3

As a beginner statistician, discussing the need to be accurate with the interpretation of statistical results with non-statistician is not an easy task. In particular, I am trying to convince some colleagues not to dive into the common mis-interpretation of confidence-interval (maybe I should not?).

For some problem, there is some illustrating and highlighting (and funny) paper as this one (http://www.jsur.org/ar/jsur_ben102010.pdf) about the need of multiple testing correction for whose I extract a small citation:

[...] we completed an fMRI scanning session with a post-mortem Atlantic Salmon as the subject. The salmon was shown the same social perspective-taking task that was later administered to a group of human subjects."

that is a terrific argument to encourage practioner to use multiple testing corrections.

Are you aware of such a paper about the common mis-interpretation of confidence interval (that are typically interpreted as credibility-interval) ? or any other than can help me in my arguments.

peuhp
  • 4,622
  • 20
  • 38
  • You might find http://stats.stackexchange.com/questions/13655/what-does-a-confidence-interval-vs-a-credible-interval-actually-express to be interesting. – abaumann Jun 13 '13 at 11:38
  • 1
    Thanks for your suggestion but ideally I am looking for published material with a real world application... but this seems difficult to find. – peuhp Jun 17 '13 at 08:43

0 Answers0