The SF-12 in Australia: cross-validation of item selection

Objective: To cross-validate the selection of the questionnaire items for the SF-12 in an Australian sample. Method: The sample comprised respondents with complete data on the SF-36 summary measures and the 12 items that comprise the SF-12 (n=17,671) from the 1995 Australian National Health Survey. Two sets of mental and physical summary scales were analysed, based on either scoring from Australian normative data or the standard United States normative data scoring. Forward stepwise linear regression analyses of weighted data provided the proportion of variance in the summary scales explained by the SF-12 items. Results: The SF-12 items predicted at least 90% of the variance in both the physical and mental summary scales of the SF-36, whether they were scored with Australian or United States normative data. Implications: The SF-12 is an appropriate substitute for the SF-36 when the summary scales are of interest but a briefer instrument is required.