Po-Yi Chen1, Grace Yao. 1. Department of Psychology, University of Kansas, 1415 Jayhawk Blvd.Fraser Hall, Room 426, Lawrence, KS, 66045-7556, USA.
Abstract
PURPOSE: Fuzzy set theory (FST) can improve various aspects of measurement with questionnaires. However, very little is known about how to use FST to measure quality of life (QOL). The main purpose of our study was to find an appropriate fuzzy measure for QOL that, while demonstrating the advantages of FST, can also be compared with mainstream QOL measures, most of which use traditional Likert-type scales. METHODS: Referring to the literature on fuzzy scoring methods, we first revised the measurement scale and scoring method of the traditional WHOQOL-BREF (i.e., a five-point Likert-type scale) to create three versions of a fuzzy WHOQOL-BREF. Then, we examined the psychometric relationships of these three fuzzy measures and the traditional WHOQOL-BREF in a within-subject design. RESULTS: Our results show that a fuzzy-scales weighted-by-membership (FSWM) version of the WHOQOL-BREF is comparable to the traditional WHOQOL-BREF in that it accepts strong invariance and shows almost perfect agreement. It also demonstrates higher reliability and face validity than the traditional WHOQOL-BREF. CONCLUSION: We recommend that future studies examine the use of FSWM to measure QOL.
PURPOSE: Fuzzy set theory (FST) can improve various aspects of measurement with questionnaires. However, very little is known about how to use FST to measure quality of life (QOL). The main purpose of our study was to find an appropriate fuzzy measure for QOL that, while demonstrating the advantages of FST, can also be compared with mainstream QOL measures, most of which use traditional Likert-type scales. METHODS: Referring to the literature on fuzzy scoring methods, we first revised the measurement scale and scoring method of the traditional WHOQOL-BREF (i.e., a five-point Likert-type scale) to create three versions of a fuzzy WHOQOL-BREF. Then, we examined the psychometric relationships of these three fuzzy measures and the traditional WHOQOL-BREF in a within-subject design. RESULTS: Our results show that a fuzzy-scales weighted-by-membership (FSWM) version of the WHOQOL-BREF is comparable to the traditional WHOQOL-BREF in that it accepts strong invariance and shows almost perfect agreement. It also demonstrates higher reliability and face validity than the traditional WHOQOL-BREF. CONCLUSION: We recommend that future studies examine the use of FSWM to measure QOL.
Authors: Georg E Matt; Maria R Turingan; Quyen T Dinh; Julie A Felsch; Melbourne F Hovell; Christine Gehrman Journal: Addiction Date: 2003-09 Impact factor: 6.526