The purpose of this study was to demonstrate a new approach in the analysis of teaching procedures, and show the importance of certain variates on the academic performance of college students. Objectives: The objectives of this study were threefold: To determine whether an ABA approach (definition on page 19) can be used effectively to identify relevant variables influencing college student's academic performance. To determine the influence of grading and exam frequency on achievement in college. To compare the effectiveness of a lectureless go-at-your-own-pace teaching procedure and lecture procedures on student performance as measured on a final exam and a six month retention test. Method Experiment One. Using an ABA reversal procedure in a Latin square design 253 students were rotated through testing conditions to determine the effects of grading and exam frequency on college students academic performance. The four testing conditions were: (a) weekly exams given which counted towards the student's grade, (b) weekly exams given which did not count toward their grade, (c) monthly exams which counted, and (d) monthly exams which did not count. An analysis of the results with respect to the students' grade point average was also carried out. Experiment Two. Three classes of introductory psychology were used in an interclass comparison. Each class was exposed to a different teaching procedure. Daily lectures with a monthly exam were used on the first class. Daily lectures with weekly exams were administered to the second class. The third class attended no lectures, took oral and written exams on each chapter when they felt read y for them, had to master one chapter before progressing to the next, and received immediate feedback on their test results. All three classes were also given a retention test six months later. Students performance on both the final exam and retention test were compared. Results Experiment One. Grading was found to significantly influence students academic performance (p .05). Though exam frequency had a consistent and systematic effect on performance, its effect was not found statistically significant. It was also noted that both variates influenced students of all grade point average ranges to the same degree rather than differentially helping only certain grade levels. Experiment Two. It was found that the students in the go-at-your-own-pace group outscored both lecture classes on the final exam and six month retention test. Conclusions The following conclusions are drawn from the results of research conducted for this report. The lack of differential effects in past comparative studies is due to the fact that the students' academic behavior was not differentially influenced. Intraclass analytic procedures can be employed which are more sensitive than the interclass comparisons employed in the past. Instructors can use grading to strongly influence college students of all grade point ranges to perform better. Two things future research in college teaching should include are: (a) more sensitive experimental procedures, and (b) a change in the variates analyzed. A stronger look should be taken at what behavioral contingencies the course includes rather than only what the teacher does in class. It is suggested that higher education should direct itself to more contingent systems of instruction rather than shifting to more self-controlled learning situations.