Reliability and Validity of Self Report of Aerobic Activity: Family Health Project

Abstract

Abstract Two studies are presented which deal with the reliability and validity of self-reports of aerobic activity. The first study compared two forms of self-report data obtained as part of a study of intervention strategy. The two forms were daily self-report and weekly retrospective report. Analyses covering the overlapping time period revealed differences in mean minutes engaged in activity, but modest and statistically significant correlations between the two forms among adults participating in the intervention group. No meaningful results were obtained among adults in the control group, nor among children in either group. The need for the development and testing of self-report forms among children was noted, and was the subject of the second study. In Study II, six different forms on which third to sixth grade children recorded their aerobic activity were compared against two days of continuous observations of their behavior. The six forms varied along two dimensions. The first dimension varied the time period of reporting: half the forms required reporting by segments of the day; the other half allowed for reporting of the day as a whole. The second dimension varied response format: one third of the forms asked for a dichotomous (yes; no) report; one third asked for a trichotomous (no; yes, but less than 20 min; yes, for 20 min or more) report; and one third asked for reporting of precise numbers of minutes. The segmented day forms resulted in significantly higher agreement with the observers' report of activity. This finding is interpreted to be congruent with recent research on the importance of context in the recall of behavior. No effect from parental assistance in child's form completion was detected.