Further evaluation of observational and mechanical measures of physical activity

The team wanted to know if cheap wearables give the same numbers as trained observers.

They watched kids play while Fitbits, heart-rate straps, and old-style pedometers ran.

Every minute they coded activity level by hand with the OSRAC tool to get a gold-standard line.

Fitbit steps and heart-rate rose and fell with the coder scores most of the time.

When kids kept both feet on the ground, Fitbit was trustworthy; when they bounced on mats it under-counted.

Pedometers shook around and gave shaky data more often.

Sutton et al. (2022) later took the same idea into homes and parks. They showed a research-grade wearable (Physilog®5) still matches lab measures in kids with Prader-Willi and Angelman syndromes, so the trick works outside the clinic too.

Verberg et al. (2022) used waist sensors to study kicking in adults with Down syndrome. Their sensors lined up with the TGMD-3 checklist, giving another vote that wearables can replace clipboards when you pick the right body part.

Wuang et al. (2009) did the opposite job: they proved the paper-and-pencil BOT-2 stays solid across weeks. Together the papers say: new gadgets are handy, but classic tests still hold their value—use both, not either-or.

You can strap a Fitbit on a client and see real-time activity spikes during recess or therapy. Just remember to double-check if the kid jumps on trampolines or swings, because airborne time kills step counts. Pair the live data with quick BOT-2 or TGMD-3 probes and you get speed plus depth without extra staff.