Salsburg, D. 2001.
``
The Lady Tasting Tea
-- How statistics revolutionized science
in the twentieth century''.
Owl Book.
Chapter 15. A Worm's-Eye View
Working for K.P.
David called Karl Pearson a wonderful lecturer.
``He lectured so well, you would sit there and let it all
soak in.'' He was also tolerant of interruptions from
students, even if one of them spotted a mistake, which
he would quickly correct. Fisher's lectures, one the other
hand, ``were awful. I couldn't understand anything.
I wanted to ask him a question, but if I asked him a
question, he wouldn't answer it because I was a female.''
So she'd sit next to one of the male students from America
and push his arm, saying, ``Ask him! Ask him!''
``After Fisher's lecture, I would go spend about three
hours in the library trying to understand what he was
up to.''
Looking back at those days, she noted,
``I'm inclined to think that I was brought in to keep
Mr. Neyman quiet. But it was a tumultuous time because
Fisher was upstairs raising hell and there was Neyman on
one side and K.P. on the other and Gosset coming every
other week.'' Her reminiscences of those years are much
too modest. She was far more than a worm ``brought in to
keep Mr.Neyman quiet.'' Her published papers (including
one very important one cowritten by Neyman on a
generalization of a seminal theorem by A.A. Markov, a
Russian mathematician of the early twentieth century)
advanced the practice and theory of statistics in many
fields. I can pull books off my shelf from almost every
branch of statistical theory and find references to
papers by F.N. David in all of them.
井田君がもちこんできた MacOS X における，
A0 サイズ PDF ファイル出力問題にとりくむ．
数時間試行錯誤してわかったことは
……
MacOS X はシステムに PDF 出力機構を内包する
(なにしろ「画面 PDF」だし)
わけだが，
その性能を左右するのはプリンタードライヴァーなどの
ドライヴァー類である．
ドライヴァーを変えると出力できる用紙サイズがかわってくる，
と
……
なんともはやですな．
ついでにいうと，
ネットワーク設定がややこしい．
内蔵モデムか LAN か，
という選択がややこしいとこに隠されており，
default ではモデム優先にされてる．
MacOS X ばてで
2330 研究室発．
2340 帰宅．
[今日の素読]
Salsburg, D. 2001.
``
The Lady Tasting Tea
-- How statistics revolutionized science
in the twentieth century''.
Owl Book.
Chapter 15. A Worm's-Eye View
War work
When war broke out in 1939, David worked at the Ministry
of Home Security, trying to anticipate the effects of
bombs that might be dropped on population centers like
London. Estimates of the number of casualities, the
effects of bombs on electricity, water, and sewage
systems, and other potential problems were determined
from statistical models she built. As a result, the
British were prepared for the German blitz of London
in 1940 and 1941 and were able to maintaion essential
services while saving livies.
Salsburg, D. 2001.
``
The Lady Tasting Tea
-- How statistics revolutionized science
in the twentieth century''.
Owl Book.
Chapter 16. Doing Away With Parameters
During the 1940s, Frank Wilcoxon, a chemist at American
Cyanamid, was bothered by a statistical problem. He had
been running hypothesis tests comparing the effects of
different treatments, using ``Student'''s t-tests and
Fisher's analysis of variance. This was the standard way
of analyzing experimental data at that time. The
statistical revolution had completely taken over the
scientific laboratory, and books of tables for
interpreting these hypothesis tests were on every
scientist's shelves. But Wilcoxon was concerned about
what often appeared to be failure of these methods.
Oh, but this was foolish! Why should a chemist like
Wilcoxon have to work out these simple but tedious
calculations? Surely, somebody in statistics had done
this already! Back he went to the statistical literature
to find this previous paper. He found no such paper.
Mainly to check on his own mathematics, he submitted a
paper to the journal
Bometrics
(not to be confused with Pearson's
Bometrika).
He still believed that this could not be original work
and he depended upon the references to konw where this
had been published before and reject his paper.
By rejeccting it, they would also notify him about these
other references. However, as far as the refrees and
editors could determine, this was original work. No one
had ever thought of this before, and his paper was
published in 1945.
