This meta-analysis (27 studies, N= 1702) examined (a) the effects of homework assignments on treatment outcome and (b) the relationship between homework compliance and therapy outcome. Results of the primary meta-analyses indicated a weighted mean effect size (r) of .36 for homework effects and .22 for homework compliance. A moderator analysis (chosen on a priori grounds) was also conducted by partitioning the sample of effect size estimations first according to the sample problem type, according to the type of homework activity administered, and according to the source and time of homework compliance assessment. We hope that the focus of future research will now be diverted from general questions of the benefit of including homework in therapy, to more specific questions regarding the relative effectiveness of different types of homework assignments for different client problems.
Available at: http://works.bepress.com/fdeane/53/