I have some time time series data which I aim to analyse against 0. To do this I intend to use a one samples t-test whilst chopping the data into interval windows. I will likely use a Bonferonni correction to strictly control for the multiple comparisons.
The research question would likely benefit from changing the sampling method to a sliding window, however, I'm aware of the huge amount of correlation I will get rendering anything which requires statistical independence useless. Is there an established method to overcome this issue?
Cheers
Looks to me like you want a longitudinal model if I understand you. See:
https://www4.stat.ncsu.edu/~davidian/st732/notes.html
Best David Boot
I can always trust research gates user to provide amazing content! Thanks David Eugene Booth
- Badges
- Science topic