Background/Purpose Advances in technology have produced many potential physical activity assessment improvements. The purpose of the research was to assess step count measurement accuracy of mobile computer pedometer applications compared to waist mounted pedometers under controlled and pseudo-controlled conditions.
Method Thirty-three participants walked on a treadmill under controlled conditions at six different speeds (40 to 107 m/min), and walked a 250m outdoor path (pseudo-controlled). Participants wore Yamax SW200 and NL1000 waist mounted pedometers, and Apple (iPod, Pedometer Free) and Android (Razr, Pedometer by Levente Bagi) pedometers.
Analysis/Results Absolute-value percent error (APE, clinically acceptable ≤ 10%) scores were computed for each pedometer, with observed steps as the criterion. APE scores were analyzed using repeated measures ANOVAs (Bonferroni post hoc analysis) for controlled (speed x pedometer) and pseudo-controlled (pedometer) conditions. Android pedometer APE scores were significantly (p ≤ .023) different from all pedometers at each speed. At speeds from 67 to 107 m/min the SW200, NL1000 and Apple pedometers did not statistically differ (p ≥ .115). APE ≤ 10% was found for the SW200, NL1000 and Apple pedometers at the speeds of 67 to 107, 67 to 107, and 54 to 107 m/min, respectively. The Android pedometer APE scores were ≥ 10% for all speeds. In the pseudo-controlled condition, Apple, SW200 and NL1000 pedometer APE scores did not significantly differ (p ≥ .136) and APE scores were ≤ 10%.
Conclusions The Apple pedometer application was comparable to waist mounted pedometer accuracy in controlled and pseudo-controlled conditions.