Evaluation of the Chinese Version of the Low Vision Quality of Life Questionnaire Using Rasch Analysis

Objective: To use Rasch analysis to analyze and refine the Chinese version of the Low Vision Quality of Life Questionnaire (CLVQOL). Methods: The CLVQOL was completed by 189 low vision patients who attended the Low Vision & Vision Rehabilitation Center of the Eye Hospital, Wenzhou Medical University from January 2017 to December 2017. Mean age of the patients was 40.9±20.2 years, and the male/female ratio was 1.6:1. Based on the score for each item, Rasch analysis was performed on CLVQOL using the following steps: Dimensionality, response ordering, local dependence, infit and outfit analyses, differential item functioning, and subject targeting by using the Winsteps software version 3.72.2. Results: The CLVQOL contains four subscales. The eigen value of first contrast showed that the unexplained responses in each dimension are all around 2; the single-dimension test is still acceptable. Response ordering in each subscale showed options 2 and 4 were not clearly separated from their adjacent options. No notable local dependence was found for any subscale. According to the results of fit analysis, the mean square infit value of Q24 was 1.55, the mean square outfit value of Q24 was 1.58, both out of the range of 0.60 to 1.40. Differential item functioning was noted for sex on Q24, with a contrast difference >5. Analysis of targeting indicated relatively poor matches. The mean person in each dimension was higher than the mean item except on subscale 3, and the test information of each dimension was relatively narrow. Conclusion: The Rasch analysis of the CLVQOL shows that CLVQOL has a good degree of reliability and validity, but the option response to each item must redesigned to fit Chinese low vision patients according to their socioeconomic status. Key words: Rasch analysis; Chinese version of the Low Vision Quality of Life Questionnaire; low vision; subscale