During Fall term, 1981, a study was done comparing the effects of two modes of teaching. The in-class model and the self-study model of teaching CIS 133: Introduction to Numerical Computation (FORTRAN programming) were studied from two perspectives.

1. The knowledge the students had at the end of the course was measured on a specially designed final exam.
2. Demographic descriptions of the students who enrolled and who succeeded were compared.

All comparison was done on a group unit, not individual basis.

Findings were:

• Between the two groups, no significant difference was found in the knowledge displayed on the final exam.
• The overall success rate for the in-class mode (86%) was much higher than for the self-study mode (61%), with success defined as completing the work and receiving a grad of A, B, C or Pass.
• The noticible difference in characteristics within the groups is that people with no prior experience enroll in self-study much less frequently, and those who do succedd less often.
• Self-study instruction absorbs fewer teaching resources, but has a lower success rate, especially for beginners at programming.