|Comments on Friday 5 December 2003:|
|The other day, I got a request to do and review an IQ test, based on my having done so before... Also that I could get the extended report free. Well-timed to coincide with my being unproductive and aimless, I opted to play along. The request was pleasantly weird, too, after all.|
The test is this one. Fairly good design - it includes a time factor, and an "I don't know" option which presumably scores better than a wrong answer. There weren't very many of my IQ test pet hate questions, the "which of these is the odd one out?" variety.
On the negative side, the test didn't have very many questions at all. Thirty questions is a bit limited when you're aiming to be judging ten distinct characteristics, and rather susceptible to the results being thrown off a lot by just one random slip of the brain.
Results-wise, it's not bad; the score is about what I would usually get on a live-normalised scale, scoring me at 136; the extended report suggests that my weak point is classification, which is fair enough since that's the category that features the stupid odd-one-out questions. The extended report also tells one the correct answers, though it was lacking an explanation, which would be a nice feature for some of the questions (again, particularly the stupid odd-one-out questions - I know what my reason is for saying something is different, but what reason does Mr Test have for selecting a different difference?)
There should be a test for this sort of question - take away the correct answer, and pick the new odd-one-out. If there is no new answer, the original question is okay. Otherwise, add the original correct answer back in - is the reason for the other correct answer invalidated? If not, the question should not be allowed, as there are two perfectly valid odd ones out. It's my opinion that more than half of such questions would be found broken, under this simple test.
Anyway, as IQ tests go, this one's fairly good. The extended 'personalised' report is rather impersonal, but what can you expect from answering 30 questions? "If you were a bird you would be a chaffinch." [06:09]
|the last few IQ tests i took i scored from 143-149 (one away from officially genius :) ) i find the odd one out questions pleasingly humorous|
|And yet all those IQ points still aren't enough to enable you to operate the shift key. How cruel life is.|
|A hundred and forty odd IQ, zero little fingers. But wait - then how did you type a colon for the smiley?|
|haha<\/A>/ - im stupid|
|I am officially in shock. Why did no one tell me I was a genius? Are you even allowed to be a genius when you're fourteen?|
|Strangely I almost never score less than 120 or so, yet got 98 on this. Mabey I'm just having a bad morning.|
|Odd. I got a 127, which is some 40 points below my last 'official' IQ test score. Are the test questions randomy selected from a larger list or always the same? I saw several of those lovely 'find the odd one out' questions. Also, I put "I don't know" one two of them, yet it gave numbers for number right, wrong, and answered, and it implied that I had answered all questions and did not appear to distinguish between omissions and actually incorrect answers.|
|further testing... I retook it, this time answering randomly the ones I had previously omitted, and received a 136. Also, the standard deviation given was the same (15), which is far too high a number at any rate. If IQs are distributed normally, as it claims, 15 would be the top tiny fraction of a percentile.|
|Ah, sorry to triple-post....|
To answer your question, Attack of the Spam: a real IQ test is only said to be accurate at age 18; that's what the test is calibrated for.
I forgot to mention, I also entered a higher age than my actual for my retest. This means that either the test does not compensate you for age, or I was extremely lucky at guessing, able to overcome the age handicap and still add 9 points to my score.
Commenting on the questions- many of these would not be considered fair for an actual IQ test. References to specific cities, for example, and expecting you to answer a question based on said knowledge, is not acceptable; the question asking which series of letters could be made to spell a word was also not a 'good' measure of IQ, as IQ tests are not meant to be biased towards native speakers of the language in which they were issued.
|agh, quadruple-posting.... silly me, I'm obviously a bit too tired for my own good. Disregard the random tense shift at the end of my last sentence, as well as any other errors in my above messages.|
|Apparently the age handicap goes the other way, Soli - the televised national IQ test of Australia, at any rate, gave a bonus 5 IQ points for being 25 or older. Being under 18 also confers a bonus - being exactly 18 is the handicap. I don't like the age handicap existing - IQ should be a measure for comparing against everyone, not "well I was probably as good as you when I was in my prime".|
The standard deviation given isn't of your result, but of the test as a whole, which is why it didn't change when your result did.
An IQ test may be biased towards native speakers of the language if it's only being issued to native speakers of the language, since then all contestants are on the same footing - especially if the test is being standardised on-the-fly. It is an unfair test to administer to non-native-speakers, but that doesn't make it flawed unless it *is* administered to non-native-speakers.
Not that I'm suggesting that this test is unflawed - just that your objections aren't necessarily right.
|Bonus 5 IQ at some scores, that is. The bonus varied depending on score, but it was always a bonus for being older, not a handicap to overcome.|
|IQ is all about a gift given to an individual by almighty, and should not be under-rated or dis-respected as misunderstood. No inventions in science would possible without it. Thumbs-up IQ especially if 250 or above.|
|You are dim.|