View Single Post

Default 

March 24th, 2013, 21:59
Yep the UK one is more recent, though also less extreme. Still a little disturbing, but nowhere nearly as much as the survey you linked.

Regarding large surveys and outliers, it's more complicated then that. If 20% of the respondents gave an extreme answer for shock value, then they would not be outliers, and a large survey would not weed that out. A large survery is not a catch all solution for methodological weaknesses in a study.

Originally Posted by JemyM View Post
I list similar, more modern research from UK in #38 in this thread.
In large surveys probability theory weeds out outliers.
(an outlier is a participant answering different than expected, such as all-yes, all-no or all-extreme).
Last edited by fadedc; March 24th, 2013 at 22:16.
fadedc is offline

fadedc

Keeper of the Watch

#48

Join Date: Apr 2011
Posts: 1,014