286
smort
(sh.itjust.works)
The purpose of IQ is to measure some sort of “g factor” which is a model of “general intelligence.” This was based on the idea that people who tend to do good at some kinds of tests tend to also be good at other kinds of tests.
The IQ test is “reliable” - ie its consistent and you’ll usually get the same results +/- an acceptable amount every time. However, there are lots of concerns about its “validity” - whether it measures what it purports to measure - ie, the “g factor.”
Of note is the “Flynn effect” - that performance on the test in the general population has been improving over time, so the test has to be renormalized. (IQ is a “normalized” test - so about 68% of the population needs to be within 1 standard deviation of the mean. I think standard deviation is about 15 - so 68% of people are going to score between 85 and 115.)
The question then would be - are people getting “smarter” or is it just that people are more adapted to taking tests on pattern recognition and mathematics/logical thinking? How would that measure the intelligence of a tribal person who has not seen abstracted geometrical shapes?
You can bring in alternative models of intelligence - like Gardner’s multiple intelligence - but then that doesn’t really have much of the psychometrics behind it.
(In general, I think a huge issue in psych research is a lack of critically examining the validity of psychometric instruments. It seems we often stop at being reliable.)
