The effect of information on the college choice to enroll at the University of Michigan.

Abstract

The purpose of this study was to determine if the final college choice of high school seniors could be affected by a post-admission letter. The person-environment fit model would posit that greater congruence between the letter and selected factors in a student's background would yield a greater impact of the letter on the student's intent to enroll. The research groups were (a) relatively high socioeconomic status, (b) relatively low socioeconomic status, and (c) a group whose stated college preferences were matched by the University of Michigan. Each group was subdivided into two treatment groups and a control group. Subgroups were equalized for gender, residency, ethnic minority status, and alumni ties. Each subgroup and a control subgroup had one hundred subjects. Thus, there were a total of 900 subjects in the study. One of four treatment letters was sent to each subgroup. The "prestige" letter described Michigan in terms of rankings and excellence. The "financial aid" letter described Michigan as a relatively friendly institution with good financial aid policies. The "high match" letter stated that Michigan matched criteria specified by the student in the ACT Student Profile Section. The "general welcome" letter, included general, positive information about the University. Analysis of the enrollment patterns indicated that the treatment letters did not generally have the desired impact on enrollment. Subjects in the control groups, who received no letter, generally were at least as likely to enroll as those subjects who received a letter. Second, the treatment letters did not differentially effect the likelihood of enrollment by research group. Third, this lack of positive treatment effect was found even for those students who had not yet decided upon a college. None of the independent variables studied, including gender, residency, ethnic minority status, significant alumni ties were significant in differentially impacting the lack of effect of the treatment letters. Implications for college admissions offices were discussed. Suggestions for future research were indicated.