Hacker News Reader

Comment by yxwvut | Hacker News Reader

This article is a nice counterpoint to what's felt like a trend of Scott's writing (outside of psychology) increasingly becoming contrarianism in search of evidence. Many pieces he's written in the last 1-2 yrs feel like he'll use some back-of-the-envelope calculation to get some ballpark (associative) estimate of some relationship and then base a worldview around it (or work backwards from the latter to the former), decades of causal inference and economics literature be damned. I guess maybe the delineation is social science vs medical science topics. When he's in his wheelhouse he's great to read.

View on HN

> contrarianism in search of evidence

But that is literally what this article is, and Scott admits as much! He says that he chose worms as the most "trollish" possible response, despite a lack of any strong evidence at all that worms are the answer.

The funny thing is that for all of their supposed smarts the bulk of the commenters in favor of the 'new' worm theory are jumping on it, just like they jumped on Ivermectin in the first place. It is in a way an interesting experience to see this happen in real-time.

If they're anything like me, they've always wanted the ivermectin thing to be worms-at-best-crap-at-worst, but have refused to accept "shut up and stop listening to the wrong science" as evidence in favor of that proposition. From that perspective their behavior is not only rational, it's far more rational and pro-science than the mainstream. Not surprising given who the audience is.

You’re totally right. I was more referring to his review of the merits of the various studies, which as a statistician I actually agreed with - I mostly read the article to understand the history of these studies, having no prior familiarity beyond the high level understanding that the evidence was weak. I expected this article to largely be pro-ivermectin-as-underdog given that the mainstream opinion is that it isn't, but I suppose this fits even more. His claim that it’s all able to be chalked up to worms is dubious at best - selection bias of one form or another (publication, metric selection, etc) seems substantially more likely. Sure enough, though his readership is credulously adopting this theory just as easily as they adopt his other contrarian points.

Reply on news.ycombinator.com