4 ms·

Why don't they add IQ tests as a parameter? Instead of SATs we can identify smart individuals from bad schools. No more privilege excuses.

jacobr1 5y ago
Because those have similar uneven racial distributions.
salt-thrower 5y ago
An IQ test is just another form of standardized test, isn't it?
Rebelgecko 5y ago
IQ tests have similar characteristics to the SAT. Practice makes you better, uneven distribution across the population, etc
- chillacy 5y ago
  Do you have a source for “practice makes you better”? Last I checked only N-back benefits from practice, the rest doesn’t.
  - Rebelgecko 5y ago
    "Retest effects in cognitive ability tests" by Scharfen & friends is a good overview that mentions other studies that go into more detail and tweak different parameters (e.g. time between tests, tests that are more repetitive than others, etc). The meta-analysis found that retaking a test can add a couple points each time. So taking an IQ test 4 times on average yields an 8 point increase, almost an entire standard deviation.
    - chillacy 5y ago
      Thanks, this actually matches my initial intuition about the matter. It's been awhile since I've looked up info in this area, partly due to the fact that a lot of IQ test proponents I spoke to seemed to have a great deal of personal identity staked in its significance.
StanislavPetrov 5y ago
Because the en vogue thought among "elites" at Harvard (and elsewhere) is that there are no smart individuals. Everything is simply a function of opportunity and privilege. Its the same sort of thinking that is behind removing gifted and talented programs and higher level math courses from schools. By that "logic" any sort of standardized test or objective standard that show differences in intelligence or learning are simply measures of privilege. Objective metrics are also the only concrete way to debunk these absurd, anti-science assertions about intelligence and learning, which makes them even more of an anathema.

jacobr1 5y ago

Because those have similar uneven racial distributions.

salt-thrower 5y ago

An IQ test is just another form of standardized test, isn't it?

Rebelgecko 5y ago

IQ tests have similar characteristics to the SAT. Practice makes you better, uneven distribution across the population, etc

chillacy 5y ago
Do you have a source for “practice makes you better”? Last I checked only N-back benefits from practice, the rest doesn’t.
- Rebelgecko 5y ago
  "Retest effects in cognitive ability tests" by Scharfen & friends is a good overview that mentions other studies that go into more detail and tweak different parameters (e.g. time between tests, tests that are more repetitive than others, etc). The meta-analysis found that retaking a test can add a couple points each time. So taking an IQ test 4 times on average yields an 8 point increase, almost an entire standard deviation.
  - chillacy 5y ago
    Thanks, this actually matches my initial intuition about the matter. It's been awhile since I've looked up info in this area, partly due to the fact that a lot of IQ test proponents I spoke to seemed to have a great deal of personal identity staked in its significance.

StanislavPetrov 5y ago

Because the en vogue thought among "elites" at Harvard (and elsewhere) is that there are no smart individuals. Everything is simply a function of opportunity and privilege. Its the same sort of thinking that is behind removing gifted and talented programs and higher level math courses from schools. By that "logic" any sort of standardized test or objective standard that show differences in intelligence or learning are simply measures of privilege. Objective metrics are also the only concrete way to debunk these absurd, anti-science assertions about intelligence and learning, which makes them even more of an anathema.