Predicting U.S. News & World Report ranking of regional universities in the South using public data

Predicting U.S. News & World Report ranking of regional universities in the South using public data, Ph.D. dissertation by Angela E. Henderson (2017)

USNWR-best-colleges
Process chart showing steps used by USNWR to calculate the 2016 Best Colleges institutional rankings.

Abstract:

Using correlational analyses and multiple regressions, this study uses U.S. News & World Report’s (USNWR) 2016 college rankings data and data from the National Center for Education Statistics’ (NCES) Integrated Postsecondary Education Data System (IPEDS) to examine variables that explain institutional peer assessment score and rank. This study focused on the 97 institutions included in USNWR’s 2016 Best Regional Universities (South) ranking list.

Analyses in this study addressed four major foci: 1) correlations between USNWR subfactor data values and selected IPEDS proxies, 2) IPEDS variables that explained variance in peer assessment score, 3) IPEDS variables that explained variance in rank, and 4) the extent to which rank could be predicted based on these results.

The results of this study indicated three main findings. First, USNWR subfactors with direct or indirect IPEDS proxies were highly correlated with the identified proxies. Second, more than 85% of variation in peer assessment score could be explained by five or fewer proxy variables, which differ dependent upon institution sector (private or public). Third, more than 85% of variation in institutional ranking could be explained by five proxy variables and without the inclusion of the peer assessment score subfactor. Collectively, findings suggest USNWR rankings are no more than a reflection of institutional outcomes and financial resources.

usnwr-predicted-ranks
Percentage of predicted ranks classified into same decile as actual ranking.

Leave a Reply

Your email address will not be published. Required fields are marked *