I saw this on XKCD this morning, and thought it was hilarious.
Comments
Thanks johnnyb. It was 11.30pm last night here when I wrote that. Didn't pay close enough attention.Janice
April 6, 2011
April
04
Apr
6
06
2011
02:52 PM
2
02
52
PM
PDT
Janice -
Pay close attention to the comic. There are 20 tests performed. One out of the 20 (the green one) has p < 0.05 (next to last one on the third row).
But that's the humor of it - out of 20 tests, one of them has a 95% confidence ratio, but for every twenty tests, one test with a 95% confidence ration will be wrong anyway!johnnyb
April 6, 2011
April
04
Apr
6
06
2011
06:21 AM
6
06
21
AM
PDT
p is supposed to be less (not greater) than 0.05 to be significant.Janice
April 6, 2011
April
04
Apr
6
06
2011
06:12 AM
6
06
12
AM
PDT
Well to be fair… scientists usually publish their results in peer reviewed journals.
Some are now talking about their research in blogs and books.
I'm not saying journalists aren't at fault - clearly some of them are. But some scientists are entirely willing to boost their research, whether through press releases, books, or otherwise. In fact, some scientists are more than willing to allow their credentials to be used for this or that group (PETA, the tobacco lobby, you name it.) They are not all (to quote Vox Day) emotionless golems powered by the pure unfiltered Spirit of the Scientific Method.nullasalus
April 6, 2011
April
04
Apr
6
06
2011
05:59 AM
5
05
59
AM
PDT
null -
Well, I think the perspective of xkcd is that *of* the scientists. So, yeah, it will largely blame the journalists.
What is probably most at fault is the funding process.johnnyb
April 6, 2011
April
04
Apr
6
06
2011
05:44 AM
5
05
44
AM
PDT
Well to be fair... scientists usually publish their results in peer reviewed journals.
Some are now talking about their research in blogs and books. But for the most part... the popular press reports on science is written by reporters and science writers. Not the actual scientists. Though I think Scientific American magazine used to have quite a few articles written by the actual scientist doing the research.utidjian
April 6, 2011
April
04
Apr
6
06
2011
05:34 AM
5
05
34
AM
PDT
What sometimes bothers me about xkcd is the suggestion (maybe I've just read the wrong comics) that if scientific results are misconstrued or overblown, journalists are responsible.
Never, you know... scientists.nullasalus
