[Scaling the SF-36 in a Chinese population].

OBJECTIVE

To facilitate the cross-cultural comparison, we employed the principles of fuzzy mathematics in scaling the SF-36 in a Chinese population.

METHODS

We stratified an urban population according to the age, sex, and occupation, and selected 650 people to participate in the study. A total of 646 valid questionnaires were returned. In the 8 domains of the SF-36, the interval estimation of the scores of different alternative answers of 6 domains which have more than 2 alternative answers were conducted.

RESULTS

The fuzzy degrees of most interval estimations were less than 0.5, except 4 estimations in the domain "physical functions". The ordering of the scores of 8 domains was more similar to the original assumption by our scaling criteria than by the US scaling criteria.

CONCLUSION

The scaling criteria set in this study is more suitable for use in Chinese population and we suggest that the alternative answers of the domain "physical functions" be simplified to "yes" and "no".