Context: I have to send my final dissertation manuscript to my committee tomorrow. My chair called me this morning and told me my stats are all wrong because I ran too many tests with too small of a sample (10 participants). When I asked for guidance on resolving this, they said "You want a PhD? Figure it out". I have spent all morning reading resources and I have gotten myself more confused with each one.

This is a single group pre-post design based on feasibility and exploring preliminary clinical outcomes. I originally had planned paired samples t-tests, but I ended up with such a small sample that it would inflate error rate. Also because of the small sample, any assumptions of underlying distributions can't be met, so I'm using exact sign tests. I modeled my stats after another published pilot study with 7 participants and 15 comparisons. They do not do a correction.

I ran the 7 exact sign tests for my outcome variables, found significance in 6/7, and also presented the data visually.…