Evaluating the Scale Functioning of The Health System Literacy Scale in Canada (HSL-CAN): A Partial Credit Model

Abstract Background Adapted from the Navigational Health Literacy scale (HLS 19 – NAV-HL), the 20-item Health System Literacy (HSL-CAN) was developed to measure health system literacy among Chinese Canadians. However, little is known about the functioning of its response categories. Method This study aimed to evaluate the proper functioning of the response categories using data from 681 adult Chinese Canadians aged 30 years or older who have resided in Canada for at least six months. Data were collected through a cross-sectional online survey. The partial credit model was utilized to evaluate the dimensionality of the scale, item- and category-level fit statistics, and the ordering of category thresholds. Results Findings supported the unidimensional construct of the scale. Most items met the item theory response (IRT) model requirements, demonstrating acceptable item fit statistics, except for three items with outfit-t statistics outside the acceptable range. Adjacent response category thresholds were increased monotonically, and threshold distances were within the recommended upper range, although relatively narrow distances were observed for several items. Person separation reliability and person separation index indicated good internal consistency and adequate discrimination among individuals with different level of health system literacy. Conclusion This study provides evidence supporting the unidimensional construct and a proper functioning of the 5-point Likert response scale, suggesting that the HSL-CAN is a psychometrically appropriate instrument for evaluating health system literacy among Chinese Canadians. Future studies are needed to examine the scale’s applications across more diverse populations in Canada.