An Investigation of College-student Performance

THE INVESTIGATION described in this article was partially motivated bya desire to focus attention on certain perceived deficiencies in computerassisted-instruction (CAl) research. One emphasis in CAl research in 1969 was on exploring and discovering new ways in which humans and computers could interact. It involved the design of special hardware and the implementation of new programming techniques (software). The reader is referred to Wexler (1970) for a brief account of the historical development of CAL In his account, one can clearly see that the primary emphasis in CAl research projects had been system development. Most of tbe projects were implemented under ideal operating conditions for small and highly motivated groups of students, while .little or no attention was given to evaluating the curriculum or the pedagogical methods used. Usually, the system was designed to simulate some intuitive concept of a "good teacher" and to "individualize instruction." The result was a large collection of complex, interesting, and, from a computer scientist's point of view, valuable instructional systems. However, little machinery was available to judge their educational value or relevance in any systematic or quantitative way. For several years, the CAl Laboratory at the Institute for Mathematical Studies in the Social Sciences (IMSSS) had been offering a course in mathematical logic. The availability of this course made it possible to collect information on student behavior in elementary mathematical logic, information which was unavailable before the advent of CAL This study was a first attempt at an in-depth analysis of some of the factors which contribute to problem difficulty in elementary mathematical logic. The major focus of the study was to develop an understanding which might eventually lead to a quantitative theory of problem solving in logic. This work was in the spirit