8

I was inspired by this question on statistics jokes and wanted to start a similar thread on statistical anecdotes.

These are famous stories from the past where statistics have been used to solve some problem in an interesting way.

I like to tell my students these stories in between, as I feel that they sometime spark interest in people that are less motivated to learn about statistics.

I'll start with three of my favorites:

Poincaré and the baker

The German tank problem

The lady tasting tea

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Gumeo
  • 3,551
  • 1
  • 21
  • 31
  • I'm voting to close this question as off-topic because it does not fit the SE model of posting questions that ought to have objectively recognizable good answers. – whuber Oct 13 '15 at 12:47
  • @whuber Ok, I will take the question down. But what makes this any different from the statistics joke question? Why was that allowed in the first place? Also what would be the correct SE venue for such a question? Academia perhaps? – Gumeo Oct 13 '15 at 12:51
  • 5
    I don't feel qualified to judge whether this is in the scope / norms of CV, but the question certainly sounded useful to me, e.g. for teaching stats. – Florian Hartig Oct 13 '15 at 14:41
  • It's not at all different from the joke question--and that's precisely why it was closed. Please see my comment to that question at http://stats.stackexchange.com/questions/1337/statistics-jokes#comment39045_1337 . I believe this kind of question is not on-topic anywhere on SE. – whuber Oct 13 '15 at 21:38
  • The standards across the entirety of the SE network are changed from what they were in the early days. Many SE sites have old posts that break the present rules or in some cases the presently accepted interpretation of what were already existing rules back then (and such posts would normally close retrospectively on that basis) but some are retained for essentially "historical reasons". New posts that break the same rules cannot survive. It would be nice if we could add a "Post Notice" to that effect on those old posts (to stop them from encouraging others), but I don't see a way to achieve it – Glen_b Oct 13 '15 at 23:14
  • Thanks for the clarification @Glen_b . Is the correct action now for me to delete the post or should it remain on hold indefinitely? – Gumeo Oct 14 '15 at 07:33
  • It shouldn't be necessary for you to do anything; but you couldn't even if you wished to because there's an upvoted answer. It's probably better if it stays around in any case, so that people who are inspired as you were and try do the same again will be likely to turn this one up in a search. If there's reason to delete it, that can be done later. Since it's on hold, some users can vote to delete it after 2 days, but the existence of upvotes on both question and answer means none of the autodelete processes will apply. – Glen_b Oct 14 '15 at 08:19
  • Thanks @Glen_b I just noticed that I couldn't delete it. I am fairly new on SE w.r.t. answering and submitting questions. Hope this serves it's purpose. – Gumeo Oct 14 '15 at 08:22
  • I would add that I think it's an excellent question and it would be likely to draw some interesting and useful answers, but it doesn't fit SE's model. – Glen_b Oct 14 '15 at 08:22
  • Ok thanks again @Glen_b I'll leave a comment here if I find another venue for this question! – Gumeo Oct 14 '15 at 08:23
  • @GuðmundurEinarsson try reddit! The statistics and ML communities there are very active and have some amazing people. See https://www.reddit.com/r/statistics and https://www.reddit.com/r/machinelearning – Pedro Tabacof Oct 14 '15 at 14:13
  • Thanks @PedroTabacof I will consider that. I just felt that the crowd here is more relevant and this community is much more active. – Gumeo Oct 14 '15 at 18:40
  • 1
    This comment by Glen_b is on the mark: "It's probably better if it stays around in any case, so that people who are inspired as you were and try do the same again will be likely to turn this one up in a search." I came across this thread because I intended to post something similar and wanted to avoid duplicating any existing threads. I would've put out a call for interesting historical anecdotes, in the same vein as the one about Wald and the bullet holes in the Allied planes. Because this was left up, I now know that it could be considered off-topic for StackExchange. – SQLServerSteve Jun 03 '16 at 13:39

1 Answers1

5

Not sure if this is too serious, but there is the Berkeley gender bias case illustrating Simpson's paradox. A very recent nearly identical case illustrating the problem seems a PNAS study on gender bias in research funding, discussed, e.g., in this Science news article.

Florian Hartig
  • 6,499
  • 22
  • 36