skewed scoring...

[thanate]

So I took one of those stupid online IQ test things the other day for no particular reason, and got back the results that out of 40 questions, I got one wrong. (It was the only one I couldn't actually figure out the answer to-- one of those "how are these 3 numbers alike, but different from those other three, and pick the other number that goes in the first group" things, which it turns out were all perfect squares, a thing that was quite obvious after it had been pointed out to me)

But anyway, because it was only 40 questions, a score of 39 correct means that I "scored 142 on [their] IQ test. This means that based on your answers, your IQ score is between 132 and 142. Most people's IQs are between 70 and 130." Presumably if I had scored 40 of 40 I would be guaranteed between 160 and 142, and if there'd been another question where I didn't track their thinking, I would have been guaranteed below 132, though presumably the interval gets narrower as one approaches the middle of the score range. Which is a lot to ask of single questions... Not of course that IQ means much more than that (in this case) I happen to be good at standardized tests, and quite a few of the premises of their questions were pretty darn dumb-- alien languages that translate word-for-word to english, for instance...

(my point, in case you had lost track of it along the way, is that with 40 questions and 160 points possible, each question on an even scoring ought to get you 4 points. Instead, for missing one question, I am dropped at least 18, or more if I feel like it. Presumably this evens out in the middle, and if you were to get only one question right, you'd be granted extra points free... except don't they already give you 20 just for playing?)

I just am always reminded in these situations of how annoyed I was when I took the SATs and, after getting the same number of questions wrong on each part, had a higher math score than verbal.

Comments

[kittymaru.livejournal.com]

I took IQ tests almost every year in Elementary school. They're just... well, I really don't have words for it. They can't tell you exactly what your IQ is anyway. Not that it even really matters. They were usually 2 hr. tests and the plus to taking them was missing class and then getting lunch and recess.

[thanate.livejournal.com]

Yeah, pretty much. IQ cannot tell you whether you're a safe driver or dig good holes in the ground. Although I found it amusing that the "these are professions at which you might excel based on your IQ type" list had archaeologist at the top.

I think the only time I had a real IQ test, as opposed to the ongoing battery of long standardized things, was in second grade, where they did that with a large-ish reading comprehension element, and with timed "figure out what this word is in under 30 seconds" flash cards, and I was just below the cut-off line for the GT center tracking program. And the school principal noticed this and told my parents to take me to a reading tutor, who gave me virtually the same test untimed and said "what are you talking about, this kid has a college reading level." It was just that I'd never actually *seen* most of the words written down before.

Which, all in all, was probably a good thing, as 2nd grade was miserable, and 3rd (in the tracking program) far far better. In retrospect, I think that as a second grader, I was smarter than my 2nd grade teacher, which was part of the problem.