Something I would quite like to see after looking at this post: a poll of LW users’ stances on polarised political issues.
There are a whole host of issues which we don’t discuss for fear of mindkilling. While I would expect opinion to be split on a lot of politically sensitive subjects, I would be fascinated to see if the LW readership came down unilaterally on some unexpected issue. I’d also be interested to see if there are any heavily polarised political issues that I currently don’t recognise as such.
I would be astonished if one result of such a poll was not quite a lot of discussion of the polarized political issues that we don’t discuss for fear of mindkilling. Whether that’s a bad thing or not depends on your beliefs about such discussion, of course.
Also, if what you’re interested in is (a) issues where we all agree, and (b) issues you don’t think of as polarized political issues in the first place, it seems a poll is neither necessary nor sufficient for your goals. For any stance S, you can find out whether S is in class (a) by writing up S and asking if anyone disagrees. And no such poll will turn up results about any issue the poll creator(s) didn’t consider controversial enough to include in the poll.
That said, I’d be vaguely interested (not enough to actually do any work to find out) in how well LW users can predict how popular various positions are on LW, and how well/poorly accuracy in predicting the popularity of a position correlates with holding that position among LW users.
0) Prohibit actual discussion of the subjects in question, with the understanding that comments transgressing this rule would be downvoted to oblivion by a conscientious readership (as they generally are already)
1) Request suggestions for dichotomies that people believe would split popular opinion. Let people upvote and downvote these on the basis of whether they’d be fit for the purpose of the poll.
2) Take the most popular dichotomies and put them in a poll, with a “don’t care” and “wrong dichotomy” option, which I hope are fairly self-explanatory.
2) a) To satisfy your curiosity on how well LW users can predict the beliefs of other LW users, also have a “what do you think most LW users would pick as an answer to this question?” option.
3) Have people vote, and see what patterns emerge.
Something I would quite like to see after looking at this post: a poll of LW users’ stances on polarised political issues.
There are a whole host of issues which we don’t discuss for fear of mindkilling. While I would expect opinion to be split on a lot of politically sensitive subjects, I would be fascinated to see if the LW readership came down unilaterally on some unexpected issue. I’d also be interested to see if there are any heavily polarised political issues that I currently don’t recognise as such.
Why would this be a bad idea?
I would be astonished if one result of such a poll was not quite a lot of discussion of the polarized political issues that we don’t discuss for fear of mindkilling. Whether that’s a bad thing or not depends on your beliefs about such discussion, of course.
Also, if what you’re interested in is (a) issues where we all agree, and (b) issues you don’t think of as polarized political issues in the first place, it seems a poll is neither necessary nor sufficient for your goals. For any stance S, you can find out whether S is in class (a) by writing up S and asking if anyone disagrees. And no such poll will turn up results about any issue the poll creator(s) didn’t consider controversial enough to include in the poll.
That said, I’d be vaguely interested (not enough to actually do any work to find out) in how well LW users can predict how popular various positions are on LW, and how well/poorly accuracy in predicting the popularity of a position correlates with holding that position among LW users.
How I imagined it going:
0) Prohibit actual discussion of the subjects in question, with the understanding that comments transgressing this rule would be downvoted to oblivion by a conscientious readership (as they generally are already)
1) Request suggestions for dichotomies that people believe would split popular opinion. Let people upvote and downvote these on the basis of whether they’d be fit for the purpose of the poll.
2) Take the most popular dichotomies and put them in a poll, with a “don’t care” and “wrong dichotomy” option, which I hope are fairly self-explanatory.
2) a) To satisfy your curiosity on how well LW users can predict the beliefs of other LW users, also have a “what do you think most LW users would pick as an answer to this question?” option.
3) Have people vote, and see what patterns emerge.