Authors: Elysia Garcia, Subhash Aryal, Emily Spence-Almaguer, Danielle Rohr, Scott T. Walters
Item Response Theory (IRT) models have been extensively used in the field of education to identify link between a response to a test item and underlying latent capability of the test taker. We demonstrate the benefit of using IRT model to analyze health data using data from M. chat program such that statisticians can use the method in lieu of traditional methods including Cronbach’s alpha, discriminant analysis and factor analysis. M. chat is a technology based health coaching program and the baseline survey from the participants in the program includes response in different but correlated domains of diet, social habits, leisure practices, mental health, substance abuse, self-sufficiency and medication adherence. We analyzed baseline data from 416 subjects using IRT models. Our results indicated that responses pertaining to alcohol and substance abuse were the most discriminating items with an average discrimination estimate of approximately 4.99 whereas the least discriminating items were the diet habits, with an average estimate of -0.476.
Journal： Open Journal of Statistics
DOI: 10.4236/ojs.2018.83034 (PDF)
Paper Id: 85251 (metadata)
See also: Comments to Paper