Graduate school choice: The effect of financial support on matriculation

This study is an empirical investigation of graduate school choice. Choice behavior, whether collegiate or graduate, is a singular phenomenon; and financial support plays a significant role in both contexts. Not only does financial support influence whether students enroll but also where students matriculate.; The study addresses two questions: (1) Does financial support influence graduate enrollment? and (2) How do competing support offers affect matriculation decisions? Part One describes the 1981-82 graduate admissions process in five academic disciplines (English, Political Science, Economics, Chemistry and Mechanical Engineering) in 18 doctorate-granting research universities in the United States. The study population includes 8700 graduate school applicants and 16,000 applications, including 6400 doctoral degree applicants. Part Two analyzes the graduate school choice behavior of admitted applicants. The study sample consists of 504 admitted doctoral applicants (and 1501 applications). A multinomial logit (MLOGIT) regression model of graduate school choice is used to estimate the effect of financial support on matriculation decisions. Several versions of the graduate school choice model are tested incorporating total support, net price, and institutional or programmatic variables as well as applicant characteristics.; Consistent with earlier studies, financial support is significantly related to probability of enrollment and to choice of institution. The effect of support (versus no support) on enrollment is substantial. Marginal differences in support also have a significant affect on matriculation decisions. Student characteristics, such as gender, ethnicity, GRE scores and previous graduate experience, are significantly related to enrollment but not to choice of institution.; Net price displays the same relationship to graduate school choice evident in studies of net price and college choice. The inclusion of institutional dummy variables substantially improves the prediction of graduate school choice over that obtained using net price alone. The net price model with institutional variables correctly predicted over 70 percent of the observed matriculation choices in those academic disciplines where support was significant. Programmatic variables, such as faculty quality, program effectiveness, proportion of doctoral graduates with national fellowship support and mean years to the doctorate do not predict matriculation as well as institutional variables do.; Policy recommendations are made based on the results.