4 ms·

Yes, yes, a thousand times yes. Years ago in our 20s my wife and I took an IQ test. I think she got in the 150s and I scored in the upper 140s. But I felt I h

Yes, yes, a thousand times yes. Years ago in our 20s my wife and I took an IQ test. I think she got in the 150s and I scored in the upper 140s. But I felt I had done much better. When we looked at the scores, for several answers I had results that were clearly consistent with an internal model I could build of the problem, but it just wasn't the model the tester wanted. So in a geometric progression, I noticed that each figure had an odd number of visible vortices. In a word analogy, I noticed that all of the words except one had an Old English root. And so on. Many times there were multiple patterns of resolution. How could one be better than another? I could certainly understand her choices, I just thought they were no worse or better than mine. The fact that IQ tests are created by people with high IQs led me to believe that what is really being measured here is some sort of academic cultural cohesion; how much we begin to solve problems in similar ways? Not necessarily a bad idea, but not what I think of when somebody says "intelligence test"

carey 14y ago
A properly _standardised_ test should at least have been administered to enough people that you were in the minority of the kind of people it was standardised against. Actually, a properly standardised test should have discarded questions like that if their answers weren’t reliable. In any case, it is quite unlikely that the test you took could reliably distinguish between scores above 145.
- cbsmith 14y ago
  Couldn't you argue part of the test is whether the test taker can determine the test maker's thinking? ;-)
matthewowen 14y ago
Did any of the other word questions have answers based upon the linguistic properties of the words (rather than on the meanings of the words)? If not, then you evidently did fail at that portion of pattern recognition.
- JoeAltmaier 14y ago
  Be fair - for math patterns you're expected to choose the simplest fit. For language, the highest-order fit (e.g. meaning, not letter count). Why? Its cultural - thats how IQ tests are done.
  - matthewowen 14y ago
    That's true, but I don't see why it is a problem. You typically get example questions in these sorts of tests - it should be trivial to understand the types of fit that the test is requesting. To be frank, I have very little time for people who claim to have overthought a test. Typically, they've thought enough to complicate the simple answer, but haven't thought enough to realise why the simple answer is correct.
    - JoeAltmaier 14y ago
      Its not always possible to know which kind of question you're answering. What is the odd one in this set? Is that a math question, or language? How far along you are in the test can change the desired outcome; I'm certain the strategy being tested switches mode as the test advances to cover more capable subjects. Short answer: you have to be familiar with the culture of testing to do well. Anecdote: I got 95+ on the GRE entering graduate school; I took advanced subject tests in subjects I had never studied. All it took was a knowledge of how the test worked.

carey 14y ago

A properly _standardised_ test should at least have been administered to enough people that you were in the minority of the kind of people it was standardised against. Actually, a properly standardised test should have discarded questions like that if their answers weren’t reliable. In any case, it is quite unlikely that the test you took could reliably distinguish between scores above 145.

cbsmith 14y ago
Couldn't you argue part of the test is whether the test taker can determine the test maker's thinking? ;-)

matthewowen 14y ago

Did any of the other word questions have answers based upon the linguistic properties of the words (rather than on the meanings of the words)? If not, then you evidently did fail at that portion of pattern recognition.

JoeAltmaier 14y ago
Be fair - for math patterns you're expected to choose the simplest fit. For language, the highest-order fit (e.g. meaning, not letter count). Why? Its cultural - thats how IQ tests are done.
- matthewowen 14y ago
  That's true, but I don't see why it is a problem. You typically get example questions in these sorts of tests - it should be trivial to understand the types of fit that the test is requesting. To be frank, I have very little time for people who claim to have overthought a test. Typically, they've thought enough to complicate the simple answer, but haven't thought enough to realise why the simple answer is correct.
  - JoeAltmaier 14y ago
    Its not always possible to know which kind of question you're answering. What is the odd one in this set? Is that a math question, or language? How far along you are in the test can change the desired outcome; I'm certain the strategy being tested switches mode as the test advances to cover more capable subjects. Short answer: you have to be familiar with the culture of testing to do well. Anecdote: I got 95+ on the GRE entering graduate school; I took advanced subject tests in subjects I had never studied. All it took was a knowledge of how the test worked.