Key points are not available for this paper at this time.
In this study data from all studies identified as meeting the selection criteria were mathematically integrated to determine the effect of grade-level retention on elementary and/or junior high school pupils. When each effect size calculated was treated equally, a grand mean effect size of −.37 was obtained indicating that, on the average, promoted children scored .37 standard deviation units higher than retained children on the various outcome measures. When the effect sizes within each study were first averaged so that each study could be given equal weight, a grand mean of −.34 was obtained. By using the effect sizes from only those studies in which the promoted and nonpromoted pupils had been matched, a grand mean of −.38 was calculated. The high degree of consistency in these measures lends credibility to the validity of these findings. In addition to the grand means, effects sizes were calculated on various dependent variable measures, including academic achievement (further subdivided into various areas), personal adjustment (which included self-concept, social adjustment, and emotional adjustment), and attitude toward school, behavior, and attendance. In all cases, the outcomes for promoted pupils were more positive than for retained pupils.
Holmes et al. (Fri,) studied this question.