A meta-analysis of the effectiveness of educational treatment programs for emotionally disturbed students

The purpose of this study was to identify and compare the effects of educational treatment programs for emotionally disturbed students. In addition, the study explored whether programmatic or demographic variables interacted with the treatment programs to affect the outcome measures.;Two hundred twenty studies were integrated using the meta-analysis technique. The analysis involved more than 2,300 research subjects and yielded 235 measures of effect size, which were calculated on outcome areas of general achievement, mathematics achievement, reading achievement, and social skills. Variables investigated were age, diagnosis, SES, IQ, length of program, program setting, and program facilitator.;It was hypothesized that (1) there are significant differences in the overall effects of educational treatment programs for emotionally disturbed students, (2) there are differences in the effects of these treatment programs upon the specific outcome measures of mathematics achievement, reading achievement, general achievement, and social skills, (3) program elements and student demographics interact with the treatment programs to significantly affect the outcomes.;Data analysis involved: (a) descriptive statistics for the data base, (b) analysis of variance to test the effectiveness of the treatment programs, and (c) regression analysis where the effect sizes were regressed onto the variables.;No statistically significant differences in the total effect sizes of the eleven treatment programs were found. There were statistically significant differences among the treatment programs upon specific outcome measures. Programmatic and demographic variables were found to significantly affect outcome areas when interacting with structured-behavioral treatment.;It was concluded that no one treatment program was superior to another in overall effects. However, a short term structured behavioral treatment program in the lower elementary grades would yield the greatest gains in both academic and social skills achievement. Further study is needed to evaluate the variable relationships