Investigating Reliability, Validity, and Measurement Invariance of the Eight-Item Vocational Rehabilitation Engagement Scale With a Sample of Multilingual Clients

Abstract

Vocational rehabilitation (VR) engagement is a vital component for successful rehabilitation outcomes. The purpose of this study is threefold: First, we investigated the item clarity of the Vocational Rehabilitation Engagement Scale (VRES); second, we explored the factorial structure; third, we examined the measurement invariance of the VRES across gender and education level for culturally diverse and bilingual clients in public VR program (i.e., clients speak English as a second language). Given the positive effects of client engagement in health and rehabilitative care on outcomes, it is important to validate this brief instrument to measure VR engagement for culturally diverse and bilingual clients. Data, collected from 16 VR clients who completed the original VRES in the focus group, were subjected to content analysis, and data, collected from 264 clients who completed the eight-item VRES and Working Alliance Inventory Short Form (WAI-S), were subjected to confirmatory and multigroup confirmatory factor analysis (MGCFA). Based on the feedback from the focus group, the first item was revised to increase clarity. However, the preliminary results indicated that the revised item had very low community value, and therefore, we removed the item from the subsequent analysis. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) results provided support for a unidimensional structure for the eight-item VRES. Measurement invariance across gender and education level was achieved after correlating error terms. The scale had strong reliability, and scores from the VRES were significantly moderately correlated with working alliance. The eight-item VRES is a reliable and valid measurement tool to use with culturally diverse and multilingual participants.