Document Type

Honors Project On-Campus Access Only


The U.S. News & World Report (USNWR) annual college ranking lists are widely-circulated measures of the comparative quality of higher education institutions. These rankings shape the admission decisions of prospective students and the admissions policies of ranked institutions alike. USNWR’s “Best National Liberal Arts Colleges” ranking for 2020 was generated by combining 14 weighted institutional statistics, or subfactors, for each institution. However, the process by which subfactor data is manipulated is not fully disclosed by the publication. In this paper, we effectively model the 2020 ranking list using an elastic net approach, forecast future subfactor values for all ranked institutions based on historical data using univariate time series simulation, and apply the elastic net model to our forecasted data to project rankings for 2021. We use these projected rankings to understand what constitutes a significant change in rank for an institution and determine the most effective course of action if Macalester College wishes to efficiently improve its rank while staying true to its mission. Through this process, we successfully predict the 2020 rank of the top 40 institutions within approximately 4 places of their published rank. In addition, we isolate the 5 subfactors that, with improvement in their values, have the highest potential to improve Macalester College’s rank in 2021: Alumni Giving Rate, Peer Assessment Score, Spending Per Full-Time-Equivalent Student, Percentage of Classes Under 20 Students, and Average Adjusted Faculty Salary.



© Copyright is owned by author of this document