On Mon, 27 Mar 2006 12:30:01 -0800, Horton, Joseph J. wrote:
>I can come close:
>"Since all models are wrong the scientist must be alert to what 
>is importantly wrong." (p. 792)
>Box, G. E. P. (1976). Science and Statistics. Journal of the 
>American Statistical Association, 71, 791-799.

Although this is close, a quick search of several databases
shows that the most likely source is:

Box, G.E.P., Robustness in the strategy of scientific model 
building, in Robustness in Statistics, R.L. Launer and G.N. 
Wilkinson, Editors. 1979, Academic Press: New York.

I found the above reference in a discussion on this website:
http://www.anecdote.com.au/archives/2006/01/all_models_are.html

I can't find an electronic version of this paper but another
search found the quote and the Box (1979) ref  in the
following article which is available online if one has access
to Sage journals:

Weakliem, D.L. (2004).  Introduction to the Special Issue on 
Model Selection. Sociological Methods & Research, vol. 33, 
no. 2, pp. 167-187, November 2004  
 
For people involved in stat/math model development, the
entire issue is pretty interesting.  One citation in the Weakliem
paper that caught my eye is:

Sala-i-Martin, Xavier X. 1997. "I Just Ran Two Million 
Regressions." American Economic Review (Papers and 
Proceedings) 87:178-83.

I think I just found a new definition of "overkill".

-Mike Palij
New York University
[EMAIL PROTECTED]


>I can send a copy of the paper if you like.
>
>Joe
>
>Joseph J. Horton Ph. D.
>In God we trust. All others must bring data.
-----Original Message-----
From: David Epstein [mailto:[EMAIL PROTECTED] 
Sent: Monday, March 27, 2006 3:12 PM
To: Teaching in the Psychological Sciences
Subject: in search of "All models are wrong..." quotation[Spam score:
8%][Scanned]

Anyone happen to have this book?

   _Robustness in Statistics_, R.L. Launer and G.N. Wilkinson,
   Editors. New York: Academic Press, 1979.

It contains a chapter by George Box called "Robustness in the strategy
of scientific model building."  And the chapter is the source of a
well-known quotation, variously rendered as:

"All models are wrong, some are useful."  (Ugh.  Comma splice.)
"All models are wrong, but some are useful."
"All models are wrong; some models are useful."

What's the correct version?

thanks,
David Epstein



---
You are currently subscribed to tips as: [email protected]
To unsubscribe send a blank email to [EMAIL PROTECTED]

Reply via email to