How Quickly Can You Know a Participant? A 3-Week Triage Point for Oura Ring Adherence in College Students

A study analyzed data from the Oura Ring wearable device from 584 first-year college students across two semesters (Fall 2022 approximately 8 weeks, Spring 2023 approximately 15 weeks). Researchers developed a logistic regression model based on device wear data from the first three weeks to identify students at risk of low engagement. The model successfully predicted students with the lowest and highest adherence with accuracy of AUC 0.81-0.86 in both semesters. When the model trained on Fall data was applied to Spring data without additional training, it achieved AUC accuracy of 0.84. Average wear time during the first three weeks was the most important factor in predicting future student engagement. Results suggest that three weeks of passive wear data is sufficient to identify students who need support in maintaining engagement with the study.