3 ms·
Extracting apparent patterns in [test answers] is a cognitively difficult task, and is thus a reasonable metric of performing cognitively difficult tasks. Give
by AbrahamParangi 4y ago
Extracting apparent patterns in [test answers] is a cognitively difficult task, and is thus a reasonable metric of performing cognitively difficult tasks.
Given that this is what we are attempting to measure, it seems completely fine. Pointing out flaws in the test (lack of generality, non-deterministic behavior) is not actually a counter-argument so long as the test correlates with at least some outcomes that we care about.
And it does! And this is very reproducible!
- mjburgess 4y agoNo no, i mean the psychometrist is the person superstitiously using the fact that answers to their choice of questions correlate, as some measure of something. The claim that IQ tests are "cognitively difficult" is either trivial or unevidenced. Its trivial in the sense that most tasks are "cognitively difficult", that doesnt make them valid measures of cognitive difficulty. It is unevidenced in the sense that there is no theory of what "cognitive difficulty" means that can be tested. My apologies, but IQ tests really are actually pseudoscience. They're a product of a discipline which superstitiously uses GUI-statistical tooling without understanding anything abotu the scientific method, nor the statistical metrics that these tools report. An animal, and its intelligence, is a complex chaotic system whose "cognitition" or "intelligence" is by no means measured by an IQ test. And such a claim is prima fascie ridiculous from any actual biological (ie., theory-building, scientific) study of intelligence.