Funny how the love question breaks the bell curve and just looks binary, presumably because you're only very upset if you're getting nothing and happy if you get anything at all, with minor variations beyond that.
I would like to hear a bit from the people who have had over 1000 lifetime sex partners. What is your life like? How many partners did you have to get through until you reached the rank of grand master of sex? (At least I would hope that someone with over 1k partners would be a gm of sex, gotta make all that grinding count for something!)
The reading comprehension clearly took a nosedive at the internet hours per day other than for work question, given the amount of answers that are >24.
Could you change the data download links? My browser complains that the current page is on https, but the download links point to an http page. I'm sure other people could have the same issue.
You only need to change to `https`, that download link appears to already work .
There’s a definite discrepancy between reported political spectrum - with the majority left or centre - and schools choice which I would rate as fairly right wing. The majority are in favour (4 or 5) and there’s a large centrist group as well. More than 70% favourable.
- 13% of readers describe themselves as working in AI
- 3.3% made 100K+ on crypto
- About 10% of readers are annoyed by the flashing "draft saved" on Gmail. Vindicated!
- As in some previous studies, if you actually survey people about eugenics, present a fair plan, and stress that you're not a Nazi, most people are in favor or at least accepting.
- 50x more people interested in donating a kidney than actually started
- Most people basically happy with social media and AI. I should ask Turkers this.
- Most people plan to regulate their kids' Internet use
- Long COVID rates slightly up since 2022, from 3.1 to 3.6%
- Face masking rate down to 4%
- Even though I tried to frame the YIMBY question fairly, people were really pro-YIMBY
I have a theory that bitcoin maximalists are also AI maximalists. That is, if you believe bitcoin will keep on going up for a long time, you believe that the total wealth of the world is going to increase, which probably means you believe that we'll enter an age of super-prosperity from AI. (Or maybe causality runs in the other direction: if you believe we'll have super-prosperity, then bitcoin might be a good way of capturing some of it.)
I think this data maybe supports that? People who work in AI earn more than those who don't (the bottom end of the distribution is curtailed... you don't see as many low-income AI workers), but income isn't predictive of cryptocurrency gains. I can't think of any other obvious confounders, and Mann-Whitney u-tests do show a difference in cryptocurrency gains between the "working in AI" respondents and "not working in AI".
1 in 6 ACX readers is formally diagnosed with depression, rising to 30% if you include self diagnosis. Average daily mood is 6.1. That seems noteworthy, does anyone have the previous survey data handy for comparison?
And I hope that everybody reading this comment (especially that 30%) has a great day.
Committed theists have a lot more children than everyone else, and are much more likely to have children. The effect is mainly Protestant/Catholic/Orthodox/Jewish. Hinduism and Buddhism seems to predict fewer children even than atheism. This might just be that people in their 40s are more likely to report being theists, but there aren't that many of them compared to atheists at any age group in this data set.
On that topic... The larger number of lukewarm theists in their 50s, 60s and 70s might be 1980s-revival related. But there's a spike around age 40 of committed theists, which seems to be a little bit young to be the echoes of the 1980s revivals. Outside of China, Korea, Africa and the Middle East, not much religious revival happened in the teenage years of today's 40-year-olds. So why are they showing up disproportionately often in ACX? (e.g. Expected 80, observed 112).
There was one religion respondent who questioned whether Mormonism, or the LDS movement generally, should be classified as Christian. I don't have a horse in this race and am presently unaffiliated with both the LDS Church and with Christianity, but I've always thought that despite their obvious heterodoxy compared to the rest of Christianity, the tenets of Mormonism should be classed as essentially Christian (they believe in G-d and in Jesus, though their cosmology is radically different from traditional trinitarians and anti-trinitarians). For that matter, I'm also of the opinion that Hong Xiuquan's Heavenly Kingdom should be properly classed as Christian.
I stopped scrolling not long after demographics, distracted by the fact that apparently the ACX community should be a huge untapped dating market for me, a single straight 39yo woman living in New York. The ratio is fantastic for me! (Yes ok my very observant Catholicism narrows the pool a bit, but still.)
Ever considered running a lonely hearts column, Scott?
The "1-5" questions would benefit from labelling of some sort.
(Number of times the Chrome tab crashed while writing this reply: 0, because I deliberately did not scroll through the comments to see if anyone else brought this up, and deliberately spent very little time making the comment)
I'm surprised how high all the proportions of "yes" answers to the "Would you sign up to/approve a program that restricted your/everyone's access to unhealthy food/recreational internet usage/social media/porn?" were. 32% of respondents found self-control with food choices enough of a bother that they'd prefer to not have a choice? 29% would forbid pornography?
I noticed this last time too: *why* does California appear multiple times in the state list (and other states too) with the same spelling, and if you select them there isn't an extra space in one of them either? Completely incomprehensible.
And moreover, what the hell is the point of the state question when it doesn't appear in the public data, and the google forms data doesn't show the percentages? All you get is a list of states, with regular multiples of the same states, and no idea what the distribution of population is.
On the "How much money have you made or lost in cryptocurrency" question, 1154 respondents report making money, 1063 "neither made nor lost very much money", and 176 report losing money.
Cryptocurrency speculating is basically zero-sum: if you made $100, that means some other people lost $100.
So either our respondents are unusually savvy speculators... or there's a very strong bias towards reporting more positive outcomes for yourself than reality indicates, to the point of making the whole survey question useless. (Well, demonstrating the strength of such a bias is useful in its own way, but you know what I mean.)
Maybe not that noteworthy but the representation by US state surprised me - primarily that Massachusetts (16th by population, fourth by responses) and Washington (13th, third) rank higher than Pennsylvania (5th, seventh) and Illinois (6th, sixth). I guess Seattle is a big tech city and MA has Boston and Cambridge? I don't know much about Pennsylvania but I didn't expect that those factors alone outweigh Chicago.
57.6% answered “yes” to the question of whether they had thought about the Roman Empire in the past 24 hours. Either the Roman Empire is more popular than I would have guessed, or some respondents interpreted the question more literally than I did. (I had thought about the Roman Empire in the past 24 hours, but only because I had just read the question. I figured that didn’t count so I answered “no.”)
I downloaded the data in order to cross-correlate the column "Do you belong to any of these ethnic groups?" with some other columns, but that column isn't in the results.
On the overgrown / dead plants question .. i dont think dead plants are nearly as bad as the other things, such as actual broken windows.
On the other hand, I have volunteered myself to tidy up the small garden of our local church.
Liker, "Come on guys, its making the neighbourhood look untidy. I'll tidy it for you if you like." "Oh, thats great thanks very much". And so here i am, volunteered for that task/
It’s a pity that the full question is not always presented on the charts. So it’s just Speech and some bar charts from 1-5.
(Presumably in that case given the audience and 1 was the plurality, 1 was full free speech and 5 restricted but it’s not clear).
Funny how the love question breaks the bell curve and just looks binary, presumably because you're only very upset if you're getting nothing and happy if you get anything at all, with minor variations beyond that.
I would like to hear a bit from the people who have had over 1000 lifetime sex partners. What is your life like? How many partners did you have to get through until you reached the rank of grand master of sex? (At least I would hope that someone with over 1k partners would be a gm of sex, gotta make all that grinding count for something!)
The reading comprehension clearly took a nosedive at the internet hours per day other than for work question, given the amount of answers that are >24.
Could you change the data download links? My browser complains that the current page is on https, but the download links point to an http page. I'm sure other people could have the same issue.
You only need to change to `https`, that download link appears to already work .
original:
http://slatestarcodex.com/Stuff/2024_public.xlsx
http://slatestarcodex.com/Stuff/2024_public.csv
suggested:
https://slatestarcodex.com/Stuff/2024_public.xlsx
https://slatestarcodex.com/Stuff/2024_public.csv
(As far as actual security gain, for xlsx this could *potentially* allow a MITM attacker to add a malicious macro, csv is likely fine regardless)
There’s a definite discrepancy between reported political spectrum - with the majority left or centre - and schools choice which I would rate as fairly right wing. The majority are in favour (4 or 5) and there’s a large centrist group as well. More than 70% favourable.
A magisterial survey of Elite Human Capital 💯 organized by one of its prime paladins.
How is 61% the answer for using Substack on a blog only available on Substack? 🤷
Was the (logarithmic?) scale on # of children intentional?
Results I find interesting at this level:
- 13% of readers describe themselves as working in AI
- 3.3% made 100K+ on crypto
- About 10% of readers are annoyed by the flashing "draft saved" on Gmail. Vindicated!
- As in some previous studies, if you actually survey people about eugenics, present a fair plan, and stress that you're not a Nazi, most people are in favor or at least accepting.
- 50x more people interested in donating a kidney than actually started
- Most people basically happy with social media and AI. I should ask Turkers this.
- Most people plan to regulate their kids' Internet use
- Long COVID rates slightly up since 2022, from 3.1 to 3.6%
- Face masking rate down to 4%
- Even though I tried to frame the YIMBY question fairly, people were really pro-YIMBY
Each extra 10 points of BMI drops you down one attractiveness point. Negative BMI almost guarantees you a 10.
I have a theory that bitcoin maximalists are also AI maximalists. That is, if you believe bitcoin will keep on going up for a long time, you believe that the total wealth of the world is going to increase, which probably means you believe that we'll enter an age of super-prosperity from AI. (Or maybe causality runs in the other direction: if you believe we'll have super-prosperity, then bitcoin might be a good way of capturing some of it.)
I think this data maybe supports that? People who work in AI earn more than those who don't (the bottom end of the distribution is curtailed... you don't see as many low-income AI workers), but income isn't predictive of cryptocurrency gains. I can't think of any other obvious confounders, and Mann-Whitney u-tests do show a difference in cryptocurrency gains between the "working in AI" respondents and "not working in AI".
1 in 6 ACX readers is formally diagnosed with depression, rising to 30% if you include self diagnosis. Average daily mood is 6.1. That seems noteworthy, does anyone have the previous survey data handy for comparison?
And I hope that everybody reading this comment (especially that 30%) has a great day.
Committed theists have a lot more children than everyone else, and are much more likely to have children. The effect is mainly Protestant/Catholic/Orthodox/Jewish. Hinduism and Buddhism seems to predict fewer children even than atheism. This might just be that people in their 40s are more likely to report being theists, but there aren't that many of them compared to atheists at any age group in this data set.
On that topic... The larger number of lukewarm theists in their 50s, 60s and 70s might be 1980s-revival related. But there's a spike around age 40 of committed theists, which seems to be a little bit young to be the echoes of the 1980s revivals. Outside of China, Korea, Africa and the Middle East, not much religious revival happened in the teenage years of today's 40-year-olds. So why are they showing up disproportionately often in ACX? (e.g. Expected 80, observed 112).
Profession -> Physics -> 137.
nice
There was one religion respondent who questioned whether Mormonism, or the LDS movement generally, should be classified as Christian. I don't have a horse in this race and am presently unaffiliated with both the LDS Church and with Christianity, but I've always thought that despite their obvious heterodoxy compared to the rest of Christianity, the tenets of Mormonism should be classed as essentially Christian (they believe in G-d and in Jesus, though their cosmology is radically different from traditional trinitarians and anti-trinitarians). For that matter, I'm also of the opinion that Hong Xiuquan's Heavenly Kingdom should be properly classed as Christian.
I stopped scrolling not long after demographics, distracted by the fact that apparently the ACX community should be a huge untapped dating market for me, a single straight 39yo woman living in New York. The ratio is fantastic for me! (Yes ok my very observant Catholicism narrows the pool a bit, but still.)
Ever considered running a lonely hearts column, Scott?
What is the data cleaning process?
Are nonsense/implausible answers removed?
If there's a clear troll/fat finger answer, is just that answer removed, or is the full survey from that respondent removed?
The "1-5" questions would benefit from labelling of some sort.
(Number of times the Chrome tab crashed while writing this reply: 0, because I deliberately did not scroll through the comments to see if anyone else brought this up, and deliberately spent very little time making the comment)
Labels are missing from the Ethnic Group results.
I'm surprised how high all the proportions of "yes" answers to the "Would you sign up to/approve a program that restricted your/everyone's access to unhealthy food/recreational internet usage/social media/porn?" were. 32% of respondents found self-control with food choices enough of a bother that they'd prefer to not have a choice? 29% would forbid pornography?
I noticed this last time too: *why* does California appear multiple times in the state list (and other states too) with the same spelling, and if you select them there isn't an extra space in one of them either? Completely incomprehensible.
And moreover, what the hell is the point of the state question when it doesn't appear in the public data, and the google forms data doesn't show the percentages? All you get is a list of states, with regular multiples of the same states, and no idea what the distribution of population is.
On the "How much money have you made or lost in cryptocurrency" question, 1154 respondents report making money, 1063 "neither made nor lost very much money", and 176 report losing money.
Cryptocurrency speculating is basically zero-sum: if you made $100, that means some other people lost $100.
So either our respondents are unusually savvy speculators... or there's a very strong bias towards reporting more positive outcomes for yourself than reality indicates, to the point of making the whole survey question useless. (Well, demonstrating the strength of such a bias is useful in its own way, but you know what I mean.)
Completely off-topic, but I just found out that Daniel Dennett has died. He was 82.
I really enjoyed his book "Breaking the Spell." May he rest in peace.
Unfortunate how the number of survey-takers has been trending down since the move to ACX, when it was flat for the last several years on SSC.
Maybe not that noteworthy but the representation by US state surprised me - primarily that Massachusetts (16th by population, fourth by responses) and Washington (13th, third) rank higher than Pennsylvania (5th, seventh) and Illinois (6th, sixth). I guess Seattle is a big tech city and MA has Boston and Cambridge? I don't know much about Pennsylvania but I didn't expect that those factors alone outweigh Chicago.
57.6% answered “yes” to the question of whether they had thought about the Roman Empire in the past 24 hours. Either the Roman Empire is more popular than I would have guessed, or some respondents interpreted the question more literally than I did. (I had thought about the Roman Empire in the past 24 hours, but only because I had just read the question. I figured that didn’t count so I answered “no.”)
I downloaded the data in order to cross-correlate the column "Do you belong to any of these ethnic groups?" with some other columns, but that column isn't in the results.
On the overgrown / dead plants question .. i dont think dead plants are nearly as bad as the other things, such as actual broken windows.
On the other hand, I have volunteered myself to tidy up the small garden of our local church.
Liker, "Come on guys, its making the neighbourhood look untidy. I'll tidy it for you if you like." "Oh, thats great thanks very much". And so here i am, volunteered for that task/
so, I'm not the only one who's never gotten COVID? I feel less special / more belonging.
Мой ДНК не раскроется до полного потенциала в Росиии
Was there any way to bin the wrong-capitalization answers? It seems like _a lot_ of answers would benefit from that (e.g. Maine and maine).
Have you considered using Differential Privacy on top of, or related to, the changes you made to preserve individually identifiable information?