Developing a Shortened Quality of Life Scale from Persian Version of the WHOQOL-100 Using the Rasch Analysis

BACKGROUND Few studies use modern approaches to develop a quality of life (QOL) questionnaire with acceptable construct validity, especially in Iran. Our main objective was to construct a new validated and uni-dimensional questionnaire, based on WHOQOL-100, using the Rasch analysis. METHODS In a population-based, cross-sectional study in 2007, 500 Tehran residents aged≥18 were randomly sampled. The Persian version of WHOQOL-100 was used to measure the participants' QOL. After using targeting and person fit analysis, we performed category/threshold ordering, item fit, and differential item functioning analyses, in succession. We used outfit or infit statistics>1.5 and <0.5 for detecting under-fit and overfit items/persons, respectively. We also deleted items with disordered category and/or threshold. Person Separation Index and test reliability were also calculated in the datasets. RESULTS Male to female ratio was 0.98 and the mean age (SD) of participants was 35.1 (12.8) years. Initial analyses showed significant differences in quality of life between age groups (P=0.002), educational levels (P=0.001), and current health status groups (P<0.001). We eliminated 67 persons for under-fitting, 38 items for category and/or threshold disordering, 6 items for under-fitting, and 26 items for item bias. Test reliability for the final 30-item scale was 0.89. CONCLUSION We prepared a shortened version of the WHOQOL-100 that is single construct, uni-dimensional and free of item bias or any disordering, according to the Rasch model.