←back to thread

American IQ scores have rapidly dropped, proving the 'Reverse Flynn effect'

(www.popularmechanics.com)
256 points hirundo | 5 comments | 10 Apr 23 13:00 UTC | HN request time: 0.001s | source
Show context
rahimnathwani ◴[10 Apr 23 16:58 UTC] No.35514446[source]▶
This blog post asserts that IQ scores didn't drop for the population as a whole, and that the drop for each individual group is due to changing composition of that group:

https://emilkirkegaard.dk/en/2023/03/new-study-didnt-really-...

For example, if the % of people who do a postgraduate degree goes doubles, it's no longer such a select group, so you'd expect the average IQ of postgraduate degree holders to go down. This doesn't mean IQ scores are going down for the population as a whole.

One more thing: why do so many papers that present charts that show how a mean or median changes over time, without also presenting charts that show how the distribution has changed over time?

replies(6): >>35514708 #>>35515280 #>>35517739 #>>35518020 #>>35518556 #>>35519141 #
tptacek ◴[10 Apr 23 21:42 UTC] No.35517739[source]▶
It's worth looking up whose blog this is before trusting any of its analysis.
replies(7): >>35517869 #>>35517966 #>>35518072 #>>35518112 #>>35518249 #>>35518570 #>>35518709 #
rahimnathwani ◴[10 Apr 23 22:05 UTC] No.35517966[source]▶
I'd never heard of this person before finding this blog post via Google. So I trust the post as much as I trust any random blog post that seems to make a reasonable argument :)
replies(2): >>35518005 #>>35518077 #
tedivm ◴[10 Apr 23 22:11 UTC] No.35518005[source]▶
You trust one of these random people enough to promote it. It's not unreasonable for people to point out that the author is quite infamous for his viewpoints.
replies(2): >>35518116 #>>35518200 #
nostrademons ◴[10 Apr 23 22:33 UTC] No.35518200[source]▶
It's an ad-hominem (in the original sense of the word: an ad-hominem fallacy is one where the truth or falsity of an argument is determined by the trustworthiness of its proponent rather than by the content of the argument itself).

Is he right? The argument is plausible: the study measures online IQ tests. Certainly in my experience the average person online has gotten dumber in the 30 years I've been on the Internet, because Internet access has expanded and it's now the general population rather than just upper-middle-class academics. But we'd need to see comparisons vs. offline IQ tests, given to a randomly-sampled selection of the population, to be sure.

replies(1): >>35518338 #
Dylan16807 ◴[10 Apr 23 22:49 UTC] No.35518338[source]▶
It's not a fallacy to attack someone's historical reliability in making arguments.
replies(7): >>35518575 #>>35518578 #>>35518579 #>>35518583 #>>35518606 #>>35519557 #>>35523808 #
Retric ◴[10 Apr 23 23:15 UTC] No.35518578[source]▶
It very much is a fallacy to attack anything outside the argument.

That said while broken clock could happen to be correct, pointing out a clock is broken or someone is a nutter is still useful information.

replies(1): >>35519767 #
1. Dylan16807 ◴[11 Apr 23 01:58 UTC] No.35519767[source]▶
Logical fallacies undermine the argument of the person using them.

How can it be a fallacy and useful information?

replies(1): >>35520132 #
2. Retric ◴[11 Apr 23 02:47 UTC] No.35520132[source]▶
You can use imperfect information. It’s the difference between heuristics and logical arguments.

Aka there’s a 1 in 500 million chance this is a winning lottery ticket is useful even if it doesn’t guarantee things one way or the other.

replies(1): >>35520300 #
3. Dylan16807 ◴[11 Apr 23 03:11 UTC] No.35520300[source]▶
I don't think it's a fallacy to use imperfect information.

All information is imperfect.

replies(1): >>35522467 #
4. Retric ◴[11 Apr 23 08:40 UTC] No.35522467{3}[source]▶
Using probabilistic logic is perfectly reasonable as long as you track the stacking likelihood it’s wrong. However, making a probabilistic argument without tracking the internal probabilities is a Fallacy.

ie: Saying A which is true 90% of the time and B which is true 90% of the time so given A and B then C sounds reasonable until you continue with given C and D which is also X% true…. The individual probability that A, B, or D is wrong makes the chain of reasoning rapidly worthless.

Thus the error is using probabilistic statements without acknowledging they weaken an argument.

replies(1): >>35524118 #
5. Dylan16807 ◴[11 Apr 23 12:19 UTC] No.35524118{4}[source]▶
Are you saying that "You should know A is true 90% of the time." is a fallacy unless I add an explicit disclaimer about probabilistic weakening? Even if I'm not making any conclusions based on it, just suggesting people will want to use that information themselves?

If you're not saying that then I don't understand how anything you just said is relevant to the current situation. Telling people to look up a guy's writing history is probabilistic but it isn't turning that into an unjustified boolean statement.