4 ms·
You have good points there. From reading the article, the inaccuracy is by inappropriately classifying people into groups, and the groupings lack sensitivity.
by singingfish 6y ago
You have good points there. From reading the article, the inaccuracy is by inappropriately classifying people into groups, and the groupings lack sensitivity. As you also point out the underlying theoretical framework is also wrong. The article does point out that the Myers Briggs does correlate reasonably well with four of the big 5 though. Having said that I would definitely discourage use of the Myers-Briggs.
The original author has a good point about the naming of the big 5 factors though.
Two things I take issue with though - IQ is reasonably well understood in that we know it measures something reliably and accurately. Exactly what that thing is though, is not really understood - and probably can't be. And how it relates to the real world is also somewhat poorly understood, and may not be possible to fully understand. But it does have good ecological and construct validity. I'm talking about good IQ tests here - any less validation than the huge amount that the Wescheler scales go through and the validation is questionable. I'm aware of the racial biases in IQ testing which is a significant problem.
Psychology isn't a particular young science - big chunks of modern statistical analysis have their roots in 19th and early 20th century psychology. So psychology is a young science in the same sense that modern statistics is. Back when I was studying psychology 20 to 30 years ago, a lot of what is now called "neuroscience" was then called "cognitive science" or more rarely "cognitive neuroscience". Neuroscience sounds sexier though.
I do have a massive issue how many social scientists abuse psychometric techniques in inappropriate ways. Management research is particularly guilty here, and I've been pressured into doing this kind of inappropriate analysis in the past.
- enkid 6y agoSo IQ measures something, but we don't know what and we don't know how it impacts the real world. How can we meaningfully say it measures something then?
- Twisol 6y agoWithout having a horse in a race, it's probably the idea that the results persist between multiple tests and vary by who's taking the test. Persistence seems like a decent lower standard to say that there's _something_ present, even if you're not quite sure what it is.
- singingfish 6y agoWell, for example, you can definitely see variations in people's short term memory capacity, basic facilities with dealing with language, and logical problem solving skills. And importantly from a clinical perspective you can see when these are mismatched. However these are very reductive variables, and in the real world there are many more things in interacting with each other than these isolated components.
- silvestrov 6y agoFor many, many years we could not explain why gravity exists and how it works. We could only experience it. We knew of gravity much before we knew of electricity. Try explain electricity to somebody in year 1300 and you'd be burned as a witch. And yet we understood electricity much earlier than gravity. How electricity works is much simpler to explain than gravity.
- mewpmewp2 6y agoWe do know what it measures and we do know what impact it has on real life scenarios.
- singingfish 6y agoTo some extent, yes. But I think it's better to emphasise the fact that we're ignorant rather than what we know because people really like abusing IQ test results.
- fmajid 6y ago> “intelligence is whatever my test measures” — Alfred Binet, inventor of the IQ test
- notahacker 6y agoThe main issue is we know IQ is meaningfully correlated with real world outcomes, but also that real world environments are meaningfully correlated with ranges of (non-normalised) IQ in a way we can't adequately explain or replicate. It's less obviously correlated with prior knowledge than most tests, but still very trainable (especially if you're a neural network with no generalised reasoning ability whatsoever...) And like all types of performance test the "something" it measures is a vector of pattern matching ability (both the component generalisable to solving more difficult problems and the component that doesn't except in the domain of career-relevant tests) but also other things obviously correlated with future performance but not fitting most people's definition of intelligence (attitude towards test, comfort in guessing, time management in timed tests, and to a degree experience), so whilst we know what the scale of the apparent impact is, we don't know really know what's behind the cause