macrocephalopod

Also, of course, the author's subsequent claims about "adjustments" are non-mathematical and incoherent. Nearly every published paper constructs CIs on parameters estimated in regressions with control variables @federicolois

I hope this is the last time I have to remind the people of the internet: please stop making extremely confident claims about technical subjects where you would likely fail introductory level coursework.

My point is unless you know the prior distribution (which you don't) you cant assume any distribution of the process generating it, therefore, when you assume that your point estimate is saying something, that is wrong.

For effective purposes the only assumption you can do and be safe from a biased analysis is uniform distribution. Any value is fair game. If you choose a prior distribution in your bayesian process you are baking in assumptions that may not hold true.

Also not acknowledging the limitations of regression analysis, among those omitting variables biasing, heterogeneity bias, selection bias, multicollinearity, plain old measurement error, etc. Is why if your crude does not hold, you cant claim anything.

You did probably the most extensive work on the issue, and I am all-in into publishing your results (that's how science works) but being responsible in what you claim and can support with the data at hand is key.

I hope this is the last time I have to remind the researcher: please stop making extremely confident claims about technical subjects where you would likely fail evidence evaluation coursework.

Again, I want to assure readers who may not know better that the above is literally gibberish -- it is technical language used without regard for the underlying meaning.