NotesFAQContact Us
Search Tips
ERIC Number: ED233086
Record Type: Non-Journal
Publication Date: 1983-Apr-13
Pages: 2
Abstractor: N/A
Reference Count: N/A
Benefits of Coaching on Test Scores Seen as Negligible.
Report on Education Research, v15 n8 p3 Apr 13 1983
THE FOLLOWING IS THE FULL TEXT OF THIS DOCUMENT: A new study by a pair of Harvard University researchers discounts earlier findings that coaching can substantially improve student performance on the Scholastic Aptitude Test (SAT). "There is simply insufficient evidence that large score increases are a result of a coaching program," write Rebecca DerSimonian and Nan Laird in the most recent issue of the Harvard Educational Review. There is "evidence in the data to support a positive effect of coaching. However, the size of the effect which we can safely attribute to coaching is too small to have much attraction either for individual examinees or for educators." DerSimonian and Laird's findings are at odds with those of two other Harvard researchers, Warner Slack and Douglas Porter, who in 1980 caused a stir when they claimed that an intensive six-week coaching course could boost SAT scores by 40 to 85 points (RR, Sept. 17, 1980). The study by DerSimonian, a Ph.D. candidate in biostatistics, and Laird, a professor of biostatistics at Harvard's School of Public Health, estimates a much smaller gain of about 10 points. DerSimonian explained that the difference had to do with the research methodology. Unlike previous studies, which lumped together all analyses of coaching effects, DerSimonian and Laird looked at well-controlled and less well-controlled studies separately. They found that the size of test score gain attributed to coaching is correlated with the design of a particular study. The less well-controlled the study, the greater the effect of coaching is shown to be. When gains achieved by coached students are compared with national norms, the effects of coaching appear four to five times greater than in studies where coached students are compared with a control group. If the controlled studies are taken as the most reliable, "then it appears that the benefits of coaching are indeed negligible," write DerSimonian and Laird. Slack, meanwhile, took issue with DerSimonian and Laird's methods and defended his own. "It's not clear that the studies they assigned most weight to were better than the others," he said. (Author)
Publication Type: Journal Articles
Education Level: N/A
Audience: Practitioners
Language: English
Sponsor: N/A
Authoring Institution: Capitol Publications, Inc., Arlington, VA.
Identifiers - Assessments and Surveys: SAT (College Admission Test)