Comparison of four habitual physical activity questionnaires in girls aged 7-15 yr.

PURPOSE: Questionnaires provide a practical approach to measuring physical activity in children and adolescents, particularly for large-scale epidemiological studies. The purpose of the present study was to compare four habitual physical activity questionnaires and to assess their long-term reliabilities.
METHODS: Subjects were female, aged 7-15 yr (and their parents), who participated in a cohort study in which predictors of the onset of menarche were being investigated. Questionnaires consisted of three single-item and one multi-item habitual physical activity questions and were sent to the 640 participants. Test-retest reliability of the questionnaires was assessed on 100 randomly selected participants 11 months later.
RESULTS: The Spearman's correlation coefficient was highest (r = 0.40) between the Godin-Shephard Score and the Perspiration Score and was lowest (r = 0.10) between the Stairs Score and the Specific Activity Score. The correlation coefficients were higher when the questionnaires were reported to have been completed by the parents alone, rather than by parents with the assistance of their daughters. The test-retest reliabilities were r = 0.44 for the Perspiration Score, r = 0.59 for the Stairs Score, r = 0.48 for the Godin-Shephard Score, and r = 0.53 for the Specific Activity Score. The reliabilities were higher when the retest was reported to have been completed by the parents alone.
CONCLUSIONS: This study suggests that the Perspiration Score, the Godin-Shephard Score, and the Specific Activity Score may provide simple and practical measures of habitual physical activity for children and adolescents.