Application of industrial engineering methods to the analysis of engineering student retention at a case study university
The issue of engineering student attrition may be perceived as a daunting and overwhelming obstacle to overcome. Hundreds of studies have been conducted in an attempt to better understand the various aspects of the engineering student attrition, including reasons for leaving, identification of attrition prediction factors, prediction modeling, retention programs and strategies, and assessment of those programs and strategies. This dissertation takes a case study approach to consider each of these factors, while utilizing a variety of industrial engineering methods for modeling and analysis. The dissertation contributed to the current literature in three major aspects. First, it introduced the application of the probabilistic neural network (PNN) to predicting student attrition, which had not been previously been done in the published literature, and compared the results to two other prediction models commonly used for attrition prediction. The PNN proved to have greater sensitivity (probability of correctly predicting a non-retained student) than the other two models. Second, in response to calls from the current literature to publish more qualitative data-driven analysis of the effectiveness of engineering summer bridge programs, this work took an objective mixed model approach to quantitatively evaluate the case study university's past engineering summer bridge program. Results were supported through bivariate correlation analysis of the qualitative data captured through a college-wide student survey. Third, this study modeled a DMAIC approach to addressing engineering student attrition, which has rarely been applied to the field of academia. A report on best practices in engineering student retention was used as a benchmarking tool, and a standard student persistence survey was used to capture student perceptions and guide recommendations for improvements.
Thesis (Ph.D.)-- Wichita State University, College of Engineering, Dept. of Industrial, Systems and Manufacturing Engineering