Time series methods to compare quantitative continuous data (e.g., heart rate) and qualitative/subjective data (e.g., self-reported stress)

Question

I work with a lot of data from sports wearables (e.g., heart rate). I ran a study where people wore a heart rate tracker whilst doing an activity. Every 5 minutes they were prompted to rate how much stress they felt on a scale of 0-10.

I'd like to see whether there is a relationship between heart rate (quantitative data) and self-reports (qualitative data). Are there any methods (e.g., for correlation) that I can try?

I have looked into dynamic time warping, but it seems it only works for the same quantitative data (e.g., voice recordings). I've also tried calculating the percent change and correlating them, but since my qualitative variable has 0 some percent changes were inf values

Answer 1

First, try some graphical analyses, it is always helpful.

Then, look at fixed effect Poisson regression; fixed effect because I understand that you have several observations per individual. Poisson because your dependent variable is made of positive integers.

At a later stage, you may be interested on fixed-effect ordered models such as fixed-effect ordered models. Some comparison and implementation (in Stata) is provided for instance in

Dickerson, A., Hole, A. R., & Munford, L. A. (2014). The relationship between well-being and commuting revisited: does the choice of methodology matter?. Regional Science and Urban Economics, 49, 321-329. https://www.sciencedirect.com/science/article/pii/S016604621400101X