More fundamentally than self-labelling, that’s an utterly false dilemma. It helps show that the results weren’t a totally random ‘people on that site then’: they show SOMETHING. But what they show must be much more open to debate. To ‘rationalist’, you can add
1) Has been exposed to LessWrong (sequences and community)
2) English-speaking (unless there were translations?)
3) Minded to take long online surveys: including at the least possibilities
3a) Egotistical enough to think that your survey results must be included
3b) Dedicated enough to the LessWrong community to wish to contribute
3c) Generally publically-minded
3d) Doesn’t have enough to do
4) Likely to overestimate one’s own IQ
It seems particularly odd to suggest these results are representative of rationalists while recognising both that the proportion of women has tripled since the last survey (and I don’t think we’re very close to working out what the true proportion is) and that men and women tend to have significantly different attitudes.
The ‘direct line to prevailing rationalist opinion’ is also straight after what I would guess is most skewed by point (1) above. I’d be shocked to see such high scores for Many Worlds, living in a simulation or cryonics amongst rationalists outside LessWrong.
Finally, could the last set of results itself have had an effect. The most likely effect would be in confirming the in-group opinions, leading to ‘evaporative cooling’ (if I may!). It seems less likely, but people could have directly calibrated too: I’d be interested in how much that page was accessed ahead of people taking this year’s survey.
If ‘rationalist’ was used just to mean ‘LessWrongian’ then please ignore the above—and take Robert Lumley’s advice!
http://lesswrong.com/lw/82s/dont_call_yourself_a_rationalist/
More fundamentally than self-labelling, that’s an utterly false dilemma. It helps show that the results weren’t a totally random ‘people on that site then’: they show SOMETHING. But what they show must be much more open to debate. To ‘rationalist’, you can add 1) Has been exposed to LessWrong (sequences and community) 2) English-speaking (unless there were translations?) 3) Minded to take long online surveys: including at the least possibilities 3a) Egotistical enough to think that your survey results must be included 3b) Dedicated enough to the LessWrong community to wish to contribute 3c) Generally publically-minded 3d) Doesn’t have enough to do 4) Likely to overestimate one’s own IQ
It seems particularly odd to suggest these results are representative of rationalists while recognising both that the proportion of women has tripled since the last survey (and I don’t think we’re very close to working out what the true proportion is) and that men and women tend to have significantly different attitudes.
The ‘direct line to prevailing rationalist opinion’ is also straight after what I would guess is most skewed by point (1) above. I’d be shocked to see such high scores for Many Worlds, living in a simulation or cryonics amongst rationalists outside LessWrong.
Finally, could the last set of results itself have had an effect. The most likely effect would be in confirming the in-group opinions, leading to ‘evaporative cooling’ (if I may!). It seems less likely, but people could have directly calibrated too: I’d be interested in how much that page was accessed ahead of people taking this year’s survey.
If ‘rationalist’ was used just to mean ‘LessWrongian’ then please ignore the above—and take Robert Lumley’s advice!