UC San Diego

The current study was an updated meta-analysis of manuscripts since the year 2000 examining the effects of homework compliance on treatment outcome. A total of 23 studies encompassing 2,183 subjects were included. Results indicated a significant relationship between homework compliance and treatment outcome suggesting a small to medium effect (r = .26; 95% CI = .19–.33). Moderator analyses were conducted to determine the differential effect size of homework on treatment outcome by target symptoms (e.g., depression; anxiety), source of homework rating (e.g., client; therapist), timing of homework rating (e.g., retroactive vs. contemporaneous), and type of homework rating (e.g., Likert; total homeworks completed). Results indicated that effect sizes were robust across target symptoms, but differed by source of homework rating, timing of homework rating, and type of homework rating. Specifically, studies utilizing combined client and therapist ratings of compliance had significantly higher mean effect size relative to those using therapist only assessments and those using objective assessments. Further, studies that rated the percentage of homeworks completed had a significantly lower mean effect size compared to studies using Likert ratings, and retroactive assessments had higher effect size than contemporaneous assessments.