### BIOGRAPHY 13.3* Egon S. Pearson* (1895 -1980)

Egon Sharpe Pearson was born in London, England, the son of Karl Pearson (Biography 14.1). Egon was educated at Cambridge University and closely followed in his father's footsteps. Early on, he joined his father's department at University College, London. In 1933, when his father resigned, he took over one of the two new positions created as replacements, the other one going to R. A. Fisher (Biography 13.1). In this position, and as editor of *Biometrika*, he contributed importantly to statistics (he himself published some 133 papers). Above all, he is known, along with Jerzy Neyman (Biography 13.2) as the developer of the modem theory of hypothesis testing (as it is found in text Chapter 13). The Neyman-Pearson approach differed considerably from Fisher's and this difference gave rise to a lifelong and bitter controversy. Unlike Fisher who viewed hypothesis testing as a procedure by which a researcher could form an opinion about some population parameter, Neyman-Pearson viewed it as a means by which a *decision maker* operating under uncertainty could *make a clear choice* between two alternatives, while at the same time controlling the chances for error (and minimizing costs associated therewith).

While Fisher hypothesized the value of only one parameter, Neyman-Pearson explicitly formulated two rival hypotheses, H_{0} and H_{A} -- a procedure suggested to them by William S. Gosset (Biography 12.1). This first step contrasts sharply with Fisher's total neglect of an alternative hypothesis. (If, under the Fisher approach, the parameter in question is judged unlikely to be true, what is the truth?). In addition, Neyman-Pearson introduced the formal acceptance/rejection rule employed throughout text Chapter 13, as well as the notion of the two error types (with probabilities a and b), and they investigated the costs of committing these errors. (Fisher, who moved in the world of the research scientist, perhaps understandably had shown no concern about such costs. Indeed, there seems to be a qualitative difference between the cost to, say, a drug manufacturer of a wrong *decision* and the cost to, say, a genetic researcher of a wrong *opinion*.) Finally, Neyman-Pearson introduced the concept of the power of a hypothesis test and noted how the cost of making observations (which depends on sample size) can be traded against the costs of type I or type II errors.

For additional information on Pearson, consult the *Encyclopedia of Statistical Sciences* (New York: Wiley-Interscience, 1982-86), vol. 6, pp. 650-653, or Bartlett, Maurice S. "Egon Sharpe Pearson, 1895-1980." *Biometrika*, April 1981, pp. 1-12. Contains a listing of 133 papers published by E. S. Pearson.