If it checks out, that’d give a way to infer Conscientiousness scores of the respondents who didn’t report their Big Five.
What would we do with this Conscientiousness data per survey respondent? Dunno. Off the top of my head, we could construct a baseball-like index of ‘most under-valued LWers by comparing their IQ and Conscientiousness against their self-reported salary & age’ :)
One cool thing we could do to check the accuracy of the Big Five scores, Conscientiousness specifically, for each user is their item non-response rate, per http://econlog.econlib.org/archives/2012/11/hedengrens_dog.html / http://dl.dropbox.com/u/6776620/Papers/The%20dog%20that%20didnt%20bark%2011-8-2012.pdf
If it checks out, that’d give a way to infer Conscientiousness scores of the respondents who didn’t report their Big Five.
What would we do with this Conscientiousness data per survey respondent? Dunno. Off the top of my head, we could construct a baseball-like index of ‘most under-valued LWers by comparing their IQ and Conscientiousness against their self-reported salary & age’ :)
EDIT: The technique seems to completely fail when I try it on the survey: http://lesswrong.com/lw/fp5/2012_survey_results/7xl5