Since my little survey about the degree to which the Less Wrong community has a preponderance of people with systematizing personality types, I’ve been collecting responses only from those people who considered taking the survey after looking at the original post, but didn’t, in order to combat nonresponse bias.
82 people responded to the initial survey, and another 186 responded after the request for non-responders to respond. In the initial survey, 26% of responders scored 32+ (which is considered to be a “high” score, and out of a group of Cambridge mathematics students, 7 out of 11 who scored over 32 were said to fit the full diagnostic criteria for aspergers syndrome after being interviewed).
In the combined survey of 82 initial responders and 186 “second”-responders, this increased to 28%. In the original survey, 5% of respondents said they had already been diagnosed with aspergers syndrome, and in the combined survey this increased to 7.5%.
Overall, this indicates that response bias is probably not significantly skewing our picture of the LW audience, though, as always, it is possible that there is a more sophisticated bias at work and that these 268 people are not representative of LW.
Aspergers Survey Re-results
Followup to: Aspergers Poll results
Since my little survey about the degree to which the Less Wrong community has a preponderance of people with systematizing personality types, I’ve been collecting responses only from those people who considered taking the survey after looking at the original post, but didn’t, in order to combat nonresponse bias.
82 people responded to the initial survey, and another 186 responded after the request for non-responders to respond. In the initial survey, 26% of responders scored 32+ (which is considered to be a “high” score, and out of a group of Cambridge mathematics students, 7 out of 11 who scored over 32 were said to fit the full diagnostic criteria for aspergers syndrome after being interviewed).
In the combined survey of 82 initial responders and 186 “second”-responders, this increased to 28%. In the original survey, 5% of respondents said they had already been diagnosed with aspergers syndrome, and in the combined survey this increased to 7.5%.
Overall, this indicates that response bias is probably not significantly skewing our picture of the LW audience, though, as always, it is possible that there is a more sophisticated bias at work and that these 268 people are not representative of LW.