Comprehensive simulation modeling in higher education : the state of the art and the stability of the ICLM

The economic climate in higher education has forced many administrators to examine some of the scientific management tools developed in industry to aid in the budgeting and decision making processes. One of the most promising is comprehensive simulation modeling which considers the university as a total system of interrelated departments, colleges, and service-units. The study has two primary purposes. First, it seeks to determine the current state of the art of comprehensive simulation modeling by considering how the models are being used, problems that are being encountered, and the perceived impacts of modeling on the budgeting and decision making processes. Second, it tests the validity of a basic assumption of this type of model—the stability of the induced course load matrix (ICLM) over time. This matrix translates numbers of students into departmental workloads. If it is too unstable, the models may be unsuitable for budgeting and decision making purposes. To study the state of the art, a questionnaire was sent to the Director of Institutional Research at 93 schools which had expressed an interest in the Resource Requirements Prediction Model-1.6 (RRPM-1,6), Seventy-one responses were received and indicate that, at the time of the study, most models were either newly implemented or would be implemented shortly. The institutions have used (or plan to use) them in budgeting and decision making and, in general, see the models as having a strong impact on each of these areas. Technical problems of model construction were found to be relatively minor, but a great deal of concern was expressed about the behavioral problems of gaining administrative support and utilization. To study the stability of the ICLM, a number of ICLMs were generated for the University of Houston for the steady enrollment period of 1971 through 1974. In general, the ICLM was judged to possess enough stability to serve as a base for models such as RRPM-1.6, although some departments and colleges appeared to be more prone to instability than others. The alternative matrix generation techniques of using headcount versus full-time equivalent student data and seven versus three student levels were examined to determine if their use contributed to stability. No significant differences were found. In terms of the level of aggregation, increasing stability was found at the college level, but matrices representing a time period of a year were not found to be more stable than those representing a semester.