Extrapolation and multicollinearity have the potential to produce some seriously skewed statistics, which was the concern that was attempted to get pointed out in the author's piece. Multiple variables to consider when compiling any sort of research, and over the years I've observed too many research efforts overlook such, which is problematic.

@SpikeTalon Another problem with "sponsored" research is the questions are written with bias so a desired outcome is achieved.

@puff And that too...

Brown's own method(s) may have also been carried out in a sloppy manner like those he was criticizing, but it was the general point he tried to make and not his own research ways that I was focusing on there, and he made a valid observation, in particular with regression analysis. Correlation does not automatically imply causation, even in regression analysis. Then throw in a mixture of too much extrapolation, the exclusion of certain predictor variables like lurking or confounding variables, and multicollinearity which can produce skewed statistics. I see that sort of thing in research efforts all too often, and yet it appears a majority of people are taking such research at face value.

Just because the author's own way of collecting and analyzing data is questionable would not remove the fact that those he was criticizing was a reasonable conclusion, and one of which I certainly have and continue to observe when it comes to the research of hot-button topics such as racism/abortion/and gun control etc. Just because people readily show/share their research with the general public, does not automatically mean they are correct in said research or more trustworthy for doing so.

@SpikeTalon Yes he had some valid points, but a very poor way of presenting them. As someone that has had to devour research papers for decades, you should learn how to read them early in your educational process. I have been highly critical of many papers due to their faulty use of their statistical samples, limited universes, selection criteria, deselection criteria, faulty or plain bad math.

Yeah... but look how long that took, and how much public outrage was first required before action was finally taken against her. People like that don't deserve to have excuses made up for them in an attempt to somehow try to justify their behavior(s).

@SpikeTalon One major problem is the institutional investment in published papers. Not just the author, but every person that was involved in writing, review, approval, submission, peer review, then publishing. It is rare that even discredit papers are removed except by the author. It is a very flawed system.

@glennlab I know... and that's precisely what concerns me.

What site did you read that on? Just curious is all.

@SpikeTalon It was someone sitting at a bar (The Dock House in Beaufort, NC)..