But I'm not gonna take a bunch of tests and use the average [...] this resolves to the first test result I get, even if it seems likely to be an over or under estimate

This seems to be at odds with what you've been doing in the comments, is there a new set of criteria you're using?

It resolves based on the first "professional IQ test", the comments are online IQ tests so they don't count towards the resolution criteria but could be useful for people betting. The comments from 8 months ago clarify this further.

Ah ok thanks

https://openpsychometrics.org/tests/VIQT/

126

https://test.mensa.no/Home/Test/

Another one. I tried to be faster this time, but still had to select random answers to 10ish questions I couldn't see a pattern in before the timer ran out.

Did significantly worse this time, 115. I wish it showed me the "right" answers, since the progressive matrices approach always makes me worry I've spotted a pattern different from the one the test-designer intended.

Oh, there's also http://antjuanfinch.com/sgiq

@IsaacKing Interesting, I just came across this market.

I took an official, real-world test when I was 10. Since then, however, I've been treated for mental illness and am able to accomplish significantly more tasks than when I was 20 years old.

I wonder if the score would go up, having been given treatment that did not exist 30 years ago. But these online tests are probably not as rigorous as the ones given in-person, which indicates to me that I probably would score higher even if I didn't actually do any better.

Just took this. Got 38/50, which it claims corresponds to an IQ of 133. That's pretty consistent with the other online test at least!

(I clearly am supposed to be answering these faster. Out of the 38 questions I answered I got them all right, but I didn't have enough time to do the last 12.)

@IsaacKing 46/50 here(not that it affects your market). I might be biased by seeing your comment though and knowing that I need to work fast. So I skipped checking answers when I usually wouldn't.

Mira just linked this, seems like fun. Should I resolve this market based on it, or is this not professional enough?

https://agctest.com/

@IsaacKing it would be nice to resolve all the IQ markets off the same test...

I'm thinking no, it's not a real test. Also I want to take it now for fun, but I'm really tried and will probably do poorly.

[Deleted, I need to learn to read things more carefully. Maybe I won't take this while I'm tired...]

Argh. Did not realize how much of a factor the time limit would be, I wasted a lot of time on a few hard questions early on. Should have just skipped them.

Frustrating that it doesn't tell you how skipped questions are scored so I didn't know if I should guess when I could narrow it down to two possibilities. Also that it doesn't tell me which ones I got wrong! Ah well.

@IsaacKing Personally, no. A huge amount of effort goes into the calibration and updating of WAIS and SB to keep them reliable, and from what I can tell, AGCT is from the 60s?

@IsaacKing I don’t see why you can’t take it twice.

[Duplicate comment, deleted]

I'm surprised people keep betting this up so high. Olivia's IQ test was 122, and from what I've seen of her on this site, I expect mine is a bit lower, at least when it comes to matters of math and programming.

@IsaacKing Why use just one data point? IIRC the average LessWrong user self-report is in the 140s. You can also look up averages by college major, for example.

@StevenK "IIRC the average LessWrong user self-report is in the 140s." 🤣 yeah and the average BuzzFeed self-report penis size was three standard deviations over the global average too. I thought LessWrong users were supposed to know about self report bias. 🤣

@bluerat I agree there's some bias there. IIRC this is self-reports of serious tests, but I don't remember for sure.

@bluerat Maybe they are so rational none of the normal biases or problematic sampling effects apply :P

@bluerat Just found this. Thanks for this comment, I was going to post just about the same reply.