A study was made of the effects of programed instruction versus the use of course objectives and reference materials on learning, transfer, and attitudes in two adult experimental groups. The subject was mathematical concepts basic to introductory statistics. Mathematical tests and attitude scales were administered before and after the experiment. Group 1 (programed instruction) learned more despite significant gains by all, but the two groups did not differ significantly on transfer or on time spent. Correlations were significant in both groups between initial scores and mathematical task attitudes, and between posttest scores and terminal attitudes toward the study of statistics. For Group 1, but not Group 2, correlations between initial task attitudes and initial performance declined during the experiment, initial attitudes toward statistics correlated strongly with pretest scores, and attitudes toward experimental treatments correlated strongly with terminal attitudes toward statistics. No significant attitude changes were found toward tasks or statistics. (author/ly)