Data presentation options to manage variability in physical activity research

Valbuena et al. (2017) wrote a how-to paper for messy step-count graphs. They showed one real data set and seven ways to clean it up.

The tricks are simple. You can add a 7-day moving average, draw a baseline line, or split the y-axis. All work in Excel.

Raw daily steps jumped up and down like a heartbeat. After adding a moving average, the trend looked smooth and clear.

The authors give step-by-step pictures so you can copy each tactic in five minutes.

Dowdy et al. (2022) found almost no one uses structured visual tools in JABA. Valbuena’s seven tactics sit on the shelf, just like the rest.

Manolov et al. (2023) went further and built a free web app that scores trend and overlap for you. Their tool automates what Valbuena still does by hand.

Fradet et al. (2025) push even further. They turn the same visual features into effect-size numbers for meta-analysis. You need clean graphs first; Valbuena shows how to make them.

Next time your client’s data look like a toddler’s crayon scribble, pick one of the seven fixes before you trust your eyes. A smooth line can save you from calling a fake change real, or missing a real change in the noise.