The reliability and validity of the adolescent physical activity recall questionnaire
PURPOSE: This study assessed the test-retest reliability and validity of the Adolescent Physical Activity Recall Questionnaire (APARQ) among 13- and 15-yr-old Australians. METHODS: Two studies were conducted using the same instrument. Self-reported participation in organized and nonorganized physical activity was summarized into four measures: a three-category measure of activity, a two-category measure, and estimated energy expenditure expressed as a continuous variable and as quintiles. The reliability study (N = 226) assessed strength of agreement for all measures between responses to two administrations of the questionnaire. The validity study (N = 2026) assessed the relationship between the APARQ and performance on the Multistage Fitness Test (MFT). RESULTS: Reliability study: for the three-category measure, percent agreement ranged 67-83% and weighted kappa ranged 0.33-0.71. For the two-category measure, percent agreement ranged 76-90% and kappa ranged 0.25-0.74. For energy expenditure expressed as a continuous variable, the intraclass correlations coefficients were generally greater than 0.6 for grade 10 students, but most were below 0.5 for grade 8 students. Validity study: for the three-category measure, mean laps were higher in the adequately and vigorously active categories than the inactive category for girls, but only the mean laps in the vigorously active and inactive categories were significantly different for boys. For the two-category measure, mean laps were higher in the active category than the inactive category for all groups. Correlations between energy expenditure and MFT laps were 0.15, 0.21, 0.14, and 0.39 for grade 8 boys, grade 8 girls, grade 10 boys, and grade 10 girls, respectively. CONCLUSION: The APARQ has acceptable to good reliability and acceptable validity, but further validation using other methods and in other population groups is required.