Validation of a Modified Version of Interpersonal Reactivity Index for Medical Students Using Rasch Analysis

Construct: Empathy has been accepted to interweave both cognitive aspects (the ability to put oneself in another person's place), and affective (or emotional) aspects, indicating an emotional reaction or response to another person's emotional state. Literature supports the positive influences of empathy on doctor-patient relationship, patient satisfaction, and positive clinical outcomes. Background: Many studies have dealt with the development of empathy measurement tools for physicians and medical students. A frequently used empathy measuring instrument for medical students is the "Interpersonal Reactivity Index" (IRI) which was designed to measure the multi-dimensional aspects of empathy in the general adult population. Most previous literature which validated IRI for medical students has used factor analysis, whilst studies applying Rasch models have been limited. Our study aimed to investigate the psychometric properties of a modified version of IRI for medical students using Rasch analysis. Approach: Medical students (1,293) from 15 medical schools in South Korea participated in an online questionnaire consisting of 28 items of the Korean translated version of IRI. We applied exploratory factor analysis (EFA) using polychoric correlation matrix to determine the optimal number of factors followed by Rasch analysis and McDonald's Omega calculation. Findings: The adapted IRI-MS (IRI for medical students) consisted of 17 items in four dimensions: empathic concern (5), fictitious situation (4), perspective taking (4), and personal distress (4). The overall fit of IRI-MS revealed an acceptable goodness-of-fit for all 17 items and a positive point measure correlation for all items. Reliability indices from the Rasch modeling and McDonald's Omega values of all four dimensions were satisfactory for research. We found the Wright-Andrich maps and category probability curves of the IRI's four dimensions to be less than optimal in measuring empathy levels with adequate precision. Conclusions: Rasch analysis of IRI-MS fell short from being able to prove satisfactory validity in measuring the multidimensional nature of empathy in medical students. However, our study applying Rasch analysis may serve as groundwork for future studies, to further develop from the shortcomings of our findings.Supplemental data for this article are available online at at www.tandfonline.com/htlm .