The Purdue experiment in mass instruction in undergraduate sociology was initiated in the fall of 1967. Lectures are combined with group discussions, and achievement in the course is evaluated mainly by essays written during the semester. Data reported were collected during three successive semesters when minor changes were made in course structure. Scores on the general sociology test indicated that students were learning a satisfactory amount of basic sociology. Questionnaires were used to assess student reactions to essays as the basis of grades, to the discussion meetings, to their assigned tutor (discussion leader) and to the lectures. Associations between these variables were studied, and correlations between over-all course evaluation and other response and personal variables were investigated while controlling on grade, approval of essay system, sex, age and semester in college. Findings were that the over-all evaluation was positive, that modifications in course structure had little effect on student responses, that the general reaction to the course was more strongly associated with reaction to lectures than to other course structure variables, and that freshman students appear to have special difficulties. (EB)