PURPOSE: The majority of existing quality of life measures are based on urban-living environments. This study aimed at exploring the validity of using an urban-lifestyle-based health questionnaire with individuals living a traditional rural lifestyle. METHODS: The Short Form-36 (SF-36) interview was administered to 1603 rural Chinese residents. Semantic ambiguity of the items was investigated using tests of internal consistency, test-retest reliability, exploratory factor analysis, and clustering and ordering of item mean scores. The self-explanations from the respondents were adopted to interpret the implications of the changes in meanings of the items. RESULTS: Cronbach's alpha reliability coefficients were high, whereas test-retest reliabilities were low. Consistent with the original factor structure, eight factors were extracted using exploratory factor analysis. However, the composition of these eight factors was not in full accordance with the priori assignment of items to scales. Seven items violated the clustering and ordering of item mean scores. The association between the identified problems in validity and the change in semantic meanings in the context of the rural lifestyle was established. CONCLUSIONS: Quality of life assessment instruments based on urban-living arrangements may not be reliably used with individuals living in rural environments.
PURPOSE: The majority of existing quality of life measures are based on urban-living environments. This study aimed at exploring the validity of using an urban-lifestyle-based health questionnaire with individuals living a traditional rural lifestyle. METHODS: The Short Form-36 (SF-36) interview was administered to 1603 rural Chinese residents. Semantic ambiguity of the items was investigated using tests of internal consistency, test-retest reliability, exploratory factor analysis, and clustering and ordering of item mean scores. The self-explanations from the respondents were adopted to interpret the implications of the changes in meanings of the items. RESULTS: Cronbach's alpha reliability coefficients were high, whereas test-retest reliabilities were low. Consistent with the original factor structure, eight factors were extracted using exploratory factor analysis. However, the composition of these eight factors was not in full accordance with the priori assignment of items to scales. Seven items violated the clustering and ordering of item mean scores. The association between the identified problems in validity and the change in semantic meanings in the context of the rural lifestyle was established. CONCLUSIONS: Quality of life assessment instruments based on urban-living arrangements may not be reliably used with individuals living in rural environments.
Authors: Peter W H Lee; Tracy T C Kwan; Dora L W Kwong; Jonathan S T Sham; Edmond H N Pow; Anne S McMillan; Gordon K H Au Journal: Cancer Date: 2007-04-01 Impact factor: 6.860