Atheism is, in almost every way, a harder choice to make than rejecting creationism, your “second place.” The oft-cited example of “smart guy who believes in god” ’round here is Robert Aumann. Another thing to consider is that only looking for people who already agree with the site is bad—we want people who don’t agree or don’t know, but who are willing to listen. So although highly correlated with being on LW, reporting as atheist/agnostic is not a good category, and you should probably find something else.
The NT distinction is reasonable when looking at the current audience of LessWrong, but probably too strict when talking about the potential audience, for the same reason as above. We don’t want to “permanently disqualify” anyone from being rational for 1 category—we want to peach the gospel, in a sense.
130 IQ? Really? Predictably Irrational was the #23 selling book on Amazon in 2008. It is astronomically unlikely that every single person who read the book had IQ over 130. Some filter for intelligence is needed, sure. But not 1.5 standard deviations, single tailed! Or, more anecdotally: I know plenty of average (less than 1 s.d. above mean) people who like reading interesting things.
Statistical things to fix
You did a generally good job controlling for correlations, with a few gaps on (30 minutes on computer)|(all the other stuff)
The real problem with “no free time”: you assume everyone’s average rather than accounting for variation, which would increase the numbers there.
Yvain’s survey. I think you’re underestimating the awesome power of response bias. And also neglecting the extra factor that all those people signed up for LW, not just spent some time reading.
I’ll try and be a bit more constructive when I have more time.
“130 IQ? Really? Predictably Irrational was the #23 selling book on Amazon in 2008. It is astronomically unlikely that every single person who read the book had IQ over 130. Some filter for intelligence is needed, sure. But not 1.5 standard deviations, single tailed! Or, more anecdotally: I know plenty of average (less than 1 s.d. above mean) people who like reading interesting things.”
I bet you don’t. Liking to read at all immediately makes someone non-average. The average USian reads four books per year ( source http://www.washingtonpost.com/wp-dyn/content/article/2007/08/21/AR2007082101045.html ), with 25% not reading at all. The same source says “The Bible and religious works were read by two-thirds in the survey, more than all other categories. Popular fiction, histories, biographies and mysteries were all cited by about half, while one in five read romance novels. Every other genre including politics, poetry and classical literature were named by fewer than five percent of readers.”
I know that the mean adult in the UK buys 1.8 books per year (they read another three or four from the library or friends).
Assuming any kind of correlation between amount of books read and intelligence (which I think is a fair assumption) and taking ‘like to read’ as meaning reading more than one book a month, say, then assuming IQ can be a reasonable proxy for intelligence (an assumption I don’t share, but leave that for present) then while I can’t find the raw data for that poll, I bet that twelve books per year is a couple of standard deviations from the mean, and probably correlates with an intelligence a couple of standard deviations above the mean (especially when you talk about ‘interesting’ books, which on this forum I take to be more likely to mean, say, Godel, Escher, Bach than The DaVinci Code)
There’s probably a selection bias present there—you probably think of people as being closer to the mean than they really are, because in general people tend to associate with people with similar tastes whenever possible.
Yeah, there’s probably some selection bias, but not so much that I can’t compensate for it. I think you misinterpreted me, or I wasn’t clear enough—nobody is average in every way, so I would have to be really stupid to mean “average everything including book-reading habits” when I said average. So what did I mean when I said “average?” Well, what seemed like the obvious interpretation to me was that I meant “average-ish IQ.” As in “I know people within 1 standard deviation of average IQ who like reading interesting things, and it’s a bad plan to respond to someone else’s personal experience with ‘you must be lying or stupid.’” A better choice would have been “then you’re very unusual” or “but I don’t, and here are some numbers.”
I bet that twelve books per year is a couple of standard deviations from the mean, and probably correlates with an intelligence a couple of standard deviations above the mean
Standard deviations may not be the best way to think about this distribution. They’ll lead you to picture it as exponential decay even though it isn’t. Plus that 1.8 books/yr number may be stuck in your head, even though that was sales (I bought 0 books this year). The median for reading was 7 books/yr in your linked survey. So if 25% read 0 books, that means the spread is big, putting 12 only around the 3rd quartile. I.e. less than 1 standard deviation above the mean of a bell-curve.
‘interesting’ books, which on this forum I take to be more likely to mean, say, Godel, Escher, Bach than The DaVinci Code
Nah, they wouldn’t be able to handle GEB. My prototype was Predictably Irrational. I’d guess about 1⁄5 of the sequence posts are that level (if “fluent” usage of the internet is also assumed), and with some editing (perhaps by some sort of directed community effort) that number could be raised to 1⁄3 (and not assuming internet fluency), containing most of the important stuff. Which is still several books worth, iirc.
I don’t really see how being smart grants you extra hours / day. Your commute, your job, your wife, your children, and your body (which wants sleep, food, sex, emotional connection, etc) all don’t care how smart you are—they want the exact same amount of time from you in all cases regardless of you being being IQ 80 or IQ 180.
Smart people maybe get a couple hours / week back by avoiding church. What other savings do rationalists realistically get in a uniform way?
ALSO—although I labeled it Atheist/Agnostic, I’m counting anyone who is unsure about the existence of god as being those two, not just the 1% of people who self-ID as atheist.
I agree that being smart doesn’t magically give you extra time (although maybe watching less TV goes here).
What I said was that you assumed everyone was the average for their group and didn’t account for variation: you discount everyone who’s employed/has kids, rather than instead looking at what percentage of those groups spend >30 minutes/day on the computer.
ALSO - ( :P ) I still think that’s too strict. You don’t even accept the full number of people who declared “none” for religion, apparently (15% in 2008 - see your link pg 5). You claim that if you believe in god then “lessons in rationality are a complete waste of time for you,” when I’d think that it would be the opposite: it is when someone is irrational that they can use lessons in rationality. The question is then not “who already agrees?” but “who is willing to listen?”
Mine does. I don’t see this thing you’re claiming about one’s spouse/partner/whatever being a timesink. Surely most people (at least most rational people) would marry someone who shares their interests, rather than detracts from them? There are plenty of things that have removed big chunks of my time (work foremost among them) but I don’t have any less free time since my marriage...
I’ll try to think of ways to integrate this with what I’ve been reading in the US time use surveys. I know you’re not American, but being married in the US typically halves peoples’ leisure time spent reading and using computers and the displaced time ends up being spent as extra time at work. This pattern matches really well to what I’ve seen from my friends and colleges at old jobs who went from being single to married… even smart ones.
So I wasn’t making a wild, sexist conjecture without context. I was just pointing to the mountains of data supporting this point and assuming that people wouldn’t find it controversial.
“the displaced time ends up being spent as extra time at work”
But you’re already counting “being in full-time work” as one of your factors, so you’re counting that twice. If someone’s in full-time work, you say it’s impossible for them to read LessWrong, and then on top of that you cut out all those who are married because they’ll be working more!
You’re also assuming that someone who will be interested in reading the site will follow normal patterns of behaviour and time use, when the very fact that they’re interested in reading the site would suggest otherwise...
I’m pretty sure I’m counting each separately and disqualifying people for having either one or both. I don’t see how that’s counting twice. Usually the extra time lost during marriage goes to work because lots of people have both a spouse and a job. But even without a job, a spouse typically diminishes free time spent on solitary leisure activities by a factor of 2 which I keep mentioning. I’m glad your personal situation is better than this.
I agree that people currently on LW are somehow re-prioritizing their time in novel ways which allow them to read the site. I’m just pointing out that there are likely 10-20x as many people out there for every current LW reader who don’t have the skills to effectively do this and this barrier keeps them from possibly reading the sequences or using LW.
Maybe trying to teach people better time management skills and prioritization would be helpful? This would allow them to have the free time to possibly read LW or do whatever it was they found exciting or helpful in life.
Two categories of comments:
Argument about which categories to choose
Atheism is, in almost every way, a harder choice to make than rejecting creationism, your “second place.” The oft-cited example of “smart guy who believes in god” ’round here is Robert Aumann. Another thing to consider is that only looking for people who already agree with the site is bad—we want people who don’t agree or don’t know, but who are willing to listen. So although highly correlated with being on LW, reporting as atheist/agnostic is not a good category, and you should probably find something else.
The NT distinction is reasonable when looking at the current audience of LessWrong, but probably too strict when talking about the potential audience, for the same reason as above. We don’t want to “permanently disqualify” anyone from being rational for 1 category—we want to peach the gospel, in a sense.
130 IQ? Really? Predictably Irrational was the #23 selling book on Amazon in 2008. It is astronomically unlikely that every single person who read the book had IQ over 130. Some filter for intelligence is needed, sure. But not 1.5 standard deviations, single tailed! Or, more anecdotally: I know plenty of average (less than 1 s.d. above mean) people who like reading interesting things.
Statistical things to fix
You did a generally good job controlling for correlations, with a few gaps on (30 minutes on computer)|(all the other stuff)
The real problem with “no free time”: you assume everyone’s average rather than accounting for variation, which would increase the numbers there.
Yvain’s survey. I think you’re underestimating the awesome power of response bias. And also neglecting the extra factor that all those people signed up for LW, not just spent some time reading.
I’ll try and be a bit more constructive when I have more time.
“130 IQ? Really? Predictably Irrational was the #23 selling book on Amazon in 2008. It is astronomically unlikely that every single person who read the book had IQ over 130. Some filter for intelligence is needed, sure. But not 1.5 standard deviations, single tailed! Or, more anecdotally: I know plenty of average (less than 1 s.d. above mean) people who like reading interesting things.”
I bet you don’t. Liking to read at all immediately makes someone non-average. The average USian reads four books per year ( source http://www.washingtonpost.com/wp-dyn/content/article/2007/08/21/AR2007082101045.html ), with 25% not reading at all. The same source says “The Bible and religious works were read by two-thirds in the survey, more than all other categories. Popular fiction, histories, biographies and mysteries were all cited by about half, while one in five read romance novels. Every other genre including politics, poetry and classical literature were named by fewer than five percent of readers.”
I know that the mean adult in the UK buys 1.8 books per year (they read another three or four from the library or friends).
Assuming any kind of correlation between amount of books read and intelligence (which I think is a fair assumption) and taking ‘like to read’ as meaning reading more than one book a month, say, then assuming IQ can be a reasonable proxy for intelligence (an assumption I don’t share, but leave that for present) then while I can’t find the raw data for that poll, I bet that twelve books per year is a couple of standard deviations from the mean, and probably correlates with an intelligence a couple of standard deviations above the mean (especially when you talk about ‘interesting’ books, which on this forum I take to be more likely to mean, say, Godel, Escher, Bach than The DaVinci Code)
There’s probably a selection bias present there—you probably think of people as being closer to the mean than they really are, because in general people tend to associate with people with similar tastes whenever possible.
Yeah, there’s probably some selection bias, but not so much that I can’t compensate for it. I think you misinterpreted me, or I wasn’t clear enough—nobody is average in every way, so I would have to be really stupid to mean “average everything including book-reading habits” when I said average. So what did I mean when I said “average?” Well, what seemed like the obvious interpretation to me was that I meant “average-ish IQ.” As in “I know people within 1 standard deviation of average IQ who like reading interesting things, and it’s a bad plan to respond to someone else’s personal experience with ‘you must be lying or stupid.’” A better choice would have been “then you’re very unusual” or “but I don’t, and here are some numbers.”
Standard deviations may not be the best way to think about this distribution. They’ll lead you to picture it as exponential decay even though it isn’t. Plus that 1.8 books/yr number may be stuck in your head, even though that was sales (I bought 0 books this year). The median for reading was 7 books/yr in your linked survey. So if 25% read 0 books, that means the spread is big, putting 12 only around the 3rd quartile. I.e. less than 1 standard deviation above the mean of a bell-curve.
Nah, they wouldn’t be able to handle GEB. My prototype was Predictably Irrational. I’d guess about 1⁄5 of the sequence posts are that level (if “fluent” usage of the internet is also assumed), and with some editing (perhaps by some sort of directed community effort) that number could be raised to 1⁄3 (and not assuming internet fluency), containing most of the important stuff. Which is still several books worth, iirc.
I don’t really see how being smart grants you extra hours / day. Your commute, your job, your wife, your children, and your body (which wants sleep, food, sex, emotional connection, etc) all don’t care how smart you are—they want the exact same amount of time from you in all cases regardless of you being being IQ 80 or IQ 180.
Smart people maybe get a couple hours / week back by avoiding church. What other savings do rationalists realistically get in a uniform way?
ALSO—although I labeled it Atheist/Agnostic, I’m counting anyone who is unsure about the existence of god as being those two, not just the 1% of people who self-ID as atheist.
I agree that being smart doesn’t magically give you extra time (although maybe watching less TV goes here).
What I said was that you assumed everyone was the average for their group and didn’t account for variation: you discount everyone who’s employed/has kids, rather than instead looking at what percentage of those groups spend >30 minutes/day on the computer.
ALSO - ( :P ) I still think that’s too strict. You don’t even accept the full number of people who declared “none” for religion, apparently (15% in 2008 - see your link pg 5). You claim that if you believe in god then “lessons in rationality are a complete waste of time for you,” when I’d think that it would be the opposite: it is when someone is irrational that they can use lessons in rationality. The question is then not “who already agrees?” but “who is willing to listen?”
“your wife… don’t care how smart you are”
Mine does. I don’t see this thing you’re claiming about one’s spouse/partner/whatever being a timesink. Surely most people (at least most rational people) would marry someone who shares their interests, rather than detracts from them? There are plenty of things that have removed big chunks of my time (work foremost among them) but I don’t have any less free time since my marriage...
I’ll try to think of ways to integrate this with what I’ve been reading in the US time use surveys. I know you’re not American, but being married in the US typically halves peoples’ leisure time spent reading and using computers and the displaced time ends up being spent as extra time at work. This pattern matches really well to what I’ve seen from my friends and colleges at old jobs who went from being single to married… even smart ones.
So I wasn’t making a wild, sexist conjecture without context. I was just pointing to the mountains of data supporting this point and assuming that people wouldn’t find it controversial.
“the displaced time ends up being spent as extra time at work” But you’re already counting “being in full-time work” as one of your factors, so you’re counting that twice. If someone’s in full-time work, you say it’s impossible for them to read LessWrong, and then on top of that you cut out all those who are married because they’ll be working more! You’re also assuming that someone who will be interested in reading the site will follow normal patterns of behaviour and time use, when the very fact that they’re interested in reading the site would suggest otherwise...
I’m pretty sure I’m counting each separately and disqualifying people for having either one or both. I don’t see how that’s counting twice. Usually the extra time lost during marriage goes to work because lots of people have both a spouse and a job. But even without a job, a spouse typically diminishes free time spent on solitary leisure activities by a factor of 2 which I keep mentioning. I’m glad your personal situation is better than this.
I agree that people currently on LW are somehow re-prioritizing their time in novel ways which allow them to read the site. I’m just pointing out that there are likely 10-20x as many people out there for every current LW reader who don’t have the skills to effectively do this and this barrier keeps them from possibly reading the sequences or using LW.
Maybe trying to teach people better time management skills and prioritization would be helpful? This would allow them to have the free time to possibly read LW or do whatever it was they found exciting or helpful in life.