In graduate admissions, as in many multiattribute decisions, evaluators must judge candidates from a flood of information, including recommendation letters, personal statements, grades, and standardized test scores. Some of this information is structured, while some is unstructured. This creates a challenge for those studying these decisions, as most theories of behavioral economics specifically focus on decisions made from highly structured information. The goal of this study was to evaluate how structured and unstructured information are used within graduate admissions decisions. We examined a uniquely comprehensive dataset of 2,231 graduate applications to the University of Kansas, containing full application packages, demographics, and final admissions decisions for each applicant. To make sense of our documents, we applied structural topic modeling, an extension of correlated topic modeling that allows topic content and prevalence to covary based on other metadata (e.g., department of study). This allowed us to examine not only what information the letters and statements contain, but also the effects of gender, race, and department on how that information was conveyed. We found that most topics in the unstructured data related to specific fields of study and were difficult to generalize outside of that field. Consequently, we found that admissions committees behaved as if they prioritized structured numeric metrics, using unstructured information to check for disqualifications if at all. Furthermore, we found that applicant race and gender influenced the prevalence of topics in their letters and statements.