Abstract
This paper explains how the discovery of a pocket notebook brings to light P G Tait's surprising involvement in statistics. Tait (1831–1901) was Professor of Mathematics at the Queen's College, Belfast and later of Natural Philosophy at the University of Edinburgh. He was a Fellow of the Royal Society of Edinburgh and a former Fellow of Peterhouse, Cambridge (senior wrangler and first Smith's prizeman in 1852).
Notes
2 Pre-1967, midwifery at St Andrews was taught on the Dundee site. In 1897 University College, Dundee (founded in 1881) became a constituent college of the University of St Andrews. In 1954 University College was renamed Queen's College; and in 1967 it gained independence from St Andrews and was renamed the University of Dundee.
3 The lectures were reproduced in the British Medical Journal in March 1883.
4 Uncited sources of biographical information on Duncan: Royal Society of Edinburgh Citation(2006), BMJ Citation(1890).
5 The coefficient of determination (the statistic) is 0.9944, meaning that 99
of the total variation in fecundity is explained by age. R produces the following linear regression model:
(3 s.f.). Tait has
. Tait's values for the
-intercept and slope fall within the 95
confidence interval R generates for the parameters, that is, a range of values for
and
such that 95
of the data is explained by the model
.
6 Tait uses for both the current age of a woman and her age at marriage. To avoid confusion, I have chosen to use
to denote age at marriage.