George E. P. Box (1919-2013)
![GeorgeEPBox](https://upload.wikimedia.org/wikipedia/commons/thumb/a/a2/GeorgeEPBox.jpg/256px-GeorgeEPBox.jpg)
George Box was born in Gravesend near London, England, where his father worked at a tailor shop and barely earned enough to support the family. As a young man, he studied chemistry until World War II began and he was drafted. Because of his studies, he was sent to work in the chemical defense experiment station where he researched antidotes for poison gases. He noticed that when he tested antidotes on a small spot on people's arms, there was a correlation between the response in the two arms of the same person. To be able to study responses more effectively, he used Fisher's book on the design of experiments to carry out and analyze appropriate experiments.
After the war, he continued learning about statistics, studying at London University, and then began to work at Imperial Chemical Industries (ICI). At the invitation of Gertrude Cox (1900-1978), he came to the United States on a one-year leave from ICI. He returned to the United States in 1956 to take over the Statistical Techniques Research Group at Princeton. He later created the Department of Statistics at the University of Wisconsin, Madison.
While at ICI he worked with people from the statistical
group at University College where he met David Cox. Their names
went so well together that they determined to someday write a
joint paper. They eventually did and published it in the
Journal of the Royal Statistical Society (1964) about
what are now known as Box-Cox transformations. These allow a
researcher to transform skewed data into an approximately
normal distribution so that assumptions for tests are met. In
addition, Box came up with the definition of robust
meaning that a statistical test yields the correct results even
when assumptions are not met, and he used this definition to
test the robustness of existing statistical tests. Before his
time, it was common practice to throw out data that appeared
wrong or that did not meet assumptions, so his work allowed
statistical inference to be performed, even on less-than-ideal
data.
![Nci-vol-8182-300 david cox](https://upload.wikimedia.org/wikipedia/commons/thumb/d/d8/Nci-vol-8182-300_david_cox.jpg/256px-Nci-vol-8182-300_david_cox.jpg)
He wrote an autobiography titled An Accidental
Statistician: The Life and Memories of George E. P. Box
which was published in the year he passed away. After reading
only a few pages, one can start to imagine his personality and
the way he viewed the world. He wrote, Good science is a form
of wit, of seeing the joke that nature is playing on us
(Box, 2013, preface).