Using Multiple Regression To Evaluate a Peer Tutoring Program for Undergraduates.

The Tutorial Assistance Program (TAP) of the University of California (Irvine) is described and a new method of evaluation is illustrated through an analysis of the performance of students in six large introductory classes. This approach to evaluation starts with a multiple regression equation for predicting course grades of those not in tutoring and applying the same equation to the TAP students to predict what they would have earned had they not attended tutoring. The multiple regression equation was obtained using: (1) high school grade point average (GPA); (2) Scholastic Aptitude Test mathematics and verbal scores; and (3) scores from the College Board tests in mathematics and English. Comparisons were made for 4,194 non-TAP students and 748 TAP students. Results demonstrate the benefits of TAP, particularly for students considered to be underprepared at college entry. TAP students actually earned higher grades than would have been expected had they not attended TAP. One of the aspects that probably accounts for the efficacy of TAP is that tutors and students attend the same course with the same instructor. Implications of these findings for program improvement are discussed and educational applications of the methodology identified in this paper are suggested. Three tables in the text and five in an appendix present study findings. An eight-item list of references is included. (SLD)