The Question
What is the efficacy of grade retention as a strategy to address the academic and social–emotional needs of low-achieving students?
The Context
Retention at grade level has been a popular method for providing remediation of low–achieving students throughout the 1900s—and particularly over the last 25 years. Proponents of retention claim that such programs motivate students to work harder in school, allow low–achieving students a year of additional study to master material they previously had difficulty with, and hold both schools and students accountable for achieving academic outcome standards. In addition, retention supporters connect the social promotion of low–achieving students with the low performance of older students, citing this connection as evidence that schools are not being required to address the academic needs of all students. Recently Florida, New York City, and Chicago have used grade retention as a remediation strategy to eliminate the promotion of low–achieving students.
Opponents of retention highlight the high risk of dropout faced by students who are retained at grade and research that shows that retained students do no better academically than students who are promoted. The high social–emotional costs faced by retained students are also a concern to retention opponents, as is the potential that retention punishes students who are learning, but may not be developmentally ready to achieve at the levels mandated by rigid grade–level structures.
Past research generally indicates that opponents of retention are right to be concerned about the harmful effects of grade retention; however, the prima facie logic of retaining low–achieving students, combined with the increased emphasis on standardized measurement and accountability inherent in current accountability systems, suggests that retention issues will continue to drive policy discussions. In an effort to inform these discussions, this ResearchBrief focuses on a study that uses meta–analysis to gather and analyze the body of research on retention.
The Details
Shane Jimerson conducted the study highlighted in this issue of ResearchBrief (see <XREF TARGET="rb_v3n01_study">below</XREF> for full citation). In an effort to condense the body of research on retention, Jimerson systematically reviewed retention research and research reviews published between 1900 and 1989. He also conducted his own meta–analysis of retention research published from 1990–99. Although many of the early studies Jimerson reviewed raised methodological concerns (including the use of inadequate comparison groups, limited analytic focus, or limited time frame), three major research reviews—conducted in 1975, 1984, and 1989, respectively—determined that using retention programs as a remediation strategy was not effective. The 1975 review determined that the body of research was inconclusive; while the 1984 and 1989 meta–analyses found that retained students generally showed lower academic achievement, poorer personal adjustment, lower self–concept, and lower school engagement than their promoted peers.
Because research methodologies continue to evolve and improve, and because previous analyses looked at research through 1989, Jimerson focused his meta–analysis on retention research conducted between 1990 and 1999. Using a database and reference keyword search, the author identified more than 400 studies for potential inclusion in the analysis. However, he limited his review to studies that had been presented in a professional publication (journal or book), focused on the efficacy of retention, included comparison or control groups, and were published between 1990–99.
- How are the comparison and control groups matched?
- In what grade are students retained, and when are the outcomes measured?
- What are the results of academic analyses?
- What are the results of social–emotional and behavioral analyses?
- What conclusions do the authors of each study included in the meta–analysis make?
In most of the studies, academic achievement was measured by standardized, norm–referenced tests. Social–emotional adjustment included a much wider range of variables, including peer competence, self–esteem, engagement, or composite variables (looking at attitudes, behaviors, and social adjustment). Most studies focused on retention in grades K–3, although six looked at grades K–8. Three different reviewers analyzed the raw data for each study (with at least two reviewers analyzing the data for each review), and there was no difference found between reviewers relative to analyzing the study’s data or their statistical significance.
Control groups were typically matched on both academic and social–emotional measures, with 45 percent of the analyzed studies controlling for IQ, 65 percent controlling for achievement; 30 percent controlling for social–emotional adjustment; 75 percent controlling for socioeconomic status; and 70 percent controlling for gender. The 20 studies yielded 175 academic achievement comparisons and 169 achievement effect size calculations. Sixteen studies included social–emotional measures, resulting in 148 comparisons.
Of the 175 academic achievement comparisons, 9 favored retained students (only three comparisons held beyond the retention year), 82 favored promoted students, and 84 showed no statistical difference. The mean effect size calculation was -0.39, meaning that retained students scored just over a third of a standard deviation below their peers who were promoted. The effect size for language arts, reading, math, composite scores, and grade point average all favored the promoted students.
Of the 148 social–emotional comparisons, 9 favored retained students, 13 favored the promoted students, and 127 showed no significant difference. The mean effect size for social–emotional adjustment was -0.22, favoring promoted students across all domains: social, emotional, behavioral, self–concept, and adjustment.
The Bottom Line
Neither retention nor social promotion appears to address the needs of low–performing students. Although across academic and social–emotional measures, promoted, low–achieving students generally outperform their retained, low–achieving peers, the majority of these measures show no statistically significant difference between groups.
Who's Affected?
Low–achieving students in grades K–8 were the focus of this study.
Caveats
Although the analysis favors promoted students, the vast majority of effects show no statistically significant difference between retained and promoted students. There are effect–size benefits for retained students during the repeated year (+0.09); however, these effects are small in comparison to the negative long–term effect (-0.31). This research generally supports the notion that educators have yet to find a successful approach to remediation of low–achieving students.
The Study
Jimerson, S. (2001). Meta–analysis of grade retention research: Implications for practice in the 21st century. School Psychology Review, 30(3). pp. 420–437.
Other Resources
<BIBLIST> <CITATION> Position Statement on Student Grade Retention and Social Promotion National Association of School Psychologists </CITATION> <CITATION> Ending Social Promotion: The Effects of Retention Consortium on Chicago School Research </CITATION> <CITATION> School Dropouts: Home and School Effects ResearchBrief, 1(9) </CITATION> <CITATION> The Dropout Process in Life Course Perspective: Early Risk Factors at Home and School (subscription required) Teachers College Record, 103(5) </CITATION> <CITATION> Retention and Social Promotion: Research and Implications for Policy ERIC Digest 161 </CITATION> <CITATION> Retention–attrition in the Nineties ERIC Digest 73 </CITATION> </BIBLIST>