Meta-analytic study on treatment effectiveness for problem behaviors with individuals who have mental retardation.

Meta-analysis of 482 empirical studies on treatment of problem behaviors of individuals with mental retardation was conducted. A metric of treatment effectiveness was computed for 1,451 comparisons between baselines and treatments, 34 topographies of problem behavior, and 64 treatment procedures. Analysis of variance with percentage of nonoverlapping data as the dependent variable and comparison as the basic unit of analysis revealed that treatment of externally destructive behaviors had significantly lower mean percentage of nonoverlapping data scores than did treatment of socially disruptive and internally maladaptive behaviors. Response contingent procedures were significantly more effective than were other procedures. No significant interactions were found. Results of a stepwise regression showed that only performing a functional analysis made a significant contribution. These results may lead to more objective assignment of treatment procedures to problem behaviors.