>>12167619>It's not a useful threshold. Every field that relies on 0.05 has a reproducibility crisis.B doesn't follow A here. We have made most progress in science thus far using p values and 0.05 as a cutoff; we learned what genes were and what genes are, stem cells, we learned how to sequence the human genome, learned what causes down-syndrome, learned what disease genes are relavent, and cured/improved outcomes of many types of cancers using p values. The reproducibility crisis is because there is entirely too many junk papers out there. Those junk papers would still exist with/without p-values.
>The p actually means the probability that you could get your data if your hypothesis was wrong. I think I understand what you're sayin here, but for clarity, the p-value is "the probability that you get a test assuming the null hypothesis is true".
Example if you t-test two datasets A and B, you're asking "assume a probability distribution (t-distrubution in this case) with parameters mean/sigma from dataset A. If you pulled random data from this distribution, how likely would it be that you would get dataset B's mean/sigma?"
Not saying you don't know this, but it was unclear what was meant.
>Sadly in academic papers, this other side is never considered.I don't think you know much about the academic world or how the scientific framework works with collection of evidence, see what I wrote
>>12167602>Consider the study that showed with p < 0.05 that eating chocolate helped you lose weight.
This is an outright falsehood and gross misunderstanding. There was no study that showed chocolate can help you lose weight with a p < 0.05. I suggest you refrain from commenting about the state of academia when its clear you know little about it.
