4 ms·

it was always my understanding that "good" IQ tests should have almost zero variance based on your vocabulary, or hell, even your mathematical abilities. A "goo

by viggity 13y ago

it was always my understanding that "good" IQ tests should have almost zero variance based on your vocabulary, or hell, even your mathematical abilities. A "good" IQ test simply tests your logic facilities. Example "good" questions would show a sequence of 6 shapes and ask you what the next one in the series should look like. or perhaps a question like "given that all boogles are quinks and some quinks are flets, are all flets boogles?". A question on a "bullshit" IQ test would be: "'Daring' is an appropriate adjective for: A. Rocks. B. Planets. C. Humans.". Another would be "x^2 = 36, solve for x" Now, I know that some standardized tests (ACT, SAT, IBTS), will sometimes try to estimate your IQ by correlating your scores, but that is fraught with problems and shouldn't be taken as gospel.

foobarbazqux 13y ago
> A "good" IQ test simply tests your logic facilities. You can dramatically improve your logic facilities by taking a course in logic. Your boogles, quinks, and flets question is trivial if you know even rudimentary set theory.
- Smaug123 13y ago
  Not even that - it's the kind of question that becomes dramatically easier to answer if you know how to trick your brain into thinking it's easy. "Given that all cats are animals and some animals are mice, are all mice cats?"
  - foobarbazqux 13y ago
    It's actually a harder question if you phrase it with cats, animals, and mice! The correct answer is "possibly" (either yes or no), but the cats, animals, and mice question predisposes you to answer with "no".
    - Smaug123 13y ago
      For me, that kind of comes under the category of "know how to answer the question": that is, "remember that you have produced an example, which does not constitute proof but merely evidence". You're right that it's still not easy to distinguish between "definitely yes" and "not {definitely no}", though I do still find it easier with nouns-which-I-know than made-up-words, I think. I should have put in an attempt to make the statement true - maybe "Given that all cats are cats, and some cats are cats, are all cats cats?"
jre 13y ago
I think logical reasoning can be improved by training too. In my first year in CS at university, we had a logic course. I was definitely better at solving that kind of problems after the course. They usually become easy once you formalize them and you start to recognize patterns once you've seen enough.
icelancer 13y ago
The WAIS-III was multimodal when I took it. Most of it was as you described, but some of it was general knowledge testing as well. Overall it was a fun experience since I like taking tests, but I felt it didn't accurately judge anything, regardless of my satisfaction with the result.

by viggity 13y ago

foobarbazqux 13y ago

> A "good" IQ test simply tests your logic facilities. You can dramatically improve your logic facilities by taking a course in logic. Your boogles, quinks, and flets question is trivial if you know even rudimentary set theory.

Smaug123 13y ago
Not even that - it's the kind of question that becomes dramatically easier to answer if you know how to trick your brain into thinking it's easy. "Given that all cats are animals and some animals are mice, are all mice cats?"
- foobarbazqux 13y ago
  It's actually a harder question if you phrase it with cats, animals, and mice! The correct answer is "possibly" (either yes or no), but the cats, animals, and mice question predisposes you to answer with "no".
  - Smaug123 13y ago
    For me, that kind of comes under the category of "know how to answer the question": that is, "remember that you have produced an example, which does not constitute proof but merely evidence". You're right that it's still not easy to distinguish between "definitely yes" and "not {definitely no}", though I do still find it easier with nouns-which-I-know than made-up-words, I think. I should have put in an attempt to make the statement true - maybe "Given that all cats are cats, and some cats are cats, are all cats cats?"

jre 13y ago

I think logical reasoning can be improved by training too. In my first year in CS at university, we had a logic course. I was definitely better at solving that kind of problems after the course. They usually become easy once you formalize them and you start to recognize patterns once you've seen enough.

icelancer 13y ago

The WAIS-III was multimodal when I took it. Most of it was as you described, but some of it was general knowledge testing as well. Overall it was a fun experience since I like taking tests, but I felt it didn't accurately judge anything, regardless of my satisfaction with the result.