Construct Validity of the Chinese version of WHOQOL-BREF & Disabilities Module in 1000 adults with disabilities: an Item Response Theory analysis

This study was to evaluate the construct validity of WHOQOL-BREF & Disabilities module for physical disabilities (PD) and intellectual disabilities (ID) people, using Item Response Theory (IRT) Graded Response Model (GRM) analysis. As one of 14 centres for the field study of the WHOQOL Disabilities module, a stratified representative sample of Guangzhou general disabled people was approached for interview. It contained 1000 respondents: 807 physical disabled people and 193 intellectual disabled people. IRT GRM analysis was used to evaluate the construct validity of WHOQOL-BREF & Disabilities module. Each domain of WHOQOL-BREF showed appropriate fit to the twoparameter IRT GRM. The Disabilities module as overall one domain had a better fitting than as three domains. Discrimination parameters of item 3, 11, 12, 14, 16 and 25 were statistically equal to one for both the PD 5-point response version and the ID 3-point response version of the scales. The domain test information curves showed that the PD group had a larger range of scale scores of information > 5 than the ID group. The construct validity of WHOQOL-BREF and Disabilities module using GRM analysis complements previous validation studies.