Comments on: Simpson’s Paradoxical Card Trick

By: Andrew Prayle

Andrew Prayle — Thu, 02 Sep 2010 19:23:02 +0000

Great post, and very succinctly explained! Always concerning in medical papers to read about post hoc subgroup analysis! Your post explains why.

By: Sex discrimination in graduate admissions? A real-life aggregation paradox | Curving Normality

Mon, 07 Jun 2010 14:39:09 +0000

[…] likelihood of being admitted is of a different sign. This is an example of Simpson’s Paradox. As I illustrated in an earlier post, this paradox can occur when two or more sub-populations are aggregated, and if one or more of the […]

By: Daniel R Hawes

Daniel R Hawes — Tue, 01 Jun 2010 00:39:16 +0000

Very nice.I was just wondering if you’ll write more about this, but now I see your earlier comment. Great stuff…

By: Rense Nieuwenhuis

Rense Nieuwenhuis — Mon, 31 May 2010 14:57:24 +0000

Thanks Chris,

I expect to continue writing some more about this paradox, for there is some more to say about it. Also, there is some literature on the conditions under which such a paradox can arise. Perhaps I should write about it. If you want me to send some literature, please let me know.

By: Chris

Chris — Mon, 31 May 2010 13:01:44 +0000

Great point Rense. I have worked on corpus linguistics and I’m always a little worried about the fact that all linguistic corpora are subsets, and small ones at that (just think of the set of all English spoken compared to the de facto standard academic resource the British National Corpus of 100 million words). I don’t know what to do about it, except be very cautious in my conclusions.