Effectiveness of school‐based psychotherapy: A meta‐analysis of dissertation research

A meta‐analysis was conducted with 65 school‐based psychotherapy and counseling dissertations over the last 10 years (1998–2008) to assess if a file‐drawer problem (i.e., studies conducted but not published that, as a whole, have different results than studies in the same area published) exists in the school‐based outcome literature. An overall mean effect size of 0.44 was found for 73 treatment interventions. This effect size is comparable to Prout and DeMartino's 1986 meta‐analysis conducted with published school‐based studies and approximately half a standard deviation smaller than Prout and Prout's 1998 meta‐analysis of school‐based intervention studies. A bias does appear to exist but seems to be smaller than the bias found in the general child and adolescent psychotherapy outcome literature. Most of the dissertation studies evaluated group interventions and used a cognitive–behavioral or skills training intervention. Skills training and interventions with elementary‐school students yielded the largest effect sizes. Comparisons of the current study with previous school‐based intervention meta‐analyses are discussed as are suggestions for future research. © 2010 Wiley Periodicals, Inc.