Kaj Malm : On the Dynamics of Undergraduate Performance and Dropout

The development and testing of a process model of undergraduate student performance and the influence of performance on dropout rates are discussed. An important feature of the model is a feedback loop that represents knowledge of results. A mathematical elaboration of traditional ideas regarding the effect of feedback on performance is included. A differential equation is obtained when the rate of change in the performance of a student at some point in time is proportional to the performance at that time minus some constant (which is a measure of the performance norm adopted by the student). The solution of the equation is the performance function, and its outstanding feature is its monotonicity. Testing of the model was undertaken with about 2,000 undergraduates enrolled in a 1-year mathematics course at the University of Helsinki, Finland, during 1975-1980. Each year, students were divided into those who failed and those who passed, and data were collected on performance and numbers of participants on seven examinations. The number who passed divided by the number in the last examination was remarkably stable through the 5 years. About 15 percent of the students participating in the last examination failed, and almost 90 percent of those who failed dropped out before the last examination. The total dropout rates were therefore more or less complementary to pass rates. Relative performance and mean standard deviations of the scores were also assessed. It is concluded that the theory explained 88 percent of the dropout variance and 81 percent of the variance of the pass rate. (SW)