<p> A primary goal of universities is to maximize student enrollment by improving course curriculum and enhancing specific programs. This is especially a challenge for smaller universities who want to offer specialized and highly diverse electives. This study aims to increase the quality and relevance of electives offered by understanding specifically what attributes students prefer more or less when choosing among alternatives. The results are used to explore how to use limited marketing and student-outreach financial resources to target students that are most likely to enter and complete courses and programs, based in part on their socioeconomic or demographic characteristics. The application is aimed at the economics unit at Michigan Technological University, which offers two programs: an undergraduate economics major and a master’s program in applied natural resource economics. Using an efficiently-designed discrete-choice experiment, this study elicits over 700 students’ stated preferences over a variety of attributes of economics courses related to the natural environment. Students were surveyed, and each student was presented with six different pairwise choice options that were developed based on 36 different alternative courses consisting of 8 attributes, such as class topics, professor rank, time of day, and research requirement. There were three significant covariates (political preference, economic interest, and domestic/foreign status). A latent-class discrete-choice random-utility model is estimated to probabilistically group students into different preference classes. Four preference classes emerge from the results that are highly heterogeneous in terms of the marginal utilities and the probabilities of being in a given class. For example, the largest class (40%) is made up of mostly conservative students, and the smallest class (12%) is made almost entirely of liberal students. While this study and the applicability of the specific results is unique to Michigan Technological University, the use of stated preference surveys and latent-class models is highly flexible and can be applied to any program at any university</p>
| Identifer | oai:union.ndltd.org:PROQUEST/oai:pqdtoai.proquest.com:10240242 |
| Date | 12 March 2017 |
| Creators | Dillingham, Reggie T. |
| Publisher | Michigan Technological University |
| Source Sets | ProQuest.com |
| Language | English |
| Detected Language | English |
| Type | thesis |
Page generated in 0.0019 seconds