Links 9/15: Linker Tailor Soldier Spy

Little-known footnote to the Scramble For Africa: Russian Somalia lasted one month before being conquered.

Latest big meta-analysis finds “little evidence” for Roy Baumeister’s ego depletion hypothesis, describes itself as “strongly challenging” the idea of willpower as a limited resource.

Why have murder rates suddenly shot up in St. Louis and Baltimore? The obvious thread linking those two cities doesn’t seem to be it; the rise appears to have started before the recent round of shootings and riots. Hopefully this doesn’t represent a halt or reversal of the secular trend towards decreasing crime – but if it does, then it could be another clue to what’s going on.

This week in neat methodologies: evidence in bias at Olympic games. Winners from last year gain a greater premium in subjective contests (eg figure skating with judges) than objective contests (eg the hundred meter dash), suggesting that judges are biased upward by an athlete’s previous reputation.

Headline of the week: The Weather Channel Reports Firenado On Kentucky Lake Brimming With Bourbon. I hope Sharknado 5 is about a fire-bourbon-shark-nado.

Also this week in neat methodologies: we don’t fully understand the genetics of intelligence, but we do know a couple of genes with small effect. Our model for natural selection on intelligence suggests that selection pressure should affect the various intelligence genes equally. That means the few genes we know offer what could be a representative sample of the entire structure that we could use to detect or refute such selection. And that in turn means that we can test for a genetic basis of intelligence differences between populations by checking how they differ on the few genes we know. Davide Piffer tries this and finds a correlation of 0.9 between genetically-predicted intelligence and actual IQ on the population level. But he gets correlations almost as high between randomly selected genes and actual IQ, which could either mean those randomly selected genes are linked to intelligence-related genes (his theory) or his methodology sucks (eternal default theory). Also, I am suspicious because you never get a real correlation of 0.9 with anything; even measurement error should be worse than that! Also, the author doesn’t do so well with my usual test of “does viewing his Twitter account increase my confidence in him as a person?”

This is probably a metaphor for something: Chimp Wielding Stick Takes Down Drone. Next time they should hold out for a craft with more experience dodging sudden giant animal attacks

The Marines recently did a study showing that women performed less well than men in combat, but as best I can tell it was fatally flawed; since the military has only recently started accepting women, all the women were new but many of the men were experienced. That just means experienced people do better than new people, which is unsurprising. Come back when you’ve got equivalent groups.

Bisexual women are more likely than straight or lesbian women to have an eating disorder, confusing everybody regardless of their theoretical perspective.

David Friedman’s Law’s Order starts off by explaining why there should be a higher penalty for armed robbery + murder than for armed robbery alone, or else armed robbers will have no incentive not to kill their victim and eliminate a potential witness. A recent Global Post article argues that the Chinese have screwed this one up by making a law that if you run over someone and they live, you have to pay for their medical care; but if you run over someone and they die, you have to pay a flat fine which is often much less. Therefore, Chinese drivers who accidentally hit someone will try to run them over again to finish the job. Snopes isn’t buying this and can’t find any sources, but I vaguely remember reading articles before in the Chinese-English media that were sort of like this.

A person who is not Ozy nevertheless analyzes the evidence on the results of gender transitioning, finds best studies show it can improve mental health.

The original Sokal hoax – critically acclaimed avante-garde artist Pierre Brassau turned out to be a chimpanzee.

Harold Lee is always thought-provoking and worthwhile. The current thing of Harold Lee’s that is thought-provoking and worthwhile is on the Manchus’ self-conscious attempts to avoid decadence in Imperial China and how it relates to modern immigration.

Clever in a Rube Goldberg sort of way: cows keep eating metal, which damages their digestive tract. The solution: feed them a big magnet that stays in their stomach and prevents metal objects from going further.

Doing your arithmetic in base phi (ie 1.618…) is less crazy than you’d think.

In 1968, Britain was skimping on celebrating the 50th anniversary of the RAF for political reasons. An RAF pilot decided to take matters into his own hands and fly a fighter jet around the city of London without permission, including passing through the Tower Bridge. Everyone had a big laugh about it and he was allowed to leave the RAF on “medical grounds” without penalty. Simpler times.

The street dogs of Moscow have adapted to their urban environment, including learning how to use the subway train system to get where they need to go. Also some speculation that the packs enforce on their members a policy of not defecating in highly-trafficked areas in order to avoid conflict with humans, and that they are gradually switching from a pattern of strongest dog as pack leader to one favoring the most intelligent dog. Also, they seem to all be well-fed and to avoid competing with each other for food so that everyone can get their share. Actually, is there any way we can put Moscow street dogs in charge of the government?

Study on the effect of investing in schools finds “very precise zero estimates” of achievement effects.

People complain that mining the Alberta oil sands is environmentally damaging. But the original plan didn’t involve the sort of inefficient mining techniques being employed today. The original plan was to detonate a hundred nuclear bombs over the oil sands, in order to boil the oil to the top. This is maybe the most 1950s idea I have ever seen.

Very nice small study with neat methodology finds that more intelligent people are more honest.

78% of the variance in national income today is predicted by the variance in national income in 1500 AD. Much of the effect remains even as far back as 1000 BC, insofar as we can measure that sort of thing.

I Have One Of The Best Jobs In Academia; Here’s Why I’m Walking Away. The first half kind of makes me want to punch the guy, but the second is a really good analysis of how we ended up in the higher education bubble and what we can do to stop it.

In case you hadn’t heard, Jeremy Corbyn did in fact win the Labour Party leadership. Some discussion of his policies on Reddit, which range from “actually sounds pretty good, not sure why everyone’s so worked up over this” to “legitimately terrible idea, but not obviously moreso than everyone else’s terrible ideas”.

In 1956, Mao Zedong declared a policy switch toward greater freedom of speech, and invited party officials and intellectuals to criticize him and his leadership so he could learn from their advice. In 1957, Mao declared that it had all been a trap, and killed everyone who had spoken up.

“I don’t think any currently published translations capture the original [spirit of the Iliad]. The Iliad we know today is the written form of an entire oral tradition unto itself, which ancient epic poets whose names and numbers we’ll never know invented on their feet, their genius and imagination driven forth by musical accompaniment that kept the dactylic-hexameter beat. The modern listener should find the subject matter familiar as well: drink, booty, fly ladies, stacks of treasure, macho violence, and ostentatious modes of transportation.” Lula writes the Iliad as a rap and IT IS AMAZING.

Ben Henry going over some problems with Vox’s article claiming that Donald Trump made no extra money over what he’d have earned by putting his inheritance in an index fund. Quick summary: it ignores consumption (!), doesn’t really account for taxes, and in one case suggests that he should have magically picked the best moment to invest beforehand.

On the other hand, Vox’s test for political bias is actually really clever and really good. It asks you a bunch of factual questions about politics, like “Do studies show capital punishment decreases crime?” and then sees whether your errors form a consistent pattern that flatters your preconceptions. Both authors are part of the broader rationalist/EA community.

A social worker with more experience doing actual therapy than me reiterates that not-having-trigger-warnings is not the same thing as exposure therapy and shouldn’t be billed as a mental health good.

I’m not sure I’m understanding this article right, but if true, it’s…different. Half of all Japanese universities will close or scale back their humanities and social sciences departments after the government orders them to “serve areas that better meet society’s needs.”

Homeopathy Conference Ends In Chaos After Delegates Take Hallucinogenic Drug. Didn’t something like this happen in a Robert Anton Wilson book?

Last week I wrote an article on politics where I claimed that the left was moving leftward and the right rightward. I got a bunch of angry comments, half of which thought I was crazy because both parties were obviously moving rightward, and half of which thought I was crazy because both parties were obviously moving leftward. Over at his own blog, Free Northerner makes the case for a general leftward shift. Suggest you take your comments there instead of inflicting them on me again.

Harvard University declares it “can’t afford” closed access journals and asks its researchers to support the open access system. Despite the fact that their reasoning is obviously a lie (Harvard could afford journals printed upon platinum leaf in ink made from the blood of endangered white tigers) this will hopefully be a big boon to the open-access research movement.

The 2016 presidential polling has some notable anomalies. And not just the fact that people are still supporting Trump.

The amount of First World inequality attributable to education is very low, and education-related policies are unlikely to impact inequality very much.

Inside the Chinese city where the average man has three girlfriends.

More really neat methodology: when it’s dark, it’s hard for police to see what race a driver is, making them less likely to racially profile. By comparing ratios of black drivers to white drivers pulled over, and how the ratio varies during different lighting conditions (summer, winter, daylights savings time, areas with more streetlights), researchers find the odds of a black driver being stopped relative to a white driver increase 15% when police are better able to distinguish their race.

US companies are less likely to be publicly-traded than in the past.

I FINALLY FOUND THE STUDY I HAVE BEEN WANTING SOMEONE TO DO FOR AGES. The Greater Male Variability Hypothesis suggests that although men and women have similar average IQ, men will end up overrepresented in high-IQ occupations like “physicist” because men have higher variability (and therefore more men will be at both the high and low extremes). Some studies have shown preliminary support for this idea, but of course IQ is really hard to study and frequently confounded by cultural stuff. But the hypothesis says this variability should exist for every biological trait (it’s because men have only one X chromosome, so all their recessive mutations are expressed). So the best way to study it would be to study a bunch of easy-to-study obviously biological traits and see if men have more variability among most of them. And they mostly do, in categories from blood bilirubin levels to 60-meter-dash-running times (women are more variable along other parameters, including BMI and thyroid hormone levels, but there are fewer in this column). Overall the paper seems to provide significant albeit inconsistent support for greater male variability.

If you’ve ever wanted a really perfect way to intuitively compare the size of two countries/states/continents without worrying about things like Mercator projection distortion, thetruesize.com is what’s been missing from your life until now.

Mitochondria have always been creepy. But now it’s starting to look like they talk to each other. Let’s hope they’re not chatting about how eukaryotes suck and it’s time to take over.

In these times of Syrian refugees, did you know that most of Haiti’s wealthiest families are Syrian immigrants?

Good (?) job opportunity for American women – make $100,000 + as a surrogate mother for a Chinese couple.

Speaking of which: in The Gambia, babies born in September have seven times the mortality rate throughout their life as babies born in June. Probably because in the first case Gambian mothers eat way more vegetables at the appropriate seasons. Unclear whether this is just folate or something more important; very unclear what lessons first-worlders can learn from this.