top | item 29600513

(no title)

SquishyPanda23 | 4 years ago

It's not sufficient that a metric be objective. The student's astrological sign is objective. But it's not a useful measure of college readiness.

Standardized test scores do very little to predict college success when controlling for other variables. And they're trivial to game.

They're a relic of back when people believed in IQ tests were useful selection criteria. Maybe there will be standardized tests in the future that are more useful, but we don't have those yet.

discuss

order

glerk|4 years ago

> They're a relic of back when people believed in IQ tests were useful selection criteria.

IQ tests are a very reliable proxy for academic performance potential. Is this even in dispute?

SquishyPanda23|4 years ago

Yes of course it's in dispute. It's so dominant a dispute in intelligence research that it's hard to imagine anyone being familiar with research on IQ and not knowing that it's in dispute.

IQ tests are maybe ok at doing population-level correlations, but not at predicting the success of any given individual. Tons of things are correlated with intelligence, and IQ has the virtue that it's easy to measure. So it has uses for things like research studies.

But as an actual filter when you care about performance it's not great. Things like high school GPA are easier to measure and more predictive of college success.

sova|4 years ago

You know, I did a google scholar search for this "fact" and in 2005 a research paper shows that self-discipline outperforms IQ by a factor of 2.

sova|4 years ago

Can you provide evidence for this claim?

nyolfen|4 years ago

standardized test scores, that is to say IQ test scores, are by far the strongest predictors of success available in any domain of the social sciences; nothing comes close, not even wealth. obviously high scores are no guarantee of anything, but there is nothing else that compares. to disregard them is to abandon the pretense of objectivity for nepotism.

throw_m239339|4 years ago

> Standardized test scores do very little to predict college success when controlling for other variables. And they're trivial to game.

needs sources, and unbiased ones.

SquishyPanda23|4 years ago

This is a very well studied field, just do a Google search

unclebucknasty|4 years ago

Do you have unbiased sources that indicate otherwise?

garbagecoder|4 years ago

This is so dangerously drunk on Kool Aid, I can barely manage to reply. The SAT was begun in the first place to even out the different schools people came from so that elite schools couldn't just slide people in, or due to grade inflation, or whatever (and race!)

You make a distinction without a difference in this case, and what other variables are you controlling for? Class rank? That can't be gamed, you think? Rich people will always game the system, whatever it is. In many cases, the alternatives are AP exams, which aren't even available in all schools.

What criterion would you suggest? A lottery?