Measures of reliability and validity of school-based pedometer step counts in Singaporean children

Abstract

The purpose of the study was to examine the reliability and validity of school-based pedometer step counts in Singaporean children. Participants were 10 boys and 10 girls aged 10-11 years old. School-based physical activity was monitored over a five-hour period using an electronic pedometer (PCB 147PDO) and a tri-axial accelerometer (RT3), fastened securely at the waist. Test-retest measurements of physical activity revealed no significant differences in step count (21.86.1 vs. 22.16.3 steps/min, p>0.05) and vector magnitude (580.470.9 vs. 581.189.4 vm!mln, p>o.o51. The girls' physical activity measurements were 77.5 % and 87.4 % that of boys'for pedometer and accelerometer readings. Typical error (TE) expressed as a coefficient of variation (CV) for repeated measurements were 9.6% for pedometry compared to 4% for accelerometry. Intra-class correlations (ICC) were 0.90 for step count and 0.92 for vector magnitude, while Bland and Altman plots showed a similar spread of activity readings for the pedometer and accelerometer. Significant correlations were established between accelerometer readings and pedometer readings (r>0.75, p<0.05). Data from the study showed that the electronic pedometer was a reliable and valid motion sensor for assessing the school-based physical activity of Singaporean children, using a variety of reliability indicators. With the reliability and validity of pedometers established, future research should focus on the efficacy of intervention programs to curtail sedentary behaviours and heighten the attractiveness of active behaviours in the context of school.