After reading the feedback I’ve made the following changes (after the first 104 entries so that anyone who has access to the data can check if there are significant differences before and after these changes):
Added an “other” option in gender
Moved “date of singularity” above question mentioning 2100 to avoid anchoring. Really I should also move the Newton question for the same reason, but I’m not going to.
changed wording of anti-agathics question to “at least one person”
added a “don’t know / no preference” to relationship style
clarified to answer probability as percent and not decimal; I’ll go back and fix anyone who got this wrong, though. If you seriously mean a very low percent, like ”.05%”, please end with a percent mark so I know not to change it. Otherwise, leave the percent mark out.
Added a “government work” option.
Deleted “divorced”. Divorced people can just put “single”
Added “economic/political collapse” to xrisk
Added “other” to xrisk
Added a question “Have you ever been to a Less Wrong meetup?” Please do NOT retake the survey to answer this question. I’ll just grab statistics from the people who answered this after it was put up, while recognizing it might be flawed.
I did NOT add an “Other” to politics despite requests to do so, because I tried this last time and ended up with people sending me manifestos. I want to encourage people to choose whichever of those categories they’re closest to. If you really don’t identify at all with any of those categories, just leave it blank.
Could you also add an ATAR/UAI, A-levels, Abitur and IB option?
(It might be better to add a box asking for marks/certificate received upon leaving high school and the name of the program; with sufficient respondents there may be enough data to say meaningful things)
Also, do I understand you correctly that the beings (conceivably) running the universe as a simulation do not count as supernatural/gods for purposes of the supernatural/gods questions?
Yeah, I thought the theism question was the worst of all. Have you ever met a theist that answered “ontologically basic mind-stuff” when asked what God is? Me neither.
I feel like several of the single-punch questions should be multi-punch. Both “profession” and “Work status” gave me pause. Also, I had to figure out what the right thing to fill in for “family religion” was, since we had several.
And there are several extremely common moral views not represented in your list of moral theories. One of the more popular is “All moral theories have some grain of truth, and we should use a combination along with our intuition”. For questions like this, you might use as your model the Philpapers survey, though I also worry that this question might not make a lot of sense to most people without at least rough definitions alongside the answer choices.
About the government work issue, if I work for an aerospace company that gets all of its business from the government, does that count as “for profit” or “government work” for purposes of the question?
I changed my estimated probabilities to reflect percentages, but didn’t mark them with a percent sign because the version of the survey I took explicitly said not to.
It’s mostly irrelevant anyway, these probabilities weren’t even accurate to two orders of magnitude.
After reading the feedback I’ve made the following changes (after the first 104 entries so that anyone who has access to the data can check if there are significant differences before and after these changes):
Added an “other” option in gender
Moved “date of singularity” above question mentioning 2100 to avoid anchoring. Really I should also move the Newton question for the same reason, but I’m not going to.
changed wording of anti-agathics question to “at least one person”
added a “don’t know / no preference” to relationship style
clarified to answer probability as percent and not decimal; I’ll go back and fix anyone who got this wrong, though. If you seriously mean a very low percent, like ”.05%”, please end with a percent mark so I know not to change it. Otherwise, leave the percent mark out.
Added a “government work” option.
Deleted “divorced”. Divorced people can just put “single”
Added “economic/political collapse” to xrisk
Added “other” to xrisk
Added a question “Have you ever been to a Less Wrong meetup?” Please do NOT retake the survey to answer this question. I’ll just grab statistics from the people who answered this after it was put up, while recognizing it might be flawed.
I did NOT add an “Other” to politics despite requests to do so, because I tried this last time and ended up with people sending me manifestos. I want to encourage people to choose whichever of those categories they’re closest to. If you really don’t identify at all with any of those categories, just leave it blank.
Should anyone retake the survey? I’d be willing to if you can cancel the my first version—I’ll give the same answers on the Newton question.
Not as good as if someone can find a satisfactory IQ test, but could you add an SAT option for intelligence measurement?
I used percents for all my probabilities, including the one which was .5.
Could you also add an ATAR/UAI, A-levels, Abitur and IB option?
(It might be better to add a box asking for marks/certificate received upon leaving high school and the name of the program; with sufficient respondents there may be enough data to say meaningful things)
Some of us are still in high school.
I’d also be willing (I’d probably rather) retake the quiz.
But there is a problem with calibration at that point, with the question about Newton.
Also, do I understand you correctly that the beings (conceivably) running the universe as a simulation do not count as supernatural/gods for purposes of the supernatural/gods questions?
Yeah, I thought the theism question was the worst of all. Have you ever met a theist that answered “ontologically basic mind-stuff” when asked what God is? Me neither.
Other than that, thanks Yvain!
I assumed when taking the survey that those running the simulation are outside our universe and so ontologically basic.
Those people outside the simulation could exist with or without ontologically basic mental features. The questions are totally orthogonal.
Is there even theoretical way of distinguishing these two cases? I’d assume the only possible answer is “they” do.
I think the percentage of LW meetup attendees is positively correlated with how quickly people take the poll, unfortunately.
I feel like several of the single-punch questions should be multi-punch. Both “profession” and “Work status” gave me pause. Also, I had to figure out what the right thing to fill in for “family religion” was, since we had several.
And there are several extremely common moral views not represented in your list of moral theories. One of the more popular is “All moral theories have some grain of truth, and we should use a combination along with our intuition”. For questions like this, you might use as your model the Philpapers survey, though I also worry that this question might not make a lot of sense to most people without at least rough definitions alongside the answer choices.
About the politics question: What if you come from a place/subculture where none of those even exist and have barely even heard of them?
About the government work issue, if I work for an aerospace company that gets all of its business from the government, does that count as “for profit” or “government work” for purposes of the question?
I believe that would usually be considered “for profit”.
I changed my estimated probabilities to reflect percentages, but didn’t mark them with a percent sign because the version of the survey I took explicitly said not to.
It’s mostly irrelevant anyway, these probabilities weren’t even accurate to two orders of magnitude.