Development of scales to assess children's perceptions of friend and parental influences on physical activity

Abstract

Background: Many children do not meet physical activity guidelines. Parents and friends are likely to influence children's physical activity but there is a shortage of measures that are able to capture these influences. Methods: A new questionnaire with the following three scales was developed 1) Parental influence on physical activity; 2) Motives for activity with friends scale; and 3) Physical activity and sedentary group normative values. Content for each scale was informed by qualitative work. One hundred and seventy three, 10-11 year old children completed the new questionnaire twice, one week apart. Participants also wore an accelerometer for 5 days and mean minutes of moderate to vigorous physical activity, light physical activity and sedentary time per day were obtained. Test-retest reliability of the items was calculated and Principal Component analysis of the scales performed and sub-scales produced. Alphas were calculated for main scales and sub-scales. Correlations were calculated among sub-scales. Correlations between each sub-scale and accelerometer physical activity variables were calculated for all participants and stratified by sex. Results The Parental influence scale yielded four factors which accounted for 67.5% of the variance in the items and had good (α > 0.7) internal consistency. The Motives for physical activity scale yielded four factors that accounted for 66.1% and had good internal consistency. The Physical activity norms scale yielded 4 factors that accounted for 67.4% of the variance, with good internal consistency for the sub-scales and alpha of .642 for the overall scale. Associations between the sub-scales and physical activity differed by sex. Although only 6 of the 11 sub-scales were significantly correlated with physical activity there were a number of associations that were positively correlated >0.15 indicating that these factors may contribute to the explanation of children's physical activity. Conclusion: Three scales that assess how parents, friends and group normative values may be associated with children's physical activity have been shown to be reliable and internally consistent. Examination of the extent to which these new scales improve our understanding of children's physical activity in datasets with a range of participant and family characteristics is needed. © 2009 Jago et al; licensee BioMed Central Ltd.