Comment by dvt | Hacker News Reader

Comment by dvt | Hacker News Reader

This article is pretty bad and it's kind of annoying that it (somehow) reached the top of HN. But I'll bite.

1) IQ is not a pseudoscience. Admittedly, IQ is a pretty shitty test for intelligence (as it's highly culturally-biased), but calling it pseudoscience is just dumb and shows an inherent misunderstanding of what it's supposed to do. Yeah, white supremacists are idiots and use it to justify racism, but that's a different point than what the author is making.

2) Author criticizes Trump's sharing of the "black on black" statistics arguing that it's fabricated. However, some cursory research shows that it is, in fact, not as fabricated as one might believe: https://ucr.fbi.gov/crime-in-the-u.s/2013/crime-in-the-u.s.-...

According to the above, 90% of African American homicides were perpetrated by African Americans. This can be compared with white homicides which were perpetrated by 83% whites. Obviously, the numbers are highly skewed in the infographic shared by POTUS, but to bury your head in the sand and pretend this isn't a problem in black communities is unfair and does a great social disservice.

3) It doesn't take some kind of genius to "spot the signs" of Neo-Nazism. Swastikas are kind of a tell-tale sign.

4) "Deo is a union Millwright and a 12 year member of the Furry Fandom who spent 6 months infiltrating multiple Alt-Right and neo-Nazi online groups observing and recording how they recruited, organized, and operated." -- laughable. "Infiltrating" = joining a few discord servers and posting on 4chan. Got it.

Completely worthless article devoid of any actual real social commentary, potential solutions, or plans of attack. Medium has been turning into a soapbox lately.

IQ tests are not culturally biased. This is a decades out of date canard. Some tests, like Raven's Progressive Matrices, don't even require the test giver and taker to share a language. IQ, as a statistic, washes out almost every other psychometric in predictive validity -- there are a staggering number of psychology papers which pointedly do not include IQ because whatever predictive power their metric purports to have is totally overridden by IQ. There are many subfields of intelligence and two major categories (fluid and crystal) so it's true that intelligence can't be defined as a single number, but this is a strawman since intelligence researchers have never said this.

Let's assume resources are also important in success and you use IQ "tests" as a way to allocate resources to candidates. Then, your biasing your results.

Yes, this is a legitimate problem because there are both genetic and environmental contributions to intelligence. We must be wary of eugenic arguments. After all, the easiest way to raise the average IQ of a society would be to kill everyone who doesn't meet a specified minimum IQ. A more politically palatable way would be to allocate resources away from those groups.

This isn't a matter of biasing results, though, because the metric validity of intelligence testing doesn't require that we play nice with making sure everyone has equal access. These are separate problems. Intelligence research is a scientific problem, and resource allocation is a political one.

We already know that environmental contributions to IQ are important, and a reasonable and moral person who is literate on intelligence research thinks that resource allocation to lower-performing people is critically important.

> 4) "Deo is a union Millwright and a 12 year member of the Furry Fandom who spent 6 months infiltrating multiple Alt-Right and neo-Nazi online groups observing and recording how they recruited, organized, and operated." -- laughable. "Infiltrating" = joining a few discord servers and posting on 4chan. Got it.

I mean I'm the first to laugh at Furries myself but how else would suggest infiltrating and online community other than by joining the community online?

Clearly whoever posted it and whoever is upvoting it finds it to have value. I wonder why it's enraged you so that you feel the need to say it's completely worthless.

Did you know that 99% of the murders in China are committed by Chinese people, whereas less than 1% are committed by Europeans? The Chinese must be murder-crazy!

Obviously if you want useful information about homicides broken down by race, you need to control for population segregation (among many other things). If black people are culturally and geographically separated and tend to only interact with each other, then of course they will tend to only murder each other. But for example they will also tend to only marry each other.

This argument is presented with the veneer of science and math, while failing to control for factors that anyone who took stats 101 would know about. That is a pseudo-science dead giveaway.

Reply on news.ycombinator.com