Background: The aim of this study was to advance physical activity (PA) surveillance in physical education (PE) by establishing a steps/min guideline that would accurately classify fifth and sixth graders as engaging in PA for 10 min or one-third of the PE lesson time.
Methods: Data were collected on 147 (11.48 +/- 0.83 y) girls and boys in 14 intact classes from five schools. PA was assessed via behavioral observation (i.e., criterion) and pedometry (i.e., predictor). Logistic and linear regression techniques were employed to generate pedometer steps/min cut points. Classification of outcome probability (c), sensitivity, specificity, and receiveroperating- characteristic (ROC) curve statistics tested the decision accuracy of generated steps/min cut points.
Results: PA measures were strongly correlated (r = 0.80, P < 0.01). A steps/min interval of 60 to 62 was the best cut point indicator of students meeting the PA guidelines.
Conclusions: Findings support steps/min as an accurate quantifier of PA time in structured PA programs. PA surveillance via pedometry in PE using empirically derived criteria is an objective, valid, and practical mechanism for assessing a primary PE and public health outcome.