Hi Timothy,

I am not an expert in this area, but it seems to me that if you take averages of steps, or peak values from each step you have created discrete values which can be treated as repeated measures - though a large amount of repeated measure could be difficult to process.

If you consider a continuous signal, time series analysis would be more appropriate. You might find an approach outlined in the following article interesting https://www.mdpi.com/1099-4300/20/10/764/htm