Loading...
Weight and See: Integration of Qualitative and Quantitative Information in Graduate Admissions
Adaryukov, James
Adaryukov, James
Citations
Altmetric:
Abstract
In graduate admissions, as in many merit-based decisions, evaluators must judge candidates from a flood of information. Some of this information is reported numerically, while some is conveyed verbally through documents, and evaluators must consider both kinds of information at once. While studies have individually examined the processes behind decisions based on verbal and numeric information, little research has examined how people process them in conjunction. The goal of this study is to evaluate how qualitative and quantitative information are used within graduate admissions decisions. We examine a unique and comprehensive data set of 2,231 graduate applicants to the University of Kansas, containing full application packages, demographics, and final admissions decisions for each applicant. To make sense of our qualitative information, we apply structural topic modeling, an extension of correlated topic modeling which allows topic content and prevalence to covary based on other metadata variables (i.e. department of study, gender, and race). We then incorporate the prevalence of derived topics as a cue in a policy capture model, predicting the actual decision whether to admit or deny applicants based on an applicant’s demographic, qualitative, and quantitative information. We find that quantitative information is overall a more consistent predictor of admission than qualitative information, but effects vary across departments. Furthermore, we find that applicant race and gender influence the prevalence of topics in qualitative materials, but negligibly predict the decision to admit on their own.
Description
Date
2022-12-31
Journal Title
Journal ISSN
Volume Title
Publisher
University of Kansas
Collections
Research Projects
Organizational Units
Journal Issue
Keywords
Quantitative psychology, Cognitive psychology, Social psychology, Decision-making, Graduate admissions, Machine learning, Multiattribute choice, Natural language processing, Policy capture