Famous statistical quotations
All models are wrong, but some are useful. (George E. P. Box)
Reference: Box & Draper (1987), Empirical model-building and response surfaces, Wiley, p. 424.
Also: G.E.P. Box (1979), "Robustness in the Strategy of Scientific Model Building" in Robustness in Statistics (Launer & Wilkinson eds.), p. 202.
I use this quote a lot to explain the difficulties in mathematicians transitioning to statistics
And this is an actual quote, as opposed to something "attributed to" Box. It appears, e.g., in Box & Draper (1987), *Empirical model-building and response surfaces*, Wiley, on page 424. Yes, I did go and look it up before using it in a paper.
Sadly, too many people use it to excuse themselves from the flaws in their models. In my personal experience, it's usage is an alarm sign.
I prefer the extended version: "...all models are approximations. Essentially, all models are wrong, but some are useful" (Box & Draper, 2007, "Response Surfaces, Mixtures, and Ridge Analyses", p. 414)
Taken out of context it is a meaningless and even misleading statement. A model helps us understand the world by simplification and by disregarding anomalies, so obviously any model is "wrong" in any literal sense of the word. Furthermore, usefulness should not generally be a criterion for model selection, e.g., a poor model ("climate change is not man made") can still be very useful for nefarious (and other) purposes.
"An approximate answer to the right problem is worth a good deal more than an exact answer to an approximate problem." -- John Tukey
I like this one, could be put as an advise when people write questions on this site ?
I remember once where a private industry company commissioned a mathematician to solve a garbage collection routing problem. Long story short, the mathematician complained that the company was only interested in finding a "close enough" solution rather than an optimal solution. I think, ultimately he was fired, and an operations researcher was brought in instead.
@dassouki I think the quote is more about the question .... something like science is not about finding good answer but about finding good questions !
This reminds me of a quote made by Edwin Jaynes. It roughly goes "...a mathematician came to me and said 'I found a brilliant solution, all I need now is the problem'..."
"Far better an approximate answer to the right question, which is often vague, than an exact answer to the wrong question, which can always be made precise." John W. Tukey 1962 The future of data analysis. Annals of Mathematical Statistics 33: 1-67 (see pp.13-14) No doubt he said similar things at other times, but that's a precise source, and the version I usually see quoted.
"To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of."
-- Ronald Fisher (1938)
The quotation can be read on page 17 of the article.
R. A. Fisher. Presidential Address by Professor R. A. Fisher, Sc.D., F.R.S. Sankhyā: The Indian Journal of Statistics (1933-1960), Vol. 4, No. 1 (1938), pp. 14-17. http://www.jstor.org/stable/40383882
I read a slightly different version of this quote by Fisher: "Hiring a physician after the data have been collected is like hiring a physician when the patient is in the morgue. He may be able to tell you what went wrong, but he is unlikely to be able to fix it."
@Peter Was it really "Hiring a physician after the data ..." or should "statistician" be in there somewhere?
87% of statistics are made up on the spot
In God we trust. All others must bring data.
(W. Edwards Deming)
Ooh, is that a new version of the Omnipotence Paradox? If god made up new data, how could you prove that it wasn't there all along?
Ironically there does not seem to be any data suggesting the quote belongs to Deming!
An attempted editor argues, "Should be B. Joiner. This is a misattribution that got multiplied because of the mistaken reference in Hastie, Tibsharni and Freedman. They never corrected it. The correct reference is Joiner, B.L. (1985). The key role of statisticians in the transformation of North American industry.American Statistician 39(3): 233–234."
It looks like it goes back earlier than either Deming or Joiner. See https://quoteinvestigator.com/2017/12/29/god-data/.
Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital.
Prediction is very difficult, especially about the future.
-- Niels Bohr
All generalizations are false, including this one.
If you torture the data enough, nature will always confess.
--Ronald Coase (quoted from Coase, R. H. 1982. How should economists chose? American Enterprise Institute, Washington, D. C.). I think most who hear this quote misunderstand its profound message against data dredging.