Multimodal Analysis of Occupational Therapy Admissions Video-Based Interviews and Predicting Reviewer Scores

• In Event: Assessment Innovation in Professions Education

Abstract

The objective of this study was to quantify how reviewers assess non-cognitive competencies of applications for admission to an Occupational Therapy graduate program. Specifically, this study investigated how these appraisals differ when the applicant's responses to a set of relevant questions are collected through different modalities, i.e., written document vs. video. The possibility of automating the assessment process was also explored by calculating the correlation between the scores assigned by reviewers and relevant personal characteristics predicted by a machine learning algorithm based on the applicants’ submissions. The study results highlight the potential value of video-based interviews and their automated analysis as an adjunct to assessing conventional document-based statements of purpose, to provide additional insights into candidates' non-cognitive competencies.

Authors

• Mashrura Tasnim, University of Alberta

• Shaniff Esmail, University of Alberta

• Susan Mulholland, University of Alberta

• Eleni Stroulia, University of Alberta

• Mary Roduta Roberts, University of Alberta