Poster Presented by Ihnwhi Heo
Heo, I., Peña, M., Depaoli, S., Prather, A., Epel, E., & Zawadzki, M. J. (2026, April 24-26). Implementing Bayesian model averaging for logistic regression to identify important predictors of meditation attrition: Evidence from the mHealth Headspace app [Poster presentation]. University of California, Riverside, Riverside, CA, USA.
Using Bayesian model averaging, we noticed meditation group assignment and income level as the strongest predictors of attrition in a Headspace mindfulness intervention. Specifically, meditation group participants were 38% more likely to drop out, while middle-to-high income individuals were 36-39% less likely, highlighting key targets for retention strategies.