Reliability of using motion sensors to measure children’s physical activity levels in exergaming

Objectives: This study examined the reliability of two objective measurement tools in assessing children’s physical activity (PA) levels in an exergaming setting. Methods: A total of 377 children (190 girls, M_age = 8.39, SD = 1.55) attended the 30-min exergaming class every other day for 18 weeks. Children’s PA levels were concurrently measured by NL-1000 pedometer and ActiGraph GT3X accelerometer, while children’s steps per min and time engaged in sedentary, light, and moderate-to-vigorous PA were estimated, respectively. Results: The results of intraclass correlation coefficient (ICC) indicated a low degree of reliability (single measures ICC = 0.03) in accelerometers. ANOVA did detect a possible learning effect for 27 classes (p < 0.01), and the single measures ICC was 0.20 for pedometers. Moreover, there was no significant positive relationship between steps per min and time spent in moderate-to-vigorous physical activity (MVPA). Finally, only 1.3% variance was explained by pedometer as a predictor using Hierarchical Linear Modeling to further explore the relationship between pedometer and accelerometer data. Conclusions: The NL-1000 pedometers and ActiGraph GT3X accelerometers have low reliability in assessing elementary school children’s PA levels during exergaming. More research is warranted in determining the reliable and accurate measurement information regarding the use of modern devices in exergaming setting.