Predicting Total Student Credit Hours Production by Cohort Stratification
The objective of this study is to develop predictive models of total student credit hours (SCH) prior to the fall semester of interest by using preregistration data from Texas Woman's University (TWU). We developed two different approaches to predict SCH for undergraduate and SCH for graduate students separately. Our first approach is based on the patterns of weekly counts of SCH observed over time. For our second approach, we developed a model that relies on an average of SCH and a total headcount. This research presents a self-contained procedure to predict headcount and includes a criterion to select a prediction model for the average of SCH. After explaining the development of each of our SCH prediction models, we compare the results and discuss their strengths and weaknesses.