DerGegner
New Member
well a little background first
I aim to be a scientist, specifically, a neuroscientist (currently an undergrad in linguistics)
unfortunately a lot of imaging studies are full of shit because the authors weren't competent in the foundations of science
I don't want to be one of them
I don't want to be full of shit
so I am reading Scientific Method in Practice by Hugh G. Gauch and more or less got through the chapter on statistics, skimming in some places
I knew about Bayesian vs frequentist issues before but the most important novel thing I have taken away from my most recent reading is that a frequentist analysis answers the question P(data | hypothesis) whereas a Bayesian analysis answers P(hypothesis | data) ... I mean I remember having learned that only rejections of null hypotheses should really be considered to mean anything as far as research is concerned but I didn't know until now that frequentist analyses don't offer any probability that a hypothesis is true
in fact I think I remember essentially being taught the opposite in a class, that a p value is simply the probability of an extreme value (without mentioning that the null hypothesis is assumed to be true beforehand!!) and I feel sort of like I've been jacked off now
and not in a good way
so now I feel like there have got to be a thousand other hurr durr moments I will be able to avoid in the future if I am on my guard
so if you have any suggestions do share