Comparative validity assessment of three android step counter applications; a semi-structured laboratory-based study

Area of Science:

• Digital Health
• Wearable Technology
• Physical Activity Measurement

Background:

• Step counting is a practical method for assessing physical activity (PA).
• Validation of PA applications (apps) is crucial for ensuring accuracy.
• Recent advancements necessitate ongoing evaluation of step-counting technologies.

Purpose of the Study:

• To compare the criterion validity of three popular step-counting apps: Pacer, Pedometer, and Google Fit.
• To assess the accuracy of these apps under controlled laboratory conditions.
• To evaluate app performance at both normal and fast walking paces.

Main Methods:

• A comparative experimental study involving 50 students using Android phones.
• Simultaneous use of Pedometer, Pacer, and Google Fit apps during walking trials.
• Video recording served as the criterion measure, with steps counted manually by two independent reviewers.

Main Results:

• No significant difference in step counts among the three apps.
• Significant difference between app-recorded steps and video criterion during normal-paced walking (p<0.05).
• Moderate Mean Absolute Percentage Errors (MAPE): Google Fit (6.6% normal pace), Pedometer/Pacer (9.2% normal pace), all apps (5.4% fast pace).

Conclusions:

• Pedometer, Pacer, and Google Fit show potential for general population step counting.
• Google Fit may offer superior accuracy for normal-paced walking.
• Future research should explore longer-term validity and diverse populations.