Key points are not available for this paper at this time.
Objectives . To assess the internal consistency, test–retest reliability and factorial structure of the SF‐36. To examine the discriminant validity of the instrument in relation to measures of physical (angina, diabetes) and psychological health (chronic GHQ, alcohol problems as defined by CAGE score). Design. A cross‐sectional survey design is used. Methods. A questionnaire containing the SF‐36 was administered to a cohort of 10 308 civil servants aged 35–55 at baseline as part of the Whitehall II study of occupational and life‐style influences upon health. Administration of the SF‐36 was repeated on a subset of 289 participants four weeks later. Clinical groups were distinguished by self‐report measures of health. Results . Data show high internal consistency (alpha .75–.85). Test‐reset reliability was poor for role limitations due to physical problems (.38), though acceptable for all other scales (range .60–.89). Orthogonal and oblique factor analyses confirm a two‐factor structure corresponding to dimensions of physical and psychological health. Validation of the scales against criterion groups defined on the basis of self‐reported health indicate that physical functioning, social functioning and general mental health have good discriminant validity. Conclusions . The study provides further evidence that the SF‐36 may be used to investigate a range of physical and psychological problems even in relatively healthy populations. However some of the scales do not exhibit high levels of reliability and need careful consideration before being used. When physical and mental component summary scores are used, it is suggested that these are best derived from oblique factor solutions.
Roberts et al. (Sat,) studied this question.