Hacker News Reader

Comment by thdrdt | Hacker News Reader

It is very hard to create unbiased IQ tests. There will always be cultural differences and differences in opportunity.

IQ tests should be taken with a grain of salt.

It also depends on what you think is important. Smart people are not always wise people.

I certainly agree with that. Nonetheless, being myself one of the "darker" skinned ethnicities, I see no offence in the statement that as a whole, darker nations seem to have lower IQ, there might be thousands of reasons for that; for some nations, colonial opressions, for some it might inherent biological reasons.

The problem with saying it, is even if you mean IQ is academically the average result on a particular test design, people understand it in common English to mean something more like 'value as a member of society'.

I dunno. There are "Raven's Progressive Matrices" which have no language component, but they have been cracked for a long time.

If you look on the net you will hear that some celebrity (say Alyssa Milano) took an IQ test and got a crazy high score.

There are just 60 of them so it is not hard to practice and I am sure there is some racket of psychologists in Hollywood who coach people to take the test and then send them to the psych across the street to get them tested.

Science should not be concerned with the interpretation of its discoveries.

Interpretation is an inseparable part of science. Take the issue of race for example. People since the 30s have repeatedly pointed out that there's no consistent biological basis for race. The concept is just flat out not biologically rooted. Yet scientists kept using it because it was convenient and they came from a worldview where race was very much a reality. Thus studies were (and continue to be) done on the basis of biological racial categories.

Not sure about races, but one can certainly tell the difference between ethnicities; I can tell apart a Russian and an Armenian for example, with more than 75% success rate. Therefore there certainly is difference between people in different geographic areas. If people have differences in their outer appearances, there might be differences (not very large, obviously) in the way their neural systems work;denying that, IMO is a very ideological position.

It's the consensus of the vast majority of relevant professionals that "humans cannot be divided into biologically distinct categories", i.e. races. That quote comes from the ASHG [1], but another relevant field is anthropology where again, the largest professional organization (the AAAS) says the same thing [2]. The AAPA concurs [3]. If you think these associations might be a little biased, someone published a national survey of anthropologists a couple years ago confirming these views [4].

Some technical caveats: So yes, on some level you can approximately group sets of people into things roughly matching modern broad racial categories on the basis of genetics. This isn't because of anything real (i.e. there's nothing all members of a group share that doesn't exist "outside" them), but simply because you can broadly categorize any large set of things, regardless of whether those categories are actualized. Race as a social construct (e.g. identifying someone as "black" or "caucasian") is absolutely, undeniably a thing. It's simply not reflecting an underlying biological reality. There are also people in the community who argue that even if racial categories aren't biologically actualized, they're still useful (or that the definition I gave previously doesn't apply), but that's a much more complicated matter for which entire libraries' worth of debate exist.

[1] https://www.cell.com/ajhg/fulltext/S0002-9297(18)30363-X [2] https://www.americananthro.org/ConnectWithAAA/Content.aspx?I... [3] https://physanth.org/about/position-statements/aapa-statemen... [4] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5299519/

Differences in opportunity, in the formative years, will actually make a person dumber, and that is likely permanent. The IQ test doesn't cause such an effect; it just reveals it.

Since tests are necessarily biased, maybe the way forward is to tune the bias to the test subjects. Then, out the window goes "standardized", at least for that interpretation of "standardized" which means everyone takes exactly the same test. (And which is never literally true, because year over year, you cannot use exactly the same test questions.)

I think tests tuned to cultural differences would never fly in America because the idea is contrary to political correctness. Everyone is equal by definition, why would they need a special test. Cultural differences, what? We are all Americans; gear the test to American culture!

And then, what if it turns out that members of a certain minority group do better on IQ tests designed for them, yet is is found that white, upper-middle class kids also do better if given those tests? Oh the horror: members of minority X are actually dumb, and the test they are given isn't actually tuned to them culturally, but nothing more than dumbed down. The conservatives would have field day.

Reply on news.ycombinator.com