Scott is well aware of all those issues. Something is, however, generally better than nothing. This is not a randomised clinical trial, not even close. It's not even a longitudinal correlational study. It's an opt in survey with a small sample. Unless you get huge effect sizes you're going to be have low certainty of anything and all it's really useful for is suggesting things for people to try or treatments to be prioritised for testing in higher powered studies.

I especially invite comments from participants here who have a professional understanding of statistical studies in clinical trials of new medicines.

Scott has an M.D. and is currently doing his residency in psychiatry.