Reliability and validity of the system for observing children\u27s activity and relationships during play (SOCARP)

Background: Children frequently engage in diverse activities that are broadly defined as play, but little research has documented children’s activity levels during play and how they are influenced by social contexts. Assessing potentially modifiable conditions that influence play behavior is needed to design optimal physical activity interventions. Methods: System for Observing Children’s Activity and Relationships during Play (SOCARP) was developed to simultaneously assess children’s physical activity, social group sizes, activity type, and social behavior during play. One hundred and fourteen children (48 boys, 66 girls; 42% overweight) from 8 elementary schools were observed during recess over 24 days, with 12 days videotaped for reliability purposes. Ninety-nine children wore a uni-axial accelerometer during their observation period. Results: Estimated energy expenditure rates from SOCARP observations and mean accelerometer counts were significantly correlated (r = .67; P < .01), and interobserver reliabilities (ie, percentage agreement) for activity level (89%), group size (88%), activity type (90%) and interactions (88%) met acceptable criteria. Both physical activity and social interactions were influenced by group size, activity type, and child gender and body weight status. Conclusions: SOCARP is a valid and reliable observation system for assessing physical activity and play behavior in a recess context.