Meta-analysis

This meta-analysis investigated the practical significance of school counselor-led
social emotional learning (SEL) interventions on outcomes for students in Grades 6–12.
The sample includes 28 studies involving 3,794 middle and high school students. The
treatment group was comprised of a total of 2,032 students, who received interventions
led by a certified school counselor. The control/comparison groups were derived from a
sample of 1,762 middle and high school students who did not receive the school
counselor-led intervention. The meta-analysis included a diverse sample of students,
with ethnicity reported as 589 (15.52%) African American, 52 (1.37%) Asian, 1,162
(30.63%) Hispanic, 1,267 (33.39%) Caucasian, 11 (0.28%) Native American, 21 (0.55%)
Pacific Islander, and 177 (4.66%) Multi-racial/Other. Of the studies included in the
meta-analysis, the ethnicities of 412 (18.86%) students were not reported. A total of 12
studies were conducted at the middle school level, 10 at the high school level, and 6
studies reported a mixed setting of Grades 6–12. The sample included almost equal
representation of 1,883 (49.63%) males and 1,847 (48.68%) females, and the genders of
69 (1.82%) students were not reported.
The overall unweighted Cohen’s d effect size (ES) of the school counselor-led
interventions was .312 (95% CI [.173, .452]). The ES for overall cognitive outcomes (d
= 0.380) was slightly larger than for overall effective role functions outcomes (d =0.377)
and affective outcomes (d = 0.356). The smallest ES reported for overall variables was
for behavioral outcomes (d = 0.228). The largest ES for a specific student outcome measure was found for standardized achievement test scores (d = 0.612) (Vernez &
Zimmer, 2007). Moderator analyses were conducted and are explored in the results and
discussion. The results address the current gap in school counseling outcome research by
broadening future research directions for comprehensive school counseling programs
(CSCP) to incorporate SEL initiatives aimed for middle and high school students.

Meta-analysis is a statistical method of combining many individual analyses. This thesis reviews the need for meta-analysis; the many statistical consideration facing the meta-analyst; and some of Hedges' results concerning the combined estimate of effect size with unequal weights from his 1981 and 1982 papers. Unequal weights used to combine estimates of effect size in meta-analysis are derived using the variances given by the large sample, normal approximation of the distribution of Hedges' unbiased estimates of effect sizes. These variances depend on the effect size and the sample sizes of both experimental and control groups. This creates circular definitions and calls for further estimates. This thesis analyzes the limiting normal approximation to derive a variance which is not dependent on effect size, and it provides guidelines for its use.