Convergent and Discriminant Validity of Quality of Life Questionnaires

Background/Purpose: Quality of life (QOL) has become a standard outcome measure in physical activity-related intervention and observational studies. Due to widespread use of QOL, many instruments have been developed and used. Few studies have compared the various QOL measures. The purpose of this study was to assess the convergent and discriminant validity of the quality of life (QOL) measures using Multitrait-Multimethod (MTMM) analysis (Campbell & Fiske, 1959). The QOL questionnaires used in the study were WHO (1998) and Roh QOL for middle-aged adults (Roh, 1988). These measures were composed of three sub-dimensions of physical, social, and psychological wellbeing.

Method: Based on the review of dissertations and articles between 2000 and 2012 published in Korea, two QOL questionnaires (WHO and Roh QOL) were selected by the order of number of applications in the previous research. Two QOL questionnaires were then administered to 297 (male = 150) college students. Seven-day test-retest reliability was acceptable for all three measures (r=.70~.79, .71~.92; WHO and Roh QOL, respectively). The MTMM analysis was used to evaluate the evidence of convergent and discriminant validity.

Analysis/Results: Convergent validity coefficient for physical wellbeing was r =.49 and the corresponding heterotrait- monomethod coefficients were r = .30 - .44 for WHO and r = .51 - .52 for Roh QOL, and heterotrait-heteromethod coefficients were r = .31 and .39. The convergent validity coefficient for social wellbeing was r =.48 and the corresponding heterotrait- monomethod coefficients were r = .43 - .44 for WHO and r = .51 - .53 for Roh QOL, and heterotrait-heteromethod coefficients were r = .34, r = .42. . The convergent validity coefficient for psychological wellbeing was r =.60 and the corresponding heterotrait- monomethod coefficients were r = .30 - .43 for WHO and r = .52 - .53 for Roh QOL, and heterotrait-heteromethod coefficients were r = .38, r = .50.

Conclusions: The patter of correlations in the MTMM matrix met the criteria of Campbell and Fiske (1959) for validity of these QOL tests to measure psychological welling but not physical and social wellbeing in order to use for the youth and adult groups. Further studies and analyses are warranted to examine the content-related validity evidence of physical, social, and psychological wellbeing sub-dimensions of the WHO and Roh QOL tests developed for the senior and adult in order to use for the youth and adult groups.

Keywords: MTMM, convergent validity, discriminant validity, the quality of life