Keep in mind that to get an average of 146 you need an implausibly huge number of >146 IQ people to balance the <146 people.
No, because there aren’t an implausibly large number of people on the list. The world is a big place. The main issue in maintaining a high average isn’t in getting the numbers of high IQ people, but in repelling the lower IQ people. But apparently, Mission Accomplished.
Note further that I was taking the 146 number as the highest reported estimate, to get the most “implausible” number, which was a mere 1/1000, and not really that rare. The 2013 survey had 138, which is 1⁄177, which is thoroughly unexciting as implausibly rare snowflakes.
It is well known and well documented that values such as IQ (or penis size) are incorrectly self reported.
Is that documented for the 146+ crowd?
I’m going by numbers I had in highschool, on two IQ tests in consecutive years which gave the same result, along with an SAT result which mapped even higher (I reported the IQ score).
It’s not too hard to remember a number, and people interested, and indeed, proud of their results, are likely paying more attention.
The one real issue I see is sampling bias—only around a third of respondents gave an IQ or SAT score, and I would expect those giving scores to skew higher.
Then again, there are probably biases associated with posting and being active as well, with the higher IQ being more confident and willing to post.
The time on LW correlated negatively with IQ… (and getting the high IQ people to come is difficult). You don’t get to invite the whole world.
Note further that I was taking the 146 number as the highest reported estimate, to get the most “implausible” number, which was a mere 1/1000, and not really that rare.
It is still rarer than many other things, e.g. extremely overinflated self assessment is not very rare.
The 2013 survey had 138, which is 1⁄177, which is thoroughly unexciting as implausibly rare snowflakes.
Well, yeah.
Is that documented for the 146+ crowd?
One can always special plead their ways out of any data. There’s two types of IQ score, one of them is about mental age, by the way.
Yes; the OP has a link to the 2013 survey data in the last line. Also note survey results for 2012, 2011, and 2009. Here’s my comment on this year’s describing what happened last year, and while this is relevant I have a memory of looking at the data, making a graph, and calling it ‘trapzeoidal,’ but I don’t know where that is, and I don’t see the image uploaded where I probably would have uploaded it- so I guess I never published that analysis. Anyway, I recommend you take a look at it yourself.
Dunno, maybe. In any case ‘repelling lower IQ people’ hypothesis seems like it ought to yield a corresponding correlation between IQ and participation, but the opposite or no correlation is observed. (albeit the writing clarity here is quite seriously low—using private terminology instead of existing words, etc. which many may find annoying and perhaps inaccessible)
No, because there aren’t an implausibly large number of people on the list. The world is a big place. The main issue in maintaining a high average isn’t in getting the numbers of high IQ people, but in repelling the lower IQ people. But apparently, Mission Accomplished.
Note further that I was taking the 146 number as the highest reported estimate, to get the most “implausible” number, which was a mere 1/1000, and not really that rare. The 2013 survey had 138, which is 1⁄177, which is thoroughly unexciting as implausibly rare snowflakes.
Is that documented for the 146+ crowd?
I’m going by numbers I had in highschool, on two IQ tests in consecutive years which gave the same result, along with an SAT result which mapped even higher (I reported the IQ score).
It’s not too hard to remember a number, and people interested, and indeed, proud of their results, are likely paying more attention.
The one real issue I see is sampling bias—only around a third of respondents gave an IQ or SAT score, and I would expect those giving scores to skew higher.
Then again, there are probably biases associated with posting and being active as well, with the higher IQ being more confident and willing to post.
The time on LW correlated negatively with IQ… (and getting the high IQ people to come is difficult). You don’t get to invite the whole world.
It is still rarer than many other things, e.g. extremely overinflated self assessment is not very rare.
Well, yeah.
One can always special plead their ways out of any data. There’s two types of IQ score, one of them is about mental age, by the way.
I thought we discovered this was driven by outliers in people who spent very little time on LW. (I’m on my phone, or I would check.)
Do you have any recollections on the source for that discovery?
Is the full survey data available, so that we could look at the distribution?
Yes; the OP has a link to the 2013 survey data in the last line. Also note survey results for 2012, 2011, and 2009. Here’s my comment on this year’s describing what happened last year, and while this is relevant I have a memory of looking at the data, making a graph, and calling it ‘trapzeoidal,’ but I don’t know where that is, and I don’t see the image uploaded where I probably would have uploaded it- so I guess I never published that analysis. Anyway, I recommend you take a look at it yourself.
Dunno, maybe. In any case ‘repelling lower IQ people’ hypothesis seems like it ought to yield a corresponding correlation between IQ and participation, but the opposite or no correlation is observed. (albeit the writing clarity here is quite seriously low—using private terminology instead of existing words, etc. which many may find annoying and perhaps inaccessible)