We all do.
and I don't have the time to read through list of all your posts and summaries.
How about reading through just the post that you asked me to simplify here
Just for simplicity's sake, are you able to sum up the problems in a few sentences?
I was hoping you could provide a short explanation of a few sentences
I can and did. The problem is one of context. For example, if you are familiar with special relativity and tensor calculus, I can sum up general relativity much more concisely and easier. If you aren't familiar with e.g., the development of hypothesis testing and why this dichotomous attempt to render specific statistical levels of significance can only determine the probability of outcomes GIVEN that the desired conclusion is false, then I need a lot more than a few sentences to explain.
If you are going to tell me there is a flaw in physics
Actually I don't really hold that position. It's more of a matter of a problem with theoretical physics, and the idea that mainstream particle physics is an empirical science backed up solely by agreement with experiment rather than a highly theoretical construct that relies greatly on mathematical abstractions. This is true even of the standard model, although the reliance isn't that significant, but as 1) the standard model is universally acknowledged to be
ad hoc and wrong, and 2) the replacements ARE fundamentally reliant on untested theories derived mathematically, the empirical support of the standard model doesn't count much when considering the empirical support of the whole of modern physics.
That said, as modern physics borrowed hypothesis testing from the social sciences, the logic remains the same: can the assumptions of normality (and so forth) of the probability functions of the phenomena in question determine whether or not the assumption that the groups characterized by the phenomena are identical combined with the fact that under said assumptions the probability that the groups are identical is less than some alpha level enables one to conclude that a factor not relevant to the assumptions or the testing is conclusive.
Hypothesis testing works like this:
1) Assume that the two groups (particles, patients, whatever) are identical. Develop and perform some experiment which "assumes" that they aren't (e.g., a placebo vs. real medicine experiment). Further assume the phenomena/phenomenon of interest that characterize the statistical nature (e.g., probability distributions) of the function(s) of said phenomena/phenomenon are given by particular models (e.g., normality assumptions).
2) Develop an experiment which allows you to test that, given two groups of particles/patients/etc. are identical, you can determine that (given the assumptions of probability distributions and so forth) the probability that the results you got are due to chance are less than some value (typically, <0.5 or smaller).
The issue is that at best, this procedure allows you only to, at best, determine that given you are wrong, your results are unlikely. It doesn't allow you to determine your assumptions are right, or (more importantly), given that you are testing hypotheses, that either one or another is more probable
and then back it up with a bunch of examples from psychiatry, I'm not going to be very convinced.
I gave you a description from particle physics. Again, the logic of hypothesis testing doesn't change.