Assessing Heterogeneity of Effect Size on Sample Size in a Meta- Analysis of Validity Studies
Download full article pdf . Doi: 10.5901/mjss.2012.03.01.219
Abstract This study examined the impact of heterogeneity of effect size on the sample size of some validity studies. Thirty (30) empirical studies were selected on the basis of empirical status and relevance; their results both quantitative and qualitative were recorded, coded and analyzed. The findings revealed that the differences in the results of the 30 selected empirical studies were a function of the sample size on which the studies were based. Mean Fisher ( Zr =0.393 , WZr=0.42347). Characteristics peculiar to each study did not affect the result of the study. Weighted Mean Fisher by sample size ( WZr) =0.398 with associated r=0.375) was an equivalent of the Mean Fisher. The results of the empirical studies were found to be significantly different in terms of their effect sizes (÷2=1444.97 p<0,05,29df). The heterogeneity of the effect size was a result of the heterogeneity of the sample size. The Weighted Mean Fisher was bigger than the Mean Fisher (WZr=0.489 ,Zr=0.393). The difference in the result was a function of differences in the sample size and not as a result of the study characteristics peculiar to each study.
Keywords: meta - analysis, effect size, heterogeneity, Sample size.