ACCURACY AND EFFICIENCY OF DIRECTLY OBSERVED CADENCE: THE SKYROCKET STUDY

Authors

Mariya Boikova, University of North Carolina at CharlotteFollow
Cayla R. McAvoy, University of North Carolina at CharlotteFollow
Agnes Bucko, University of North Carolina at CharlotteFollow
Taylor Fiorentino, University of North Carolina at CharlotteFollow
Trudy Moore-Harrison, University of North Carolina at CharlotteFollow
Michae Dulin, University of North Carolina at CharlotteFollow
Laura H. Gunn, University of North Carolina at CharlotteFollow
Catrine Tudor-Locke FACSM, University of North Carolina at CharlotteFollow

Abstract

BACKGROUND: The criterion measure for determining cadence (steps/min) is direct observation, where steps are counted manually in real-time (original count) and often video recorded for verification purposes. The Consumer Technology Association (CTA) has recommended that at least two researchers review all video footage to recount and verify accuracy of the original count. This recommendation was made without a clear understanding of the magnitude of deviation potentially associated with direct observation and original counting. It also did not consider how much time, and associated personnel costs, may be required to implement the recommended reviewing and recounting process. METHODS: The SKyRoCKeT Study is a five-year, NIH-funded R01 study that aims to assess cadence and physical activity intensity in 6-20-year-olds. The study is composed of three lab visits with up to 23 distinct 5-minute activity bouts conducted across each of 360 participants. To evaluate accuracy, a single researcher reviewed and recounted video recordings from 45 activity bouts. The formula for mean absolute percentage deviation (MAPD=1/n Σ | (recount - original)/original| x 100%) was applied, where n was the total number of activity bouts conducted. To evaluate the time, it would take one full-time (40 hours/week) researcher to perform a single round of reviews/recounts, the 23 possible activity bouts were multiplied by a 5-minute bout duration for the study total of 360 participants. RESULTS: It took less than four cumulative hours to review/recount the 45 activity bouts for this analysis, and the MAPD indicating potential deviation between the original count and the review/recount was 0.56%. It would take one full-time researcher 17.25 weeks, or approximately 4 months, to complete a single round of recounts for all possible activity bouts implemented in this study. This time and personnel requirement would be doubled if the CTA recommendation was enacted as intended. CONCLUSIONS: Although video recording activity bouts is prudent as a redundant backup strategy, the MAPD value calculated herein was less than 1%, making mandatory and all-inclusive systematic reviewing/recounting processes potentially unnecessary and an inefficient use of resources. These data-driven findings are useful for planning and organizing future research projects. Grant or funding information: NIH NICHD - R01HD105768

Recommended Citation

Boikova, Mariya; McAvoy, Cayla R.; Bucko, Agnes; Fiorentino, Taylor; Moore-Harrison, Trudy; Dulin, Michae; Gunn, Laura H.; and Tudor-Locke, Catrine FACSM (2024) "ACCURACY AND EFFICIENCY OF DIRECTLY OBSERVED CADENCE: THE SKYROCKET STUDY," International Journal of Exercise Science: Conference Proceedings: Vol. 16: Iss. 3, Article 37.
Available at: https://digitalcommons.wku.edu/ijesab/vol16/iss3/37