2003 年 46 巻 4 号 p. 225-234
Exploratory and confirmatory analyses were used to examine the Satisfaction With Life Scale in a sample of Hong Kong university students. The exploratory analyses showed that the Satisfaction With Life Scale had good internal consistency reliability and that it appeared to measure a unidimensional construct. Comparison of Hong Kong students’ scores on the Satisfaction With Life Scale with the normative data for American students and mainland Chinese and Korean students showed that Hong Kong students were more similar, on average, to the American students than to the Asian students in their overall satisfaction with life. Further analysis of the scale’s structure using confirmatory factor analysis found that a modified two-factor model defined by present and past life satisfaction items and which allowed the error variances for two items to correlate fit the data best. However, inspection of the fit indices for the modified two-factor model suggested that it over fit the data, while a modified one-factor model, with correlated errors, also showed acceptable fit. Since the modified one-factor model was more parsimonious and agreed with the theoretical development of the scale as a unidimensional measure of general overall life satisfaction, it was retained. It is suggested that because of the Satisfaction With Life Scale’s good reliability and cross-cultural validity along with its brevity and broad-band nature in assessing overall satisfaction with life it could have wide applicability as an ancillary measure in a number of research areas.