Significance testing: a picture (well, cartoon) is worth 1000 words.

As of the 23rd May 2022 this website is archived and will receive no further updates.

understandinguncertainty.org was produced by the Winton programme for the public understanding of risk based in the Statistical Laboratory in the University of Cambridge. The aim was to help improve the way that uncertainty and risk are discussed in society, and show how probability and statistics can be both useful and entertaining.

Many of the animations were produced using Flash and will no longer work.

Statistical significance testing is a pretty tricky concept. We're planning to post an article on it soon, but until we get round to it, here's a link to something excellent on the topic from the xkcd Web comic. They're illustrating exactly what people misunderstand. Maybe we needn't bother with the article after all...

Levels: 

Comments

Weren't they lucky that exactly 1 of their 20 colour tests gave a result significant at the 5% level?

Almost everyone I've shown this cartoon to is unable to clearly describe why or what it's showing - as far as it goes it's a in joke for people who understand the misunderstanding....

I love xkcd. What does R say? > dbinom(1,20,0.05) [1] 0.3773536 So it happens about 38% of the time if the null is true.