3 ms·
The best way to see if someone will perform well at a job, is to have them do that job and see. The second best way is to ask them for proof they succeeded at
by alphazard 4y ago
The best way to see if someone will perform well at a job, is to have them do that job and see. The second best way is to ask them for proof they succeeded at a similar job. That proof is usually a combination of references and a discussion about work they did at the previous job.
If you can't do any of those things, like for some recent grads, then the next best predictors of job performance are going to be psychometrics, specifically IQ and trait conscientiousness. Colleges heavily select for these traits during admissions, but less so with the actual course work. These traits are most predictive when you expect to teach a person the job on the job, with their prior experience being irrelevant.
So it's not that there is no evidence for these questions being useful. It's just that there are usually better signals on offer.
- vagrantJin 4y agoThat was a whole lot of psychobabble.
- plusminusplus 4y agoSeemed cogent to me. At the very least, it contributed to discussion more than you dismissing something using a very tired cliche.
- anonymouskimmer 4y ago> "specifically IQ and trait conscientiousness. Colleges heavily select for these traits during admissions" Out of curiousity, can you cite these claims? I mean I buy the conscientiousness, but I wouldn't guess they heavily select for it or IQ. Or at least not both together, not past a point.
- alphazard 4y agoYou can look at the standardized tests (SAT, ACT) and their respective correlations to IQ tests. I'm not claiming that colleges give IQ tests themselves, just that they care a lot about other tests which are highly correlated. There are many studies on the correlations here, and you should do your own survey of the literature if you are curious, not trust a cherry picked study posted by some guy on HN.
- anonymouskimmer 4y agoI've looked at various ones over the years. I even still have the reference for at least one (that shows a negative correlation between ACT SR+R:M+E ratio and success in college, and recommends that colleges should select against such ACT composite scores). I just wanted to also read whatever papers or articles you were thinking of. I also know that, all else held equal, super-intelligence starts to show a negative correlation with success in college if the super-intelligent individual is curricularly mainstreamed in pre-college school. So the correlation with intelligence, at least, does break down in various cases. I'm sure it does with conscientiousness as well. Whether this is at the level of admissions? I know that selective schools tend to pass on students who score wildly above their average, because the student is less likely to attend if accepted (and for college rankings purposes this looks bad).