Talk:Randomized response
This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||||||||||
|
Doesn't work with heads/tails
editAs stated, I think it's wrong. With heads/tails (or any other 50/50 split), you always get the same result, regardless of the underlying properties. Work it out for yourself. Stevage 07:48, 11 June 2010 (UTC)
- Have you read the article? It says "For mathematical reasons chance cannot be "fair" (½ and ½)." --Arno Matthias (talk) 10:28, 11 June 2010 (UTC)
- Actually I was mistaken, but not for the reason you mention (you're referring to the "Original Version" - I was referring to the "Example"). I misread it as answering truthfully/falsely depending on the coin, but it's actually tell the truth/answer yes. Stevage 05:38, 17 June 2010 (UTC)
- May we hope then that you won't be using sneer like "Work it out for yourself" a lot in the future? --Arno Matthias (talk) 09:19, 18 June 2010 (UTC)
- You read that as sneering? My apologies, it certainly wasn't intended as such. Rather, "try the experiment yourself, and see if you agree with me". Stevage 06:47, 24 June 2010 (UTC)
- May we hope then that you won't be using sneer like "Work it out for yourself" a lot in the future? --Arno Matthias (talk) 09:19, 18 June 2010 (UTC)
"Half the people"
editand "the true fraction is". These two phrases make it seem like nothing probabilistic is happening. Maybe it should say something like "approximalety half the people" and then "the true fraction is expected to be" — Preceding unsigned comment added by 71.33.5.50 (talk) 03:08, 31 March 2013 (UTC)
Sample Size
editWhat is the desired sample size for using the randomized response method? Is there a formula for calculating this described in the academic literature on this topic? Usually entries on statistical methods in research (such as ANOVA) state the minimum number of cases, ways of calculating that, etc. Thelema418 (talk) 18:21, 28 May 2015 (UTC)
- The randomized response technique is a method of collecting data and as such not tied to a particular sample size, see Sample size determination. --Arno Matthias (talk) 12:49, 29 May 2015 (UTC)
Potential ref
editRef about marijuana survey carried out by US http://files.eric.ed.gov/fulltext/ED133636.pdf Jonpatterns (talk) 13:24, 28 February 2016 (UTC)