From 00709ef6f24257a8ee71925ecccd551a1e802232 Mon Sep 17 00:00:00 2001 From: "M. Taylor Saotome-Westlake" Date: Mon, 17 Feb 2020 18:39:01 -0800 Subject: [PATCH] Human Diversity review Presidential sprint session 3: univariate fallacy Is it bad to treat a topic in passing when you still want to cover it in more depth in a future post? (I wanted to work in the mathematics of the "typical set" into this graf, but I didn't see how to make it fit.) Maybe it's good. --- content/2018/untitled-metablogging-26-december-2018.md | 2 +- content/drafts/book-review-human-diversity.md | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/content/2018/untitled-metablogging-26-december-2018.md b/content/2018/untitled-metablogging-26-december-2018.md index f3a37f0..a2f0c9b 100644 --- a/content/2018/untitled-metablogging-26-december-2018.md +++ b/content/2018/untitled-metablogging-26-december-2018.md @@ -20,7 +20,7 @@ I guess I haven't made any new posts here in almost two months?—which is not g * I argue that this isn't always practical given the _far_-less-than-perfect information available in many social situations. Since not all traits can be cheaply, precisely, and verifiably measured, sometimes people might want to use (perceived) sex as a proxy, or as a [Schelling point](https://www.lesswrong.com/posts/Kbm6QnJv9dgWsPHQP/schelling-fences-on-slippery-slopes) for coordination. * I hopefully-accurately summarize Ozy as arguing that gender, like money, is socially constructed by collective agreement. It's coherent to argue that gender should be fully consensual, attributed on the basis of self-identity. * I argue that just as not all possible money systems are feasible (in particular, you couldn't run an economy in which anyone could arbitrarily declare what they thought other people should categorize as a _dollar_), not all possible gender systems are feasible. Fully consensual gender _sounds_ like a good idea when you phrase it like that (what kind of monster could possibly be against "consent"??), but doesn't reflect the structure of probabilistic inferences people actually make in the real world when they have some information about people's sex. - * I need to write an in-depth post about the overlap-along-one-dimension-does-not-imply-overlap-in-the-entire-configuration-space statistical phenomenon ([standard diagram](https://en.wikipedia.org/wiki/File:Pattern_classification_medium.JPG)) of which I have decided that ["univariate fallacy"](https://twitter.com/sapinker/status/1071245692180578305) is a better name than ["Lewontin's fallacy"](https://en.wikipedia.org/wiki/Human_Genetic_Diversity:_Lewontin's_Fallacy) (working title: "High-Dimensional Social Science and the Conjunction of Small Effect Sizes") + * I need to write an in-depth post about the overlap-along-one-dimension-does-not-imply-overlap-in-the-entire-configuration-space statistical phenomenon ([standard diagram](https://en.wikipedia.org/wiki/File:Pattern_classification_medium.JPG)) of which I have decided that ["univariate fallacy"](https://twitter.com/sapinker/status/1071245692180578305) is a better name than ["Lewontin's fallacy"](https://en.wikipedia.org/wiki/Human_Genetic_Diversity:_Lewontin's_Fallacy) (working title: "High-Dimensional Social Science and the Conjunction of Small Effect Sizes") * a technical post about how imperfect measurements are subject to [regression to the mean](https://en.wikipedia.org/wiki/Regression_toward_the_mean), which (unfortunately! _really genuinely_ unfortunately!) quantitatively weakens the standard reassurance of, "Oh, no one should feel threatened by discussion of group differences, because the statistics obviously don't apply to any one individual" * I haven't done any serious math in a while and I'm afraid that learning and explaining the details here could take me _many_ hours * a technical post about using [naïve Bayes models](http://lesswrong.com/lw/o8/conditional_independence_and_naive_bayes/) for sex categorization diff --git a/content/drafts/book-review-human-diversity.md b/content/drafts/book-review-human-diversity.md index 01f0fc8..e935a55 100644 --- a/content/drafts/book-review-human-diversity.md +++ b/content/drafts/book-review-human-diversity.md @@ -18,7 +18,7 @@ _Human Diversity_ is divided into three parts corresponding to the topics in the The first (short) chapter is mostly about explaining [Cohen's _d_](https://en.wikiversity.org/wiki/Cohen%27s_d) [effect sizes](https://en.wikipedia.org/wiki/Effect_size), which I think are solving a very important problem! When people say "Men are taller than women" you know they don't mean _all_ men are taller than _all_ women (because you know that they know that that's obviously not true), but that just raises the question of what they _do_ mean. Saying they mean it "generally", "on average", or "statistically" doesn't really solve the problem, because that covers everything between-but-not-including "No difference" to "Yes, literally all women and all men". Cohen's _d_ is the summary statistic that lets us _quantify_ statistical differences in standardized form: once you can [visualize the overlapping distributions](https://rpsychologist.com/d3/cohend/), whether the reality of the data should be summarized in English words as a "large difference" or a "small difference" becomes a _much less interesting_ question. -[multivariate effect sizes and the Marco del Guidice fan club, https://www.lesswrong.com/posts/cu7YY7WdgJBs3DpmJ/the-univariate-fallacy ] +Murray also addresses the issue of aggregating effect sizes—something [I've been meaning to get around to blogging about](/2018/Dec/untitled-metablogging-26-december-2018/#high-dimensional-social-science-and-the-conjunction-of-small-effect-sizes) more exhaustively for a while in this context of group differences (although at least, um, my favorite author on _Less Wrong_ [covered it in the purely abstract setting](https://www.lesswrong.com/posts/cu7YY7WdgJBs3DpmJ/the-univariate-fallacy)): small effect sizes in any single measurement can amount to a _big_ difference when you're considering many measurements at once. That's how people can distinguish female and male faces at 96% accuracy, even though there's no single measurement (like "eye width" or "nose height") offers that much predictive power. Subsequent chapers address sex differences in personality, cognition, interests, and the brain. It turns out that women are more warm, empathetic, æsthetically discerning, and cooperative than men are! You might think that this is due to socialization, but then it's hard to explain why the same differences show up in different countries—and why the differences are seem _larger_ in richer, more feminist countries. @@ -90,7 +90,7 @@ Then Appendix 3 ("Sex Differences in Brain Volumes and Variance") has tables of ... and that's the book review that I would prefer to write. A science review of a science book, for science nerds. The kind of thing that would have no reason to draw your attention if you're not _genuinely interested_ in Mahanalobis _D_ effect sizes or adaptive introgression or Falconer's formula, for their own sake, or for the sake of [compressing the length of the message needed to encode your observations](https://en.wikipedia.org/wiki/Minimum_message_length). -But that's not why you're reading this. That's not why I'm writing this. +But that's not why you're reading this. That's not why Murray wrote the book. That's not even why _I'm_ writing this. -- 2.17.1