

Health Promotion

  • Mental Health
  • Other

Risk Reduction

  • Prevent/reduce alcohol use/abuse
  • Prevent/reduce tobacco use
  • Prevent/reduce illegal drug use/abuse
  • Prevent/reduce misuse of medications
  • Prevent violence
  • Other

Specific Activities of the Intervention

  • Group process/program

Priority/Target Population for Intervention Delivery

Life Stage

  • Children (age 6-12 years)

Settings

Educational Settings

  • Elementary school

Community Setting

    N/A

    Outcomes

    Outcomes and Impact Chart
    Level of ImpactDescription of OutcomeEquity Focus
    Individual LevelFor the first cohort, there was a reduction in lifetime alcohol abuse/dependence disorders in the GBG group compared with controls (13% for GBG versus 20% for internal GBG controls, p = 0.08; 29% for all controls, p = 0.03, unadjusted for baseline or classroom effects). The overall effects of the GBG appeared to be similar for both males and females. In the second cohort, no significant impact of the GBG on alcohol abuse/ dependence disorders was found.Outcomes are reported for people living in conditions of disadvantage, and are not compared to people living in more advantaged conditions.
    Individual Level

    Males in the first cohort in GBG first-grade classrooms who had been highly aggressive, disruptive were more likely to graduate from high school (75%) compared with the highly aggressive, disruptive males in control classrooms (20% in internal controls, p = 0.03; 40% in all controls, p = 0.04, unadjusted for classroom and baseline effects). There were no significant differences found in the second cohort for high school graduation rates.

    Outcomes are reported for people living in conditions of disadvantage, and are not compared to people living in more advantaged conditions.
    Individual LevelYoung adults in the first cohort who were in GBG first-grade classrooms had lower levels of lifetime drug abuse/dependence disorders compared with those in the control classrooms (12% for GBG versus 21% for internal controls, p = 0.04; 19% for all controls, p = 0.03, unadjusted for child baseline and classroom effects). This reduction was significant for males; 19% of GBG males reported drug abuse/dependence disorders compared with 38% of internal male controls (p = 0.01) and 30% of all control males (p = 0.05, unadjusted for child baseline and classroom effects). The GBG did not significantly reduce drug dependence/ abuse disorders in females alone. We found a much greater GBG effect for males who were rated as highly aggressive, disruptive by teachers in first grade. For these highly aggressive, disruptive males, the rate of drug dependence/abuse disorders was 29% in the GBG groups, 83% in the internal controls (p = 0.02), and 68% in the combined control group (p = 0.02, unadjusted for child baseline and classroom effects). In the second cohort, there was also a reduction in drug abuse/dependence disorders for GBG males (19%) compared to internal GBG controls (32%, p = 0.10, not controlling for classroom effects). The overall benefit of the GBG compared to low aggressive, disruptive internal controls was significant at p = 0.016 but not compared to all controls (24%, p = 0.41).Outcomes are reported for people living in conditions of disadvantage, and are not compared to people living in more advantaged conditions.
    Individual Level

    Overall rates of ASPD in the first cohort were lower for youth in the GBG groups (17%) than they were for internal controls (25%, p = 0.07) and all controls (25%, p = 0.03, unadjusted for classroom- and child-level effects). The GBG also reduced ASPD among males who were rated as highly aggressive, disruptive in first grade compared with similar high aggressive, disruptive males in the control groups (38% for GBG aggressive, disruptive males versus 80% for internal controls, p = 0.10, and 70% for all aggressive, disruptive male controls, p = 0.05, also unadjusted for classroom- and child-level effects).

    In Cohort 2, the impact of GBG on ASPD was not significant.

    Outcomes are reported for people living in conditions of disadvantage, and are not compared to people living in more advantaged conditions.
    Individual LevelThe rate of regular smoking was lower among those in the GBG classrooms (6%) than it was among those in control classrooms (10% for internal GBG controls, p = 0.15; 14% for all controls, p = 0.002, unadjusted for baseline or classroom effects). This effect was larger among males – 6% of the GBG were regular smokers compared with 19% of male internal controls (p = 0.03) and 20% of all control males (p = 0.004, unadjusted for baseline or classroom effects) – than it was for females – 5% were regular smokers compared to 3% of internal female controls (p = 0.55) and 9% of all female controls (p = 0.20; unadjusted for baseline or classroom effects). These reductions in regular smoking were highly apparent among males who began first grade with high aggressive, disruptive teachers’ ratings; for the high aggressive, disruptive males, none were regular smokers in the GBG groups (0%) compared with 40% of internal controls (p = 0.008) and 25% in the all control groups (p = 0.03, unadjusted for classroom effects). In the second cohort, both GBG males and females showed 30–40% lower rates of regular smoking compared with their counterparts; this reduction was non-significant but consistent across baseline levels of aggressive, disruptive behavior as well as control conditions.Outcomes are reported for people living in conditions of disadvantage, and are not compared to people living in more advantaged conditions.

    Adaptability

    Implementation History

    • Multiple implementations - Different settings/populations/providers - The intervention showed significant adaptability as it has been implemented in different settings or with different populations or by different provider(s). This can include multiple implementations during the same time period. Each implementation of the intervention must have been substantially the same and must have demonstrated positive results for the primary objectives of the intervention.

    Expertise Required for Implementation within the Context of the Intervention

    • Specialized skills training provided as part of the Intervention - The intervention does not require individuals or groups with highly specialized training, but requires that individuals or groups be trained as part of the implementation of the intervention.

    Are there supports available for implementation

    Yes.

    Are there resources and/or products associated with the interventions

    Yes.

    American Institutes for Research (AIR) works with district and school staff to build a team to effectively implement the Good Behavior Game. AIR staff identify and clearly articulate each role—for the principal, district and school administration, teachers, and support staff—so the Game is fully supported and its components are incorporated into the school and district culture and climate.

    There are two strands of professional development: one for teachers and one for local GBG coaches and trainers. Teachers receive one year of GBG training that consists of group-based sessions enhanced by the support of a coach in the classroom. Local coaches are trained in situ over one year as they work with teachers. Teachers receive 3 days of on-site training: a 2-day Initial GBG Training and a 1-day Booster Session. Training for program Coaches includes attending the sessions above plus a one day Initial Coach Training. Coaches also receive at least two on-site Implementation Audits and up to 100 hours of technical assistance by phone and email in their first year.