Studies on the effects of retention follow this design: Researchers identify a group of students who have been retained in a grade. Then they find another group, matched in relevant characteristics to the retained group, which had instead been promoted to the next grade. Usually, the groups are matched on achievement test scores so that they wereequally low before the retention or promotion occured. Then, the achievement test scores of the two group are compared at the end of the following year or at the end of the next grade. Two meta-analyses have been conducted, one on the studiesup to 1989and the other on studies conducted between 1990 and 1999. In the first meta-analyses, Holmes found out that the overall difference in mean achievement across 63 studies was- 31. This means that, one or two years later, the group that have been eligible to be retainedbut was instead promoted had a better level of average achievement than the peer group which had been retained. 54 of the 63 studies favored the promoted group. Such a one-sided division of studiesis rarely observed in education research. In the second meta-analyses, Jimerson integrated 18 studies of the effects of the retention conducted between 1990-1999. Only about 10% of the studies favored the retained group. The overall effect size was- 39. This means roughly that after the retention or promotion the retained group members were about three school months behind their initially comparable peers. Any early advantage that had been observed for the retained group disappeared overtime.
*Alexander, K., Entwisle, D., & Kabbani, N. (1999). Grade retention, social promotion, and “third way” alternatives. Paper presented at the National Invitational Conference hosted by the Laboratory for Student Success at Temple University for Research in Human Development and Education. Alexandria, Virginia (November 29–December 1).
Foreign Literature
Amidst an era emphasizing