An evaluation of TICCIT (Time-shared Interactive, Computer-controlled, Information Television) involved over 5,000 community college students in introductory algebra and English composition courses. Comparisons between computer-assisted instruction and lecture-discussion sections of the same courses focused on four aspects of student performance: course completion rates, achievement, attitudes, and activities (time allocation). Other evaluation questions focused on teacher attitudes, teacher role, and program implementation; including administration, site management, and courseware design goals. It was decided to construct objective and essay tests specifically for this program, to measure both end-of-course achievement and immediate learning. TICCIT had a significant positive impact on achievement. The dramatic decreases noted in course completion rates may be inherent in self-paced instruction because students who have trouble managing their own instruction are risks. Student attitudes towards TICCIT were often less favorable than toward conventional teaching methods, but attitudes improved when TICCIT couses were supplemented by small group discussion with an instructor. Results suggest that TICCIT may be inappropriate for community colleges since only those students with a strong initial grasp in the subject matter benefited substantially. In itself, computer assisted instruction is no panacea; results depend on factors involved in the instructional process. (CP)