English version of the computer vision symptom scale (cvss17): translation and rasch analysis-based cultural adaptation

Abstract

Background Because the CVSS17 was originally developed in Spanish, the objective of this study was to adapt it linguistically and culturally into English while evaluating its psychometric properties. Methods After translating and adapting the CVSS17 to English, 441 participants (aged 18 to 65 years) from a general population, recruited from an on-line panel, completed the English version (CVSS17ENG). To determine the measurement properties of CVSS17ENG, we used the partial credit model. To assess convergent validity, coefficients of correlation between CVSS17ENG and the Ocular Comfort Index or Visual Discomfort Scale were calculated. A subset of 218 subjects was tested for test-retest reliability. In addition, differences between CVSS17ENG and CVSS17 were tested through Differential Item Functioning (a Rasch sta-tistic used to check item bias). Results A total of 441 responses to CVSS17ENG (average age, 38.57; age range, 19–65; females, 50.24%) showed good fit to the Rasch model, good precision (person separation index = 2.73), and suboptimal targeting (-1.43). Residual principal component analysis suggested multidimensionality, but this was ruled out by a disattenuated correlation coefficient value of 0.82, and no DIF according to sex or age was found. Pearson correlation for CVSS17ENG-VDS was 0.54 (p < 0.01) and Spearman correlation for CVSS17ENG-Ocular Comfort Index was 0.66 (p < 0.001). For test– retest reliability, the limits of agreement were 9.39 and -8.61. Rasch analysis results were similar for CVSS17 and CVSS17ENG and there was no evidence of item bias based on gender or age. Conclusion The English version of CVSS17 demonstrates comparable performance to the original, indicating its suitability for clinicians and researchers to reliably assess Digital Eye Strain among English-speaking users of screen-based electronic devices.