2012 Survey Results

Thank you to ev­ery­one who took the 2012 Less Wrong Sur­vey (the sur­vey is now closed. Do not try to take it.) Below the cut, this post con­tains the ba­sic sur­vey re­sults, a few more com­pli­cated analy­ses, and the data available for down­load so you can ex­plore it fur­ther on your own. You may want to com­pare these to the re­sults of the 2011 Less Wrong Sur­vey.

Part 1: Population

How many of us are there?

The short an­swer is that I don’t know.

The 2011 sur­vey ran 33 days and col­lected 1090 re­sponses. This year’s sur­vey ran 23 days and col­lected 1195 re­sponses. The av­er­age num­ber of new re­sponses dur­ing the last week was about five per day, so even if I had kept this sur­vey open as long as the last one I prob­a­bly wouldn’t have got­ten more than about 1250 re­sponses. That means at most a 15% year on year growth rate, which is pretty abysmal com­pared to the 650% growth rate in two years we saw last time.

About half of these re­sponses were from lurk­ers; over half of the non-lurker re­main­der had com­mented but never posted to Main or Dis­cus­sion. That means there were only about 600 non-lurk­ers.

But I am skep­ti­cal of these num­bers. I hang out with some peo­ple who are very closely as­so­ci­ated with the greater Less Wrong com­mu­nity, and a lot of them didn’t know about the sur­vey un­til I men­tioned it to them in per­son. I know some peo­ple who could plau­si­bly be de­scribed as fo­cus­ing their lives around the com­mu­nity who just never took the sur­vey for one rea­son or an­other. One les­son of this sur­vey may be that the com­mu­nity is no longer limited to peo­ple who check Less Wrong very of­ten, if at all. One friend didn’t see the sur­vey be­cause she hangs out on the #less­wrong chan­nel more than the main site. Another mostly just goes to mee­tups. So I think this rep­re­sents only a small sam­ple of peo­ple who could justly be con­sid­ered Less Wrongers.

The ques­tion of “how quickly is LW grow­ing” is also com­pli­cated by the high turnover. Over half the peo­ple who took this sur­vey said they hadn’t par­ti­ci­pated in the sur­vey last year. I tried to break this down by com­bin­ing a few sources of in­for­ma­tion, and I think our 1200 re­spon­dents in­clude 500 peo­ple who took last year’s sur­vey, 400 peo­ple who were around last year but didn’t take the sur­vey for some rea­son, and 300 new peo­ple.

As ex­pected, there’s lower turnover among reg­u­lars than among lurk­ers. Of peo­ple who have posted in Main, about 75% took the sur­vey last year; of peo­ple who only lurked, about 75% hadn’t.

This view of a very high-turnover com­mu­nity and lots of peo­ple not tak­ing the sur­vey is con­sis­tent with Vladimir Nesov’s data show­ing http://​​less­wrong.com/​​lw/​​e4j/​​num­ber_of_mem­bers_on_less­wrong/​​77xz 1390 peo­ple who have writ­ten at least ten com­ments. But the sur­vey in­cludes only about 600 peo­ple who have at least com­mented; 800ish of Vladimir’s ac­counts are ei­ther gone or didn’t take the cen­sus.

Part 2: Cat­e­gor­i­cal Data

SEX:
Man: 1057, 89.2%
Wo­man: 120, 10.1%
Other: 2, 0.2%)
No an­swer: 6, 0.5%

GENDER:
M (cis): 1021, 86.2%
F (cis): 105, 8.9%
M (trans f->m): 3, 0.3%
F (trans m->f): 16, 1.3%
Other: 29, 2.4%
No an­swer: 11, 0.9%

ORIENTATION:
Hetero­sex­ual: 964, 80.7%
Bi­sex­ual: 135, 11.4%
Ho­mo­sex­ual: 28, 2.4%
Asex­ual: 24, 2%
Other: 28, 2.4%
No an­swer: 14, 1.2%

RELATIONSHIP STYLE:

Pre­fer monog­a­mous: 639, 53.9%
Pre­fer polyamorous: 155, 13.1%
Uncer­tain/​no prefer­ence: 358, 30.2%
Other: 21, 1.8%
No an­swer: 12, 1%

NUMBER OF CURRENT PARTNERS:
0: 591, 49.8%
1: 519, 43.8%
2: 34, 2.9%
3: 12, 1%
4: 5, 0.4%
6: 1, 0.1%
7, 1, 0.1% (and this per­son added “re­ally, not trol­ling”)
Con­fus­ing or no an­swer: 20, 1.8%

RELATIONSHIP STATUS:
Sin­gle: 628, 53%
Re­la­tion­ship: 323, 27.3%
Mar­ried: 220, 18.6%
No an­swer: 14, 1.2%

RELATIONSHIP GOALS:
Not look­ing for more part­ners: 707, 59.7%
Look­ing for more part­ners: 458, 38.6%
No an­swer: 20, 1.7%

COUNTRY:
USA: 651, 54.9%
UK: 103, 8.7%
Canada: 74, 6.2%
Aus­tralia: 59, 5%
Ger­many: 54, 4.6%
Is­rael: 15, 1.3%
Fin­land: 15, 1.3%
Rus­sia: 13, 1.1%
Poland: 12, 1%

Th­ese are all the coun­tries with greater than 1% of Less Wrongers, but other, more ex­otic lo­cales in­cluded Kenya, Pak­istan, and Ice­land, with one user each. You can see the full table here.

This data also al­lows us to calcu­late Less Wrongers per cap­ita:


Fin­land: 1366,666
Aus­tralia: 1389,830
Canada: 1472,972
USA: 1483,870
Is­rael: 1533,333
UK: 1603,883
Ger­many: 11,518,518
Poland: 13,166,666
Rus­sia: 111,538,462

RACE:
White, non-His­panic 1003, 84.6%
East Asian: 50, 4.2%
His­panic 47, 4.0%
In­dian Sub­con­ti­nen­tal 28, 2.4%
Black 8, 0.7%
Mid­dle Eastern 4, 0.3%
Other: 33, 2.8%
No an­swer: 12, 1%

WORK STATUS:
Stu­dent: 476, 40.7%
For-profit work: 364, 30.7%
Self-em­ployed: 95, 8%
Unem­ployed: 81, 6.8%
Aca­demics (teach­ing): 54, 4.6%
Govern­ment: 46, 3.9%
Non-profit: 44, 3.7%
In­de­pen­dently wealthy: 12, 1%
No an­swer: 13, 1.1%

PROFESSION:
Com­put­ers (prac­ti­cal): 344, 29%
Math: 109, 9.2%
Eng­ineer­ing: 98, 8.3%
Com­put­ers (aca­demic): 72, 6.1%
Physics: 66, 5.6%
Fi­nance/​Econ: 65, 5.5%
Com­put­ers (AI): 39, 3.3%
Philos­o­phy: 36, 3%
Psy­chol­ogy: 25, 2.1%
Busi­ness: 23, 1.9%
Art: 22, 1.9%
Law: 21, 1.8%
Neu­ro­science: 19, 1.6%
Medicine: 15, 1.3%
Other so­cial sci­ence: 24, 2%
Other hard sci­ence: 20, 1.7%
Other: 123, 10.4%
No an­swer: 27, 2.3%

DEGREE:
Bach­e­lor’s: 438, 37%
High school: 333, 28.1%
Master’s: 192, 16.2%
Ph.D: 71, 6%
2-year: 43, 3.6%
MD/​JD/​pro­fes­sional: 24, 2%
None: 55, 4.6%
Other: 15, 1.3%
No an­swer: 14, 1.2%

POLITICS:
Liberal: 427, 36%
Liber­tar­ian: 359, 30.3%
So­cial­ist: 326, 27.5%
Con­ser­va­tive: 35, 3%
Com­mu­nist: 8, 0.7%
No an­swer: 30, 2.5%

You can see the ex­act defi­ni­tions given for each of these terms on the sur­vey.

RELIGIOUS VIEWS:
Athe­ist, not spiritual: 880, 74.3%
Athe­ist, spiritual: 107, 9.0%
Ag­nos­tic: 94, 7.9%
Com­mit­ted the­ist: 37, 3.1%
Luke­warm the­ist: 27, 2.3%
Deist/​Pan­the­ist/​etc: 23, 1.9%
No an­swer: 17, 1.4%

FAMILY RELIGIOUS VIEWS:
Luke­warm the­ist: 392, 33.1%
Com­mit­ted the­ist: 307, 25.9%
Athe­ist, not spiritual: 161, 13.6
Ag­nos­tic: 149, 12.6%
Athe­ist, spiritual: 46, 3.9%
Deist/​Pan­the­ist/​Etc: 32, 2.7%
Other: 84, 7.1%

RELIGIOUS BACKGROUND:
Other Chris­tian: 517, 43.6%
Catholic: 295, 24.9%
Jewish: 100, 8.4%
Hindu: 21, 1.8%
Tra­di­tional Chi­nese: 17, 1.4%
Mor­mon: 15, 1.3%
Mus­lim: 12, 1%

Raw data is available here.

MORAL VIEWS:

Con­se­quen­tial­ism: 735, 62%
Virtue Ethics: 166, 14%
Deon­tol­ogy: 50, 4.2%
Other: 214, 18.1%
No an­swer: 20, 1.7%

NUMBER OF CHILDREN
0: 1044, 88.1%
1: 51, 4.3%
2: 48, 4.1%
3: 19, 1.6%
4: 3, 0.3%
5: 2, 0.2%
6: 1, 0.1%
No an­swer: 17, 1.4%

WANT MORE CHILDREN?

No: 438, 37%
Maybe: 363, 30.7%
Yes: 366, 30.9%
No an­swer: 16, 1.4%

LESS WRONG USE:
Lurk­ers (no ac­count): 407, 34.4%
Lurk­ers (with ac­count): 138, 11.7%
Posters (com­ments only): 356, 30.1%
Posters (com­ments + Dis­cus­sion only): 164, 13.9%
Posters (in­clud­ing Main): 102, 8.6%

SEQUENCES:
Never knew they ex­isted un­til this mo­ment: 99, 8.4%
Knew they ex­isted; never looked at them: 23, 1.9%
Read < 25%: 227, 19.2%
Read ~ 25%: 145, 12.3%
Read ~ 50%: 164, 13.9%
Read ~ 75%: 203, 17.2%
Read ~ all: 306, 24.9%
No an­swer: 16, 1.4%

Dear 8.4% of peo­ple: there is this col­lec­tion of old blog posts called the Se­quences. It is by Eliezer, the same guy who wrote Harry Pot­ter and the Meth­ods of Ra­tion­al­ity. It is re­ally good! If you read it, you will un­der­stand what we’re talk­ing about much bet­ter!

REFERRALS:
Been here since Over­com­ing Bias: 265, 22.4%
Referred by a link on an­other blog: 23.5%
Referred by a friend: 147, 12.4%
Referred by HPMOR: 262, 22.1%
No an­swer: 35, 3%

BLOG REFERRALS:

Com­mon Sense Athe­ism: 20 peo­ple
Hacker News: 20 peo­ple
Red­dit: 15 peo­ple
Unequally Yoked: 7 peo­ple
TV Tropes: 7 peo­ple
Marginal Revolu­tion: 6 peo­ple
g­w­ern.net: 5 peo­ple
Ra­tion­alWiki: 4 peo­ple
Shtetl-Op­ti­mized: 4 peo­ple
XKCD fora: 3 peo­ple
Ac­cel­er­at­ing Fu­ture: 3 peo­ple

Th­ese are all the sites that referred at least three peo­ple in a way that was ob­vi­ous to dis­en­tan­gle from the raw data. You can see a more com­plete list, in­clud­ing the long tail, here.

MEETUPS:
Never been to one: 834, 70.5%
Have been to one: 320, 27%
No an­swer: 29, 2.5%

CATASTROPHE:
Pan­demic (bio­eng­ineered): 272, 23%
En­vi­ron­men­tal col­lapse: 171, 14.5%
Un­friendly AI: 160, 13.5%
Nu­clear war: 155, 13.1%
Eco­nomic/​Poli­ti­cal col­lapse: 137, 11.6%
Pan­demic (nat­u­ral): 99, 8.4%
Nan­otech: 49, 4.1%
As­teroid: 43, 3.6%

The word­ing of this ques­tion was “which dis­aster do you think is most likely to wipe out greater than 90% of hu­man­ity be­fore the year 2100?”

CRYONICS STATUS:
No, don’t want to: 275, 23.2%
No, still think­ing: 472, 39.9%
No, pro­cras­ti­nat­ing: 178, 15%
No, un­available: 120, 10.1%
Yes, signed up: 44, 3.7%
Never thought about it: 46, 3.9%
No an­swer: 48, 4.1%

VEGETARIAN:
No: 906, 76.6%
Yes: 147, 12.4%
No an­swer: 130, 11%

For com­par­i­son, 3.2% of US adults are veg­e­tar­ian.


SPACED REPETITION SYSTEMS
Don’t use them: 511, 43.2%
Do use them: 235, 19.9%
Never heard of them: 302, 25.5%

Dear 25.5% of peo­ple: spaced rep­e­ti­tion sys­tems are nifty, mostly free com­puter pro­grams that al­low you to study and mem­o­rize facts more effi­ciently. See for ex­am­ple http://​​ankisrs.net/​​

HPMOR:
Never read it: 219, 18.5%
Started, haven’t finished: 190, 16.1%
Read all of it so far: 659, 55.7%

Dear 18.5% of peo­ple: Harry Pot­ter and the Meth­ods of Ra­tion­al­ity is a Harry Pot­ter fan­fic about ra­tio­nal think­ing writ­ten by Eliezer Yud­kowsky (the guy who started this site). It’s re­ally good. You can find it at http://​​www.hp­mor.com/​​.


ALTERNATIVE POLITICS QUESTION:

Pro­gres­sive: 429, 36.3%
Liber­tar­ian: 278, 23.5%
Re­ac­tionary: 30, 2.5%
Con­ser­va­tive: 24, 2%
Com­mu­nist: 22, 1.9%
Other: 156, 13.2%

ALTERNATIVE ALTERNATIVE POLITICS QUESTION:
Left-Liber­tar­ian: 102, 8.6%
Pro­gres­sive: 98, 8.3%
Liber­tar­ian: 91, 7.7%
Prag­ma­tist: 85, 7.2%
So­cial Demo­crat: 80, 6.8%
So­cial­ist: 66, 5.6%
Anar­chist: 50, 4.1%
Futarchist: 29, 2.5%
Moder­ate: 18, 1.5%
Mold­bug­gian: 19, 1.6%
Ob­jec­tivist: 11, 0.9%

Th­ese are the only ones that had more than ten peo­ple. Other re­sponses no­table for their un­usu­al­ness were Monar­chist (5 peo­ple), fas­cist (3 peo­ple, plus one who was up for fas­cism but only if he could be the leader), con­ser­va­tive (9 peo­ple), and a bunch of peo­ple tel­ling me poli­tics was stupid and I should feel bad for ask­ing the ques­tion. You can see the full table here.

CAFFEINE:
Never: 162, 13.7%
Rarely: 237, 20%
At least 1x/​week: 207, 17.5
Daily: 448, 37.9
No an­swer: 129, 10.9%

SMOKING:
Never: 896, 75.7%
Used to: 1-5, 8.9%
Still do: 51, 4.3%
No an­swer: 131, 11.1%

For com­par­i­son, about 28.4% of the US adult pop­u­la­tion smokes

NICOTINE (OTHER THAN SMOKING):
Never used: 916, 77.4%
Rarely use: 82, 6.9%
>1x/​month: 32, 2.7%
Every day: 14, 1.2%
No an­swer: 139, 11.7%

MODAFINIL:
Never: 76.5%
Rarely: 78, 6.6%
>1x/​month: 48, 4.1%
Every day: 9, 0.8%
No an­swer: 143, 12.1%

TRUE PRISONERS’ DILEMMA:
Defect: 341, 28.8%
Co­op­er­ate: 316, 26.7%
Not sure: 297, 25.1%
No an­swer: 229, 19.4%

FREE WILL:
Not con­fused: 655, 55.4%
Some­what con­fused: 296, 25%
Con­fused: 81, 6.8%
No an­swer: 151, 12.8%

TORTURE VS. DUST SPECKS
Choose dust specks: 435, 36.8%
Choose tor­ture: 261, 22.1%
Not sure: 225, 19%
Don’t un­der­stand: 22, 1.9%
No an­swer: 240, 20.3%

SCHRODINGER EQUATION:
Can’t calcu­late it: 855, 72.3%
Can calcu­late it: 175, 14.8%
No an­swer: 153, 12.9%

PRIMARY LANGUAGE:
English: 797, 67.3%
Ger­man: 54, 4.5%
French: 13, 1.1%
Fin­nish: 11, 0.9%
Dutch: 10, 0.9%
Rus­sian: 15, 1.3%
Por­tuguese: 10, 0.9%

Th­ese are all the lan­guages with ten or more speak­ers, but we also have ev­ery­thing from Marathi to Ti­be­tan. You can see the full table here..

NEWCOMB’S PROBLEM
One-box: 726, 61.4%
Two-box: 78, 6.6%
Not sure: 53, 4.5%
Don’t un­der­stand: 86, 7.3%
No an­swer: 240, 20.3%

ENTREPRENEUR:
Don’t want to start busi­ness: 447, 37.8%
Con­sid­er­ing start­ing busi­ness: 334, 28.2%
Plan­ning to start busi­ness: 96, 8.1%
Already started busi­ness: 112, 9.5%
No an­swer: 194, 16.4%

ANONYMITY:
Post us­ing real name: 213, 18%
Easy to find real name: 256, 21.6%
Hard to find name, but wouldn’t bother me if some­one did: 310, 26.2%
Anonymity is very im­por­tant: 170, 14.4%
No an­swer: 234, 19.8%

HAVE YOU TAKEN A PREVIOUS LW SURVEY?
No: 559, 47.3%
Yes: 458, 38.7%
No an­swer: 116, 14%

TROLL TOLL POLICY:
Dis­ap­prove: 194, 16.4%
Ap­prove: 178, 15%
Haven’t heard of this: 375, 31.7%
No opinion: 249, 21%
No an­swer: 187, 15.8%

MYERS-BRIGGS
INTJ: 163, 13.8%
INTP: 143, 12.1%
ENTJ: 35, 3%
ENTP: 30, 2.5%
INFP: 26, 2.2%
INFJ: 25. 2.1%
ISTJ: 14, 1.2%
No an­swer: 715, 60%

This in­cludes all types with greater than 10 peo­ple. You can see the full table here.

Part 3: Numer­i­cal Data

Ex­cept where in­di­cated oth­er­wise, all the num­bers be­low are given in the for­mat:

mean+stan­dard_de­vi­a­tion (25% level, 50% level/​me­dian, 75% level) [n = num­ber of data points]

INTELLIGENCE:

IQ (self-re­ported): 138.7 + 12.7 (130, 138, 145) [n = 382]
SAT (out of 1600): 1485.8 + 105.9 (1439, 1510, 1570) [n = 321]
SAT (out of 2400): 2319.5 + 1433.7 (2155, 2240, 2320)
ACT: 32.7 + 2.3 (31, 33, 34) [n = 207]
IQ (on iqtest.dk): 125.63 + 13.4 (118, 130, 133) [n = 378]

I am go­ing to harp on these num­bers be­cause in the past some peo­ple have been pretty quick to ridicule this sur­vey’s in­tel­li­gence num­bers as com­pletely use­less and im­pos­si­ble and so on.

Ac­cord­ing to IQ Com­par­i­son Site, an SAT score of 1485/​1600 cor­re­sponds to an IQ of about 144. Ac­cord­ing to Ivy West, an ACT of 33 cor­re­sponds to an SAT of 1470 (and thence to IQ of 143).

So if we con­sider self-re­port, SAT, ACT, and iqtest.dk as four mea­sures of IQ, these come out to 139, 144, 143, and 126, re­spec­tively.

All of these are pretty close ex­cept iqtest.dk. I ran a cor­re­la­tion be­tween all of them and found that self-re­ported IQ is cor­re­lated with SAT scores at the 1% level and iqtest.dk at the 5% level, but SAT scores and IQTest.dk are not cor­re­lated with each other.

Of all these, I am least likely to trust iqtest.dk. First, it’s a ran­dom In­ter­net IQ test. Se­cond, it cor­re­lates poorly with the other mea­sures. Third, a lot of peo­ple have com­plained in the com­ments to the sur­vey post that it ex­hibits some weird be­hav­ior.

But iqtest.dk gave us the low­est num­ber! And even it said the av­er­age was 125 to 130! So I sug­gest that we now have pretty good, pretty be­liev­able ev­i­dence that the av­er­age IQ for this site re­ally is some­where in the 130s, and that self-re­ported IQ isn’t as ter­rible a mea­sure as one might think.

AGE:
27.8 + 9.2 (22, 26, 31) [n = 1185]

LESS WRONG USE:
Karma: 1078 + 2939.5 (0, 4.5, 136) [n = 1078]
Months on LW: 26.7 + 20.1 (12, 24, 40) [n = 1070]
Minutes/​day on LW: 19.05 + 24.1 (5, 10, 20) [n = 1105]
Wiki views/​month: 3.6 + 6.3 (0, 1, 5) [n = 984]
Wiki ed­its/​month: 0.1 + 0.8 (0, 0, 0) [n = 984]

PROBABILITIES:
Many Wor­lds: 51.6 + 31.2 (25, 55, 80) [n = 1005]
Aliens (uni­verse): 74.2 + 32.6 (50, 90, 99) [n = 1090]
Aliens (galaxy): 42.1 + 38 (5, 33, 80) [n = 1081]
Su­per­nat­u­ral: 5.9 + 18.6 (0, 0, 1) [n = 1095]
God: 6 + 18.7 (0, 0, 1) [n = 1098]
Reli­gion: 3.8 + 15.5 (0, 0, 0.8) [n = 1113]
Cry­on­ics: 18.5 + 24.8 (2, 8, 25) [n = 1100]
An­ti­a­gath­ics: 25.1 + 28.6 (1, 10, 35) [n = 1094]
Si­mu­la­tion: 25.1 + 29.7 (1, 10, 50) [n = 1039]
Global warm­ing: 79.1 + 25 (75, 90, 97) [n = 1112]
No catas­trophic risk: 71.1 + 25.5 (55, 80, 90) [n = 1095]
Space: 20.1 + 27.5 (1, 5, 30) [n = 953]

CALIBRATION:
Year of Bayes’ birth: 1767.5 + 109.1 (1710, 1780, 1830) [n = 1105]
Con­fi­dence: 33.6 + 23.6 (20, 30, 50) [n= 1082]

MONEY:
In­come/​year: 50,913 + 60644.6 (12000, 35000, 74750) [n = 644]
Char­ity/​year: 444.1 + 1152.4 (0, 30, 250) [n = 950]
SIAI/​CFAR char­ity/​year: 309.3 + 3921 (0, 0, 0) [n = 961]
Aging char­ity/​year: 13 + 184.9 (0, 0, 0) [n = 953]

TIME USE:
Hours on­line/​week: 42.4 + 30 (21, 40, 59) [n = 944]
Hours read­ing/​week: 30.8 + 19.6 (18, 28, 40) [n = 957]
Hours writ­ing/​week: 7.9 + 9.8 (2, 5, 10) [n = 951]

POLITICAL COMPASS:
Left/​Right: −2.4 + 4 (-5.5, −3.4, −0.3) [n = 476]
Liber­tar­ian/​Author­i­tar­ian: −5 + 2 (-6.2, −5.2, −4)

BIG 5 PERSONALITY TEST:
Big 5 (O): 60.6 + 25.7 (41, 65, 84) [n = 453]
Big 5 (C): 35.2 + 27.5 (10, 30, 58) [n = 453]
Big 5 (E): 30.3 + 26.7 (7, 22, 48) [n = 454]
Big 5 (A): 41 + 28.3 (17, 38, 63) [n = 453]
Big 5 (N): 36.6 + 29 (11, 27, 60) [n = 449]

Th­ese scores are in per­centiles, so LWers are more Open, but less Con­scien­tious, Agree­able, Ex­traverted, and Neu­rotic than av­er­age test-tak­ers. Note that peo­ple who take on­line psy­cho­me­t­ric tests are prob­a­bly a pretty skewed cat­e­gory already so this tells us noth­ing. Also, sev­eral peo­ple got con­fus­ing re­sults on this test or found it differ­ent than other tests that they took, and I am pretty un­satis­fied with it and don’t trust the re­sults.

AUTISM QUOTIENT
AQ: 24.1 + 12.2 (17, 24, 30) [n = 367]

This test says the av­er­age con­trol sub­ject got 16.4 and 80% of those di­ag­nosed with autism spec­trum di­s­or­ders get 32+ (which of course doesn’t tell us what per­cent of peo­ple above 32 have autism...). If we trust them, most LWers are more autis­tic than av­er­age.

CALIBRATION:

Rev­erend Thomas Bayes was born in 1701. Sur­vey tak­ers were asked to guess this date within 20 years, so any­one who guessed be­tween 1681 and 1721 was recorded as get­ting a cor­rect an­swer. The per­cent of peo­ple who an­swered cor­rectly is recorded be­low, strat­ified by the con­fi­dence they gave of hav­ing guessed cor­rectly and with the num­ber of peo­ple at that con­fi­dence level.

0-5: 10% [n = 30]
5-15: 14.8% [n = 183]
15-25: 10.3% [n = 242]
25-35: 10.7% [n = 225]
35-45: 11.2% [n = 98]
45-55: 17% [n = 118]
55-65: 20.1% [n = 62]
65-75: 26.4% [n = 34]
75-85: 36.4% [n = 33]
85-95: 60.2% [n = 20]
95-100: 85.7% [n = 23]

Here’s a clas­sic cal­ibra­tion chart. The blue line is perfect cal­ibra­tion. The or­ange line is you guys. And the yel­low line is av­er­age cal­ibra­tion from an ex­per­i­ment I did with un­trained sub­jects a few years ago (which of course was based on differ­ent ques­tions and so not di­rectly com­pa­rable).

The re­sults are atro­cious; when Less Wrongers are 50% cer­tain, they only have about a 17% chance of be­ing cor­rect. On this prob­lem, at least, they are as bad or worse at avoid­ing over­con­fi­dence bias as the gen­eral pop­u­la­tion.

My hope was that this was the re­sult of a lot of lurk­ers who don’t know what they’re do­ing stum­bling upon the sur­vey and mak­ing ev­ery­one else look bad, so I ran a sec­ond anal­y­sis. This one used only the num­bers of peo­ple who had been in the com­mu­nity at least 2 years and ac­cu­mu­lated at least 100 karma; this limited my sam­ple size to about 210 peo­ple.

I’m not go­ing to post ex­act re­sults, be­cause I made some minor mis­takes which means they’re off by a per­centage point or two, but the gen­eral trend was that they looked ex­actly like the re­sults above: atro­cious. If there is some core of elites who are less bi­ased than the gen­eral pop­u­la­tion, they are well past the 100 karma point and prob­a­bly too rare to feel con­fi­dent even de­tect­ing at this kind of a sam­ple size.

I re­ally have no idea what went so wrong. Last year’s re­sults were pretty good—en­courag­ing, even. I won­der if it’s just an es­pe­cially bad ques­tion. Bayesian statis­tics is pretty new; one would ex­pect Bayes to have been born in rather more mod­ern times. It’s also pos­si­ble that I’ve han­dled the statis­tics wrong on this one; I wouldn’t mind some­one dou­ble-check­ing my work.

Or we could just be re­ally hor­rible. If we haven’t even learned to avoid the one bias that we can mea­sure su­per well and which is most sus­cep­ti­ble to train­ing, what are we even do­ing here? Some re­me­dial time at Pre­dic­tionBook might be in or­der.

HYPOTHESIS TESTING:

I tested a very few of the pos­si­ble hy­poth­e­sis that were pro­posed in the sur­vey de­sign threads.

Are peo­ple who un­der­stand quan­tum me­chan­ics are more likely to be­lieve in Many Wor­lds? We perform a t-test, check­ing whether one’s prob­a­bil­ity of the MWI be­ing true de­pends on whether or not one can solve the Schrod­inger Equa­tion. Peo­ple who could solve the equa­tion had on av­er­age a 54.3% prob­a­bil­ity of MWI, com­pared to 51.3% in those who could not. The p-value is 0.26; there is a 26% prob­a­bil­ity this oc­curs by chance. There­fore, we fail to es­tab­lish that peo­ple’s prob­a­bil­ity of MWI varies with un­der­stand­ing of quan­tum me­chan­ics.

Are there any in­ter­est­ing biolog­i­cal cor­re­lates of IQ? We run a cor­re­la­tion be­tween self-re­ported IQ, height, ma­ter­nal age, and pa­ter­nal age. The cor­re­la­tions are in the ex­pected di­rec­tion but not sig­nifi­cant.

Are there differ­ences in the ways men and women in­ter­act with the com­mu­nity? I had sort of vaguely got­ten the im­pres­sion that women were pro­por­tion­ally younger, newer to the com­mu­nity, and more likely to be referred via HPMOR. The av­er­age age of women on LW is 27.6 com­pared to 27.7 for men; ob­vi­ously this differ­ence is not sig­nifi­cant. 14% of the peo­ple referred via HPMOR were women com­pared to about 10% of the com­mu­nity at large, but this differ­ence is pretty minor. Women were on av­er­age newer to the com­mu­nity − 21 months vs. 39 for men—but to my sur­prise a t-test was un­able to de­clare this sig­nifi­cant. Maybe I’m do­ing it wrong?

Does the amount of time spent in the com­mu­nity af­fect one’s be­liefs in the same way as in pre­vi­ous sur­veys? I ran some cor­re­la­tions and found that it does. Peo­ple who have been around longer con­tinue to be more likely to be­lieve in MWI, less likely to be­lieve in aliens in the uni­verse (though not in our galaxy), and less likely to be­lieve in God (though not re­li­gion). There was no effect on cry­on­ics this time.

In ad­di­tion, the clas­sic cor­re­la­tions be­tween differ­ent be­liefs con­tinue to hold true. There is an ob­vi­ous cluster of God, re­li­gion, and the su­per­nat­u­ral. There’s also a scifi cluster of cry­on­ics, an­ti­a­gath­ics, MWI, aliens, and the Si­mu­la­tion Hy­poth­e­sis, and catas­trophic risk (this also seems to in­clude global warm­ing, for some rea­son).

Are there any differ­ences be­tween men and women in re­gards to their be­lief in these clusters? We run a t-test be­tween men and women. Men and women have about the same prob­a­bil­ity of God (men: 5.9, women: 6.2, p = .86) and similar re­sults for the rest of the re­li­gion cluster, but men have much higher be­liefs in for ex­am­ple an­ti­a­gath­ics (men 24.3, women: 10.5, p < .001) and the rest of the scifi cluster.

DESCRIPTIONS OF LESS WRONG

Sur­vey users were asked to sub­mit a de­scrip­tion of Less Wrong in 140 char­ac­ters or less. I’m not go­ing to post all of them, but here is a rep­re­sen­ta­tive sam­ple:

- “Prob­a­bly the most sen­si­ble philo­soph­i­cal re­source ava­ialble.”
- “Con­tains the great Se­quences, some of Luke’s posts, and very lit­tle else.”
- “The cur­rently most in­ter­est­ing site I found ont the net.”
- “EY cult”
- “How to think cor­rectly, pre­cisely, and effi­ciently.”
- “HN for even big­ger nerds.”
- “So­cial skills philos­o­phy and AI the­o­rists on the same site, not notic­ing each other.”
- “Cool place. Any oth­ers like it?”
- “How to avoid pre­dictable pit­falls in hu­man psy­chol­ogy, and un­der­stand hard things well: The Web­site.”
- “A bunch of peo­ple try­ing to make sense of the wold through their own lens, which hap­pens to be one of calcu­la­tion and rigor”
- “Nice.”
- “A font of brilli­ant and un­con­ven­tional wis­dom.”
- “One of the few sane places on Earth.”
- “Robot god apoc­a­lypse cult spinoff from Harry Pot­ter.”
- “A place to con­verse with in­tel­li­gent, rea­son­ably open-minded peo­ple.”
- “Cal­la­han’s Crosstime Saloon”
- “Amaz­ing ra­tio­nal tran­shu­man­ist calming ad­dict­ing Su­per Red­dit”
- “Still wrong”
- “A fo­rum for helping to train peo­ple to be more ra­tio­nal”
- “A very bright com­mu­nity in­ter­ested in am­a­teur eth­i­cal philos­o­phy, math­e­mat­ics, and de­ci­sion the­ory.”
- “Dy­ing. So­cial games and bul­lshit now >50% of LW con­tent.”
- “The good kind of strange, ad­dic­tive, so much to read!”
- “Part gen­uinely use­ful, part men­tal mas­tur­ba­tion.”
- “Mostly very bright and starry-eyed adults who never quite grew out of their sci­ence-fic­tion ad­dic­tion as ado­les­cents.”
- “Less Wrong: Sav­ing the world with MIND POWERS!”
- “Perfectly pat­tern­matches the ‘young-peo­ple-with-all-the-an­swers’ cliche”
- “Ra­tion­al­ist com­mu­nity ded­i­cated to self-im­prove­ment.”
- “Sper­glord hip­sters pre­tend­ing that be­ing a sper­glord hip­ster is cool.” (this per­son’s Autism Quo­tient was two points higher than LW av­er­age, by the way)
- “An in­ter­est­ing per­spec­tive and valuable database of men­tal tech­niques.”
- “A web­site with ker­nels of in­for­ma­tion hid­den among aspy non­sense.”
- “Ex­clu­sive, elitist, in­ter­est­ing, po­ten­tially use­ful, per­sonal de­pres­sion trig­ger.”
- “A group blog about ra­tio­nal­ity and re­lated top­ics. Tends to be overzeal­ous about cryo­gen­ics and other pet ideas of Eliezer Yud­kowsky.”
- “Things to read to make you think bet­ter.”
- “Ex­cel­lent ra­tio­nal­ity. New-age self-help. Wor­ry­ing group­think.”
- “Not a cult at all.”
- “A cult.”
- “The new thing for peo­ple who would have been Ran­dian Ob­jec­tivists 30 years ago.”
- “Fas­ci­nat­ing, well-started, risk­ing bloat and failure modes, best as archive.”
- “A fun, in­sight­ful dis­cus­sion of prob­a­bil­ity the­ory and cog­ni­tion.”
- “More in­ter­est­ing than use­ful.”
- “The most pro­duc­tive and ac­cessible mind-fuck­ery on the In­ter­net.”
- “A blog for ra­tio­nal­ity, cog­ni­tive bias, fu­tur­ism, and the Sin­gu­lar­ity.”
- “Robo-Protes­tants at­tempt­ing nat­u­ral the­ol­ogy.”
- “Orderly quag­mire of tan­ta­l­iz­ing ideas drawn from dis­agree­able pri­ors.”
- “An­a­lyze ev­ery­thing. And I do mean ev­ery­thing. In­clud­ing anal­y­sis. Espe­cially anal­y­sis. And anal­y­sis of anal­y­sis.”
- “Very in­ter­est­ing and some­times use­ful.”
- “Where peo­ple dis­cuss and try to im­ple­ment ways that hu­mans can make their val­ues, ac­tions, and be­liefs more in­ter­nally con­sis­tent.”
- “Eliezer Yud­kowsky per­son­al­ity cult.”
- “It’s like the Mor­mons would be if ev­ery­one were an athe­ist and good at math and didn’t ab­stain from sub­stances.”
- “Seems wacky at first, but grad­u­ally be­gins to seem nor­mal.”
- “A varied group of peo­ple in­ter­ested in philos­o­phy with high Open­ness and a me­thod­i­cal yet am­a­teur ap­proach.”
- “Less Wrong is where hu­man al­gorithms go to de­bug them­selves.”
- “They’re kind of like a cult, but that doesn’t make them wrong.”
- “A com­mu­nity blog de­voted to nerds who think they’re smarter than ev­ery­one else.”
- “90% sane! A new record!”
- “The Se­quences are great. LW now slowly de­gen­er­at­ing to just an­other sci­ence fo­rum.”
- “The meetup groups are where it’s at, it seems to me. I re­serve judg­ment till I at­tend one.”
- “All I re­ally know about it is this long sur­vey I took.”
- “The royal road of ra­tio­nal­ity.”
- “Tech­ni­cally cor­rect: The best kind of cor­rect!”
- “Full of an­gry priv­ilege.”
- “A sinister in­stru­ment of billion­aire Peter Thiel.”
- “Danger­ous apoc­a­lypse cult bent on the sys­tem­atic era­sure of tra­di­tional val­ues and cul­ture by any means nec­es­sary.”
- “Often in­ter­est­ing, but I never feel at home.”
- “One of the few places I truly feel at home, know­ing that there are more peo­ple like me.”
- “Cur­rently the best in­ter­net source of in­for­ma­tion-dense ma­te­rial re­gard­ing cog sci, de­bi­as­ing, and ex­is­ten­tial risk.”
- “Pro­lific and eru­dite writ­ing on prac­ti­cal tech­niques to en­hance the effec­tive­ness of our rea­son.”
- “An em­bar­rass­ing In­ter­net com­mu­nity formed around some gen­uinely great blog writ­ings.”
- “I book­marked it a while ago and com­pletely for­got what it is about. I am tak­ing the sur­vey to while away my in­som­nia.”
- “A some­what in­timi­dat­ing but re­ally in­ter­est­ing web­site that helps re­fine ra­tio­nal think­ing.”
- “A great col­lec­tion of ways to avoid sys­tem­atic bias and come to true and use­ful con­clu­sions.”
- “Ob­nox­ious self-serv­ing, fool­ish trol­ling de­hu­man­iz­ing pseu­doin­tel­lec­tu­al­ism, aes­thet­i­cally bankrupt.”
- “The cut­ting edge of hu­man ra­tio­nal­ity.”
- “A pur­veyor of ex­ceed­ingly long sur­veys.”

PUBLIC RELEASE

That last com­menter was right. This sur­vey had vastly more data than any pre­vi­ous in­car­na­tion; al­though there are many more analy­ses I would like to run I am pretty ex­hausted and I know peo­ple are anx­ious for the re­sults. I’m go­ing to let CFAR an­a­lyze and re­port on their ques­tions, but the rest should be a com­mu­nity effort. So I’m re­leas­ing the sur­vey to ev­ery­one in the hopes of get­ting more in­for­ma­tion out of it. If you find some­thing in­ter­est­ing you can ei­ther post it in the com­ments or start a new thread some­where.

The data I’m pro­vid­ing is the raw data EXCEPT:

- I deleted a few cat­e­gories that I re­moved halfway through the sur­vey for var­i­ous rea­sons
- I deleted 9 en­tries that were du­pli­cates of other en­tries, ie some­one pressed ‘sub­mit’ twice.
- I deleted the times­tamp, which would have made peo­ple ex­tra-iden­ti­fi­able, and sorted peo­ple by their CFAR ran­dom num­ber to re­move time or­der in­for­ma­tion.
- I re­moved one per­son whose in­for­ma­tion all came out as weird sym­bols.
- I nu­mer­al­ized some of the non-nu­meric data, es­pe­cially on the num­ber of months in com­mu­nity ques­tion. This is not the ver­sion I cleaned up fully, so you will get to ex­pe­rience some of the same plea­sure I did work­ing with the rest.
- I deleted 117 peo­ple who ei­ther didn’t an­swer the pri­vacy ques­tion or who asked me to keep them anony­mous, leav­ing 1067 peo­ple.

Here it is: Data in .csv for­mat , Data in Ex­cel format