1 Oscar Sheynin History of Statistics Berlin, 2012 ISBN 978-3 ...

sheynin.de
  • No tags were found...

1 Oscar Sheynin History of Statistics Berlin, 2012 ISBN 978-3 ...

8. PoissonPoisson introduced the concepts of random variable and distributionfunction. He contributed to limit theorems and brought into use the lawof large numbers proving it for the case of Poisson trials, studiedcriminal statistics and systematically determined the significance ofempirical discrepancies which proved essential for the development ofstatistics and stressed the difference between subjective and objectiveprobabilities.Key words: law of large numbers, criminal statistics, medicalstatistics, null hypothesis8.1. Probability and Statistics8.1.1. General Statements. Libri Carruci et al (1834, p. 535)advocated the need for basing statistics on the theory of probability:The most sublime problems of the arithmétique sociale can only beresolved with the help of the theory of probability.For their part, Double et al (1835, p. 174) noted its connection withtreating mass phenomena:Statistics carried into effect always is, after all, the functioningmechanism of the calculus of probability, necessarily concerning infinite [?]masses, an unrestricted number of facts; and (p. 176) [with respect to theapplicability of mathematics] the state of the medical sciences is not worsethan, not different from the situation with all the physical and naturalsciences, jurisprudence, moral and political sciences etc.Social arithmetic denoted demography, medical statistics and actuarialscience. In both cases Poisson was co-author.He (1837a, pp. 1 and 36) remarked that in the 18 th century,probability became a main branch of mathematics with respect both tothe number and utility of its applications and the kind of analysiswhich it engendered. And, too categorically: No other part ofmathematics is capable of more or more immediately usefulapplications.On pp. 35 – 36 Poisson defined the aims of the theory of probabilityas determining the cause for believing that a thing is true, and forcomparing those causes for widely differing problems. He added thatThese principles should be regarded as a necessary supplement tologic. The connection of probability with logic originated with Leibnizand Lambert and was developed in the 19 th century by Boole, Jevonsand Venn.Poisson (1837a) consistently demanded to check the significance ofempirical discrepancies, for example between results of different seriesof observations; along with Bienaymé, he was therefore the Godfatherof the Continental direction of statistics (§ 14.1). True, his approachwas restricted (§ 8.5).73

More magazines by this user
Similar magazines