Hmm. I got 110. And then because that’s ridiculous, and I have an ego, I took it a second—and third—time, subsequently scoring 126 and 140. (I reported 125 on the survey because I know 110 isn’t right.)
And while I was trying harder on the second and third attempts (as a result of realizing ‘oh, I guess most of these actually are easy to everyone else, not just me, so I shouldn’t be so leisurely’), I wasn’t superbly focused on any—for example, I became distracted on the third attempt with something in another tab for more than 10 minutes before remembering it.
Back in grade school, I took several real-life IQ tests and usually scored in the high 130′s to low 140′s. I’d heard of Raven’s Progressive Matrices, but this was the first time I’d taken that type of test. It was quite humbling. I got 122 on iqtest.dk. From what I’ve heard in #lesswrong, most people score low on this test.
I opened the test again in a different browser, VPN’d from a different country. It gave the same questions. That means your subsequent tests aren’t valid. You already knew many of the answers. Worse, you knew which questions had stumped you before. You were probably thinking about those questions before you started the test a second or third time.
Errr… Did you notice that on iqtest.dk you were allowed to skip to earlier and later questions by clicking on the numbers on the left? The first time I took that test, about a year ago, I didn’t, so I wasted plenty of time on certain difficult questions before giving up.
Yes, I noticed I could skip around. I mostly did the questions in order, since they got progressively harder. Still, I ran out of time and had to guess on the last two.
Hmm, yes. But it’s not quite that bad. 1) I re-took the test right away (less time to ‘think about the questions’), and 2) I had figured out many of the questions, but I didn’t receive actual feedback on which I got right or wrong. Anyway, I’d be surprised if my actual IQ were below 130.
It suffers the usual problems of tests, among which are that test-taking is itself a skill.
That said, I don’t think re-taking the test produces a valid result—a lot of the time I spent on the test was figuring out the rules of the puzzles as much as solving them. The problematic nature of the initial result is a reflection of the weakness of the test, as you noted, but re-taking the test simply introduces a new suite of problems.
Hmm. I got 110. And then because that’s ridiculous, and I have an ego, I took it a second—and third—time, subsequently scoring 126 and 140. (I reported 125 on the survey because I know 110 isn’t right.)
I’m pretty sure you’re not supposed to do that—IIRC tests are designed to give reasonably accurate results in absence of practice effects. I had taken this same test one year ago and I’m pretty sure I answered certain questions faster than I would have if I had never seen them before (though this effect was almost exclusively in the easy, early questions, which took a very small part of the 40 minutes anyway—I did score 9 points higher than last year but I had a headache (and hadn’t realized I could go back to previous questions) back then so that sounds about right).
Also, I failed to answer quite a few questions when I got 110, thinking I’d be penalized for wrong answers… Apparently I failed at reading the directions which state you should answer all of them facepalm
I reported my first try answer (which also seemed unrealistically low to me). I think on balance it might be best for everyone to just report their first try answer accepting the test is normed low and then for macro analysis it can be adjusted / compared with another test like SAT scores
Calibration for other people looking at this comment: I took the test and got 10 points higher than my self-reported IQ. I think it picks up on a different kind of reasoning than the usual type of IQ test!
Took it—I hadn’t taken an IQ test before, and I found it interesting (and, for the final few questions, quite difficult).
Addendum: I don’t recall seeing the question about the “troll toll” when I took the survey; my response would be that V snibe vg, ohg thneqrqyl.
Hmm. I got 110. And then because that’s ridiculous, and I have an ego, I took it a second—and third—time, subsequently scoring 126 and 140. (I reported 125 on the survey because I know 110 isn’t right.)
And while I was trying harder on the second and third attempts (as a result of realizing ‘oh, I guess most of these actually are easy to everyone else, not just me, so I shouldn’t be so leisurely’), I wasn’t superbly focused on any—for example, I became distracted on the third attempt with something in another tab for more than 10 minutes before remembering it.
All I’m saying is I’m dubious of this IQ test.
Back in grade school, I took several real-life IQ tests and usually scored in the high 130′s to low 140′s. I’d heard of Raven’s Progressive Matrices, but this was the first time I’d taken that type of test. It was quite humbling. I got 122 on iqtest.dk. From what I’ve heard in #lesswrong, most people score low on this test.
I opened the test again in a different browser, VPN’d from a different country. It gave the same questions. That means your subsequent tests aren’t valid. You already knew many of the answers. Worse, you knew which questions had stumped you before. You were probably thinking about those questions before you started the test a second or third time.
I got 135… Was I the only one who realized I could go back to previous questions, or something?
Does anyone know what IQ it gives to a perfect set of answers? Its picture of a bell curve superimposed on a scale tops out at about 140.
I got 138, and I’m fairly certain I got one question wrong. I don’t think it tops out at 140.
See this comment of mine.
SD was 15 and the tests were geared for high-IQ people. I’ve taken tests meant for average people and gotten hilarious results (163).
Errr… Did you notice that on iqtest.dk you were allowed to skip to earlier and later questions by clicking on the numbers on the left? The first time I took that test, about a year ago, I didn’t, so I wasted plenty of time on certain difficult questions before giving up.
Yes, I noticed I could skip around. I mostly did the questions in order, since they got progressively harder. Still, I ran out of time and had to guess on the last two.
Hmm, yes. But it’s not quite that bad. 1) I re-took the test right away (less time to ‘think about the questions’), and 2) I had figured out many of the questions, but I didn’t receive actual feedback on which I got right or wrong. Anyway, I’d be surprised if my actual IQ were below 130.
It suffers the usual problems of tests, among which are that test-taking is itself a skill.
That said, I don’t think re-taking the test produces a valid result—a lot of the time I spent on the test was figuring out the rules of the puzzles as much as solving them. The problematic nature of the initial result is a reflection of the weakness of the test, as you noted, but re-taking the test simply introduces a new suite of problems.
I’m pretty sure you’re not supposed to do that—IIRC tests are designed to give reasonably accurate results in absence of practice effects. I had taken this same test one year ago and I’m pretty sure I answered certain questions faster than I would have if I had never seen them before (though this effect was almost exclusively in the easy, early questions, which took a very small part of the 40 minutes anyway—I did score 9 points higher than last year but I had a headache (and hadn’t realized I could go back to previous questions) back then so that sounds about right).
Also, I failed to answer quite a few questions when I got 110, thinking I’d be penalized for wrong answers… Apparently I failed at reading the directions which state you should answer all of them facepalm
I reported my first try answer (which also seemed unrealistically low to me). I think on balance it might be best for everyone to just report their first try answer accepting the test is normed low and then for macro analysis it can be adjusted / compared with another test like SAT scores
Calibration for other people looking at this comment: I took the test and got 10 points higher than my self-reported IQ. I think it picks up on a different kind of reasoning than the usual type of IQ test!