These statements don’t necessarily contradict each other. Even if average-intelligence people don’t read Less Wrong, perhaps they could. Personally, I suspect it’s more because of a lack of interest (and perhaps a constellation of social factors).
I bet the average LessWrong person has a great sense of humour and feels things more than other people, too.
Seriously, every informal IQ survey amongst a group/forum I have seen reports very high IQ. My (vague) memories of the LessWrong one included people who seemed to be off the scale (I don’t mean very bright. I mean that such IQs either have never been given out in official testing rather than online tests, or possibly that they just can’t be got on those tests and people were lying).
There’s always a massive bias in self-reporting: those will only be emphasised on an intellectual website that starts the survey post by saying that LessWrongers are, on average, in the top 0.11% for SATs, and gives pre-packaged excuses for not reporting inconvenient results - “Many people would prefer not to have people knowing their scores. That’s great, but please please please do post it anonymously. Especially if it’s a low one, but not if it’s low because you rushed the test”, (my emphasis).
If there’s a reason to be interested in average IQ beyond mutual ego-massage, I guess the best way would be to have an IQ test where you logged on as ‘Less Wrong member X’ and then it reported all the results, not just the ones that people chose to share. And where it revealed how many people pulled out halfway through (to avoid people bailing if they weren’t doing well).
Selection bias—which groups and forums actually asked about IQ?
Your average knitting/auto maintenance/comic book forum probably has a lower average IQ but doesn’t think to ask. And of course we’re already selecting a little just by taking the figures off of web forums, which are a little on the cerebral side.
True. I don’t think I can define the precise level of inaccuracy or anything. My point is not that I’ve detected the true signal: it’s that there’s too much noise for there to be a useful signal.
Do I think the average LessWronger has a higher IQ? Sure. But that’s nothing remotely to do with this survey. It’s just too flawed to give me any particularly useful information. I would probably update my view of LW intelligence more based on its existence than its results. In that reading the thread lowers my opinion of LW intellgence, simply because this forum is usually massively more rational and self-questioning than every other forum I’ve been on, which I would guess is associated with high IQ, and people taking the survey seriously is one of the clearest exceptions.
BTW, I’m not sure your assessments of knitting/auto maintenance/comic books/web forums are necessarily accurate. I’m not sure I have enough information on any of them to reasonably guess their intelligence. Forums are particularly exceptional in terms of showing amazing intelligence and incredible stupidity side by side.
If there’s a reason to be interested in average IQ beyond mutual ego-massage, I guess the best way would be to have an IQ test where you logged on as ‘Less Wrong member X’ and then it reported all the results, not just the ones that people chose to share.
Would still suffer from selection effects. People that thought they might not do so well would be disinclined to do it, and people who knew they were hot shit would be extra inclined to do it. The phrase “anonymous survey” doesn’t really penetrate into our status-aware hindbrains.
Better: randomly select a group of users (within some minimal activity criteria) and offer the test directly to that group. Publicly state the names of those selected (make it a short list, so that people actually read it, maybe 10-20) and then after a certain amount of time give another public list of those who did or didn’t take it, along with the results (although don’t associate results with names). That will get you better participation, and the fact that you have taken a group of known size makes it much easier to give outer bounds on the size of the selection effect caused by people not participating.
You can also improve participation by giving those users an easily accessible icon on Less Wrong itself which takes them directly to the test, and maybe a popup reminder once a day or so when they log on to the site if they’ve been selected but haven’t done it yet. Requires moderate coding.
I would find such a feature to be extraordinarily obnoxious, to the point that I’d be inclined to refused such a test purely out of anger (and my scores are not at all embarrassing). I can’t think of any other examples of a website threatening to publicly shame you for non-compliance.
Wasn’t the average IQ here from the survey something like 130+?
The average self-reported IQ.
If we really wanted to measure LWs collective IQ, I’d suggest using the education data as a proxy; we have fairly good information about average IQs by degree and major, and people with less educational history will likely be much less reticent to answer than those with a low IQ test result since there are so many celebrated geniuses who didn’t complete their schooling.
The average tested IQ on the survey was about 125, which is close to my estimate of the true average IQ around here; I don’t entirely trust the testing site that Yvain used, but I think it’s skewing low, and that ought to counteract some of the reporting bias that I’d still expect to see.
125 is pretty much in line with what you’d expect if you assume that everyone here is, or is going to be, a four-year college graduate in math, philosophy, or a math-heavy science or engineering field (source). That’s untrue as stated, of course, but we do skew that way pretty hard, and I’m prepared to assume that the average contributor has that kind of intellectual chops.
I think that’s a fair assessment, although it might be because my guess was around 120 to start with. I never meant to say we’re not smart around here, far from it, but I don’t think we’re all borderline geniuses either. It’s important to keep perspective and very easy to overestimate yourself.
Find those reasons to be sufficient to spend time reading it
Read it
Put forth the cognitive effort to understand it (reading something and putting forth cognitive resources to understand it are not the same thing)
Succeed in understanding it
Intelligence is just one component of knowledge acquisition, and probably less important than affective issue. Often, intelligence acts indirectly by affecting affect, but in such cases, those effects can be counteracted. The mistaking of performance of cognitive tasks for intelligence is, I believe, often an aspect of the fundamental attribution error.
Not anymore, though only just. The 2012 survey reports a mean of 138 and change with a SD of 12.7. It was 140 or higher on the 2011 and 2009 surveys, though.
All the usual self-reporting caveats apply, of course.
Wasn’t the average IQ here from the survey something like 130+?
These statements don’t necessarily contradict each other. Even if average-intelligence people don’t read Less Wrong, perhaps they could. Personally, I suspect it’s more because of a lack of interest (and perhaps a constellation of social factors).
I bet the average LessWrong person has a great sense of humour and feels things more than other people, too.
Seriously, every informal IQ survey amongst a group/forum I have seen reports very high IQ. My (vague) memories of the LessWrong one included people who seemed to be off the scale (I don’t mean very bright. I mean that such IQs either have never been given out in official testing rather than online tests, or possibly that they just can’t be got on those tests and people were lying).
There’s always a massive bias in self-reporting: those will only be emphasised on an intellectual website that starts the survey post by saying that LessWrongers are, on average, in the top 0.11% for SATs, and gives pre-packaged excuses for not reporting inconvenient results - “Many people would prefer not to have people knowing their scores. That’s great, but please please please do post it anonymously. Especially if it’s a low one, but not if it’s low because you rushed the test”, (my emphasis).
If there’s a reason to be interested in average IQ beyond mutual ego-massage, I guess the best way would be to have an IQ test where you logged on as ‘Less Wrong member X’ and then it reported all the results, not just the ones that people chose to share. And where it revealed how many people pulled out halfway through (to avoid people bailing if they weren’t doing well).
Selection bias—which groups and forums actually asked about IQ?
Your average knitting/auto maintenance/comic book forum probably has a lower average IQ but doesn’t think to ask. And of course we’re already selecting a little just by taking the figures off of web forums, which are a little on the cerebral side.
True. I don’t think I can define the precise level of inaccuracy or anything. My point is not that I’ve detected the true signal: it’s that there’s too much noise for there to be a useful signal.
Do I think the average LessWronger has a higher IQ? Sure. But that’s nothing remotely to do with this survey. It’s just too flawed to give me any particularly useful information. I would probably update my view of LW intelligence more based on its existence than its results. In that reading the thread lowers my opinion of LW intellgence, simply because this forum is usually massively more rational and self-questioning than every other forum I’ve been on, which I would guess is associated with high IQ, and people taking the survey seriously is one of the clearest exceptions.
BTW, I’m not sure your assessments of knitting/auto maintenance/comic books/web forums are necessarily accurate. I’m not sure I have enough information on any of them to reasonably guess their intelligence. Forums are particularly exceptional in terms of showing amazing intelligence and incredible stupidity side by side.
People with high IQ have extra power to be exceptionally stupid.
Would still suffer from selection effects. People that thought they might not do so well would be disinclined to do it, and people who knew they were hot shit would be extra inclined to do it. The phrase “anonymous survey” doesn’t really penetrate into our status-aware hindbrains.
Yep! But it’s the best way I can imagine that someone could plausibly create on the forum.
Better: randomly select a group of users (within some minimal activity criteria) and offer the test directly to that group. Publicly state the names of those selected (make it a short list, so that people actually read it, maybe 10-20) and then after a certain amount of time give another public list of those who did or didn’t take it, along with the results (although don’t associate results with names). That will get you better participation, and the fact that you have taken a group of known size makes it much easier to give outer bounds on the size of the selection effect caused by people not participating.
You can also improve participation by giving those users an easily accessible icon on Less Wrong itself which takes them directly to the test, and maybe a popup reminder once a day or so when they log on to the site if they’ve been selected but haven’t done it yet. Requires moderate coding.
I would find such a feature to be extraordinarily obnoxious, to the point that I’d be inclined to refused such a test purely out of anger (and my scores are not at all embarrassing). I can’t think of any other examples of a website threatening to publicly shame you for non-compliance.
btw, in Markdown use double asterisks at each end for bold, like this **bold text.
with two at the end also.
The average self-reported IQ.
If we really wanted to measure LWs collective IQ, I’d suggest using the education data as a proxy; we have fairly good information about average IQs by degree and major, and people with less educational history will likely be much less reticent to answer than those with a low IQ test result since there are so many celebrated geniuses who didn’t complete their schooling.
The average tested IQ on the survey was about 125, which is close to my estimate of the true average IQ around here; I don’t entirely trust the testing site that Yvain used, but I think it’s skewing low, and that ought to counteract some of the reporting bias that I’d still expect to see.
125 is pretty much in line with what you’d expect if you assume that everyone here is, or is going to be, a four-year college graduate in math, philosophy, or a math-heavy science or engineering field (source). That’s untrue as stated, of course, but we do skew that way pretty hard, and I’m prepared to assume that the average contributor has that kind of intellectual chops.
I think that’s a fair assessment, although it might be because my guess was around 120 to start with. I never meant to say we’re not smart around here, far from it, but I don’t think we’re all borderline geniuses either. It’s important to keep perspective and very easy to overestimate yourself.
To comprehend a text, a person must:
Become aware of it
Have some reason for reading it
Find those reasons to be sufficient to spend time reading it
Read it
Put forth the cognitive effort to understand it (reading something and putting forth cognitive resources to understand it are not the same thing)
Succeed in understanding it
Intelligence is just one component of knowledge acquisition, and probably less important than affective issue. Often, intelligence acts indirectly by affecting affect, but in such cases, those effects can be counteracted. The mistaking of performance of cognitive tasks for intelligence is, I believe, often an aspect of the fundamental attribution error.
140+
Not anymore, though only just. The 2012 survey reports a mean of 138 and change with a SD of 12.7. It was 140 or higher on the 2011 and 2009 surveys, though.
All the usual self-reporting caveats apply, of course.