The effect of commercial coaching on Scholastic Aptitude Test (SAT) scores was analyzed, using 1974-1977 test results of 2,500 non-coached students and 1,568 enrollees in two coaching schools. (The Stanley H. Kaplan Educational Center, Inc., and the Test Preparation Center, Inc.). Multiple regression analysis was used to control for student demographic variables. Separate analyses were conducted for the two schools, for students who were coached before their first SAT, or between their first and second SAT administrations. Coaching was effective at the Kaplan School, where students tended to be underachievers--i.e. they scored lower on standardized examinations than would have been predicted by personal and demographic variables. Kaplan students who were not underachievers also benefited from coaching; however, this finding was based on a limited sample, covering only a single administration of the verbal section of the SAT. Coaching at the Test Preparation Center was not found to be effective. However, these students were not underachievers. Results of this study were inconclusive, partly because of the non-experimental design. It was impossible to consider all possible effects of self-selection--for example, the possibility that students who chose to enroll in coaching schools were especially motivated to achieve higher scores. (Author/CP)