Publication: Test-retest reliability of adolescents’ self-reported physical activity item in two consecutive surveys

Research Impact: Measures of self-reported physical activity among adolescents are used frequently in public health. However, researchers have suggested the need for scrutiny in the validity of the measures. One type of validity is the test-retest of the measures. In this paper, researchers in Finland used the same self-report physical activity question in two consecutive studies. This was a novel approach to test retest reliability because the respondents (n=2572) completed two surveys on the same day. Recency effects were removed through cognitive overloading by completing personal questions about health and health behaviours in the WHO collaborative Health Behaviour in School Aged Children (HBSC) study prior to completing the Finnish School-aged Physical Activity (F-SPA) study. The data were interpreted through Cohen’s Kappa and Intraclass correlation coefficients of a single item physical activity item. There was moderate agreement when the item was grouped into four (k=0.50) and two (k=0.60) categories. In addition, there was good to excellent reliability (range 0.694–0.765) for boys and girls aged 13 to 15 years, but only fair (0.565) for boys aged 11. A single item was proved to have good intra-rater reliability when measuring the frequency, intensity and time of physical activity among young adolescents. This has practical implications for national surveys. This particular item is used in the recently collected Health Behaviour in School-aged Children study, collected in Ireland. It is also one of two items in the recently completed Children’s Sport Participation and Physical Activity (CSPPA) study. The paper is open access.

Ng, K., Hämylä, R., Tynjälä, J., Villberg, J., Tammelin, T., Kannas, L., and Kokko, S. “Test-retest reliability of adolescents’ self-reported physical activity item in two consecutive surveys”.  Archives of Public Health. 2019. DOI: 0.1186/s13690-019-0335-3

Tagged with: