Hacker News Reader

Comment by kevinh | Hacker News Reader

> For example, try posting FBI murder rate statistics by race of perpetrator and victim to any forum where political correctness is highly active. Instant downvotes and censorship to maintain ideological hygiene. This is the constant positive pressure that is needed to alleviate the friction between reality and political correctness.

I find it hard to believe that no one has mentioned the factors that make murder rate statistics misleading to you. At the very least, that it's being put out there with the implication that correlation is causation, and ignores other factors that have much more basis for a differing murder rate, like socioeconomic status.

It's probably the set of information that's put out there the most often in order to muddy the waters, often by people have racist motivations.

View on HN

And you instantly went to assume the intentions of the author and their lack of knowledge, etc?

The author said "posting the data" is enough for backlash, they didn't say anything about judging the data, let alone that they'd judge the data like you say they would.

And even if they would draw "misleading" conclusions from the data, it is not helpful to rant at them and punish them, but rather assume their good faith (until proven otherwise) and explain to them why they were mislead by the data. Even if it's a bad white supremacist who'd never change their mind, you might still positively influence other people reading the same exchange instead of making other people root for the "underdog" (in that conversation) who was viciously and personally attacked just for posting data.

Yes, because that would be a reasonable explanation for the downvotes.

Most of the time people talk about “black on black crime” they’re not trying to make a sound statistical argument, they’re using it in service of some larger racist point. So yeah, when you start talking about “what about Chicago” or the shape of people’s heads, people tend to put you in the same bucket at everyone else who seems to be really interested in certain facts...

When the last hundred people you've seen post that data do so in order to apply a racist judgement, bayesian reasoning says that this person is likely doing the same and engaging seriously isn't going to go well.

This is a poor example for making that point, in the sense that it lacks a moment of realization or revelation. All this example carries is a technicality of yet-to-be-inidentified faith, but not even a hint of justification for considering the contrary. An open mind is something to test and critique but not merely punish or condemn, as this example attempts. A human should be expected to act on probability and defend against risk to the best of their ability and awareness, and it is the point’s failure to sympathize with that which makes it moot.

Well how about the friction that occurs when a crackdown on crime is viewed as racist because more people who commit murders are arrested which statistically includes more minorities? The flip side of this is fewer minorities, who are much more likely to be murdered by other minorities, are murdered by those minorities. So the people who benefit the most from these crackdowns are minorities who are victims of crime perpetrated by other minorities. Thus, this policy is actually beneficial to law abiding minorities, though it is viewed as against their interest by politically correct advocates who refuse to acknowledge the actual statistics of crime. They are ideologically obliged to ignore the FBI statistics and to state that a policy that arrests more minorities is thus racist because the arrest rate of all races are not equal. This exact thing actually happened in New York City with the Giuliani administration when they had the big crime crackdown in the 90s.

Giuliani's record is mixed. http://www.politifact.com/truth-o-meter/article/2007/sep/01/...

That's the big problem with censorship. If op would post the true statistics that black crime rate is higher than other races then op would be downvoted. op would then go on to think that they are right, and that the political opponents are afraid of having their political bias exposed. If however op was not downvoted for making an innocent observation, but rather was told that the statistic could be used to show a higher correlation between poverty and crime, maybe op, and other readers, would change their mind about the situation.

The problem with this approach is that if op was trolling then they could spit out a huge list of seemingly correct correlations that nobody would have time to debunk all of them. That would allow some of op's bad correlation to be left not debunked and appear that it is right.

In online forms you have to figure out if someone is trolling and then behave in one of two opposite ways based on it they are out aren't.

> I find it hard to believe that no one has mentioned the factors that make murder rate statistics misleading to you. At the very least, that it's being put out there with the implication that correlation is causation, and ignores other factors that have much more basis for a differing murder rate, like socioeconomic status.

Absolutely socioeconomic and cultural factors affect the statistics. But they also affect statistics like representation in the tech industry.

If someone objects to bringing up one then the same person should object equally strenuously to bringing up the other. Which is not what happens in practice.

>It's probably the set of information that's put out there the most often in order to muddy the waters, often by people have racist motivations.

Lower cognitive ability is related with racism[0]. So it might not be malice, but simple that they cannot grasp the social underpinnings of such statistics.

[0] http://www.fraw.org.uk/library/anarchism/hodson_busseri_2012...

Reply on news.ycombinator.com