>>11063973>calculate overlap of control and experiment distributions numerically then use that to estimate probability Again I'm a stats retard so forgive me but is that not the point behind statistical tests?
>>11063993>sample size is absolute dogshitAgreed. It's not my data set, I've been told to analyse it. Originally there were 6 repeats but 2 didn't work. The people who gave it to me said they have analysed it already to get results (this is from a lab with a dedicated statistician so must be legit). Looking at it it does seem to be statistically significant despite the small sample size; multiple times both controls have no protein expressed with the experimental group showing a lot of protein expressed.
