All models are wrong

George Box’s 1976 paper named “Science and Statistics” is famous for that “all models are wrong” quote. I’m sure that the number of people who knows this quote is quite larger than the number of people who have read the paper entirely.

The paper provides interesting (and insightful) discussions that still apply to the field of Statistics these days. IMO, one of the most remarkable things Dr. Box pointed out is about “Mathematistry”. Which he defines (or explains) as follows

“Mathematistry is characterized by development of theory for theory’s sake, which since it seldom touches down with practice, has a tendency to redefine the problem rather than solve it.”

Later on, in the same paper, he mentions how this “malady” (in his words) is harmful to statistics as a field.

“An even more serious consequence of mathematistry concerns the training of statisticians. We have recently been passing through a period where nothing very much was expected of the statistician. A great deal of research money was available and one had the curious situation where the highest objective of the teacher of statistics was to produce a student who would be another teacher of statistics. It was thus possible for successive generations of teachers to be produced with no practical knowledge of the subject whatever. Although statistics departments in universities are now commonplace there continues to be a severe shortage of statisticians competent to deal with real problems. But such are needed.”

I think that we still face this challenges (on many others he mentioned in that paper) in Statistics. Therefore, I thought it would be nice to share this to make people think about it.

Lucas Godoy
PhD Candidate / TA /GA

I’m a PhD Candidate in Stats interested in R, Open Data, and the most diverse applications of statistics.

comments powered by Disqus