OBJECTIVES: Three studies were undertaken to establish the reliability and validity of the Scottish physical activity questionnaire (SPAQ), developed to aid seven day recall of leisure and occupational physical activity. METHODS: To establish reliability, SPAQs (n = 34) were completed on a Monday and the following Wednesday. Thus each questionnaire measured four identical days. To establish concurrent validity, 94 participants completed a SPAQ and an adapted stage of exercise behaviour change questionnaire. Responses to SPAQ were then analysed by stage of exercise behaviour change. In a further study of criterion validity, 30 volunteers wore a Caltrac motion sensor for four consecutive days, after which they completed a SPAQ. RESULTS: In the first study, total physical activity had a coefficient of repeatability (R) of 53 minutes. Occupational physical activity showed a similar variance (R = 54.6 minutes) but leisure physical activity was more reliable (R = 29.3 minutes). The main variation in occupational physical activity was found to be walking (R = 39.8 minutes). In study 2, a one way analysis of variance showed the expected relation between physical activity and stage of exercise behaviour change, confirming the concurrent validity of SPAQ with the stage of exercise behaviour change model. In study 3, several erroneous recordings affected both SPAQ and the Caltrac results (kcal). After relevant corrections had been made, the correlation between the two measurement devices was 0.52 (p < 0.05). CONCLUSIONS: SPAQ has been shown to be reliable and to hold strong concurrent validity and limited criterion validity. The main limitation in SPAQ appears to be the measurement of occupational walking. It is therefore recommended that further work be conducted to refine the measurement of this physical activity component. It is evident nonetheless that SPAQ can be used with confidence to measure outcomes in physical activity interventions when account is taken of its limitations.
