interesting article although you lost all validity when you started calculating probabilities with…
Albert Higgins

Hi Albert

Well, if I’ve lost your respect over something like this, then the respect reservoir I enjoyed with you was pretty shallow to begin with :).

In all seriousness, I actually like what I did there. Nothing about it is “random”: I’m trying to make explicit the epistemic reality that social science is not engineering. Journalists, for all their good work, usually assume that findings within social science are somehow deterministic in nature, but they are not; there is always a large amount of uncertainty lingering in every single analytic step.

In the example from the text, I’m actually concatenating as few steps as possible and being almost comically overconfident (believing that each step has a 90% probability of being true).

The story is more complex than the journalists who wrote about it realize, and, more importantly, the things that are presented to be true are really quite uncertain.