-
作者：Welch, BL
摘要：Three methods of testing whether 2 means differ significantly are discussed, (i) The t test, appropriate when the 2 population variances are equal, may sometimes be misleading if these variances are unequal. (ii) A similar test, based on 2 separate estimates of the population variances, is less liable to bias. (iii) A test proposed from considerations of fiducial probability does not appear to be valid. The comparison of regressions is considered from the same points of view.
-
作者：Wishart, J
摘要：Figures for average growth-rate and its rate of change over the 16-week period of a nutrition exp. on 30 individually-fed bacon pigs were calculated, replacing the usual figures of live-weight gain. Analysis of variance, and covariance on initial weight, demonstrated the significance of a fall in average growth-rate with increasing protein percentage in the ration, and of a sex difference in favor of the gilts in both the average growth-rate and its rate of change.
-
作者：Neyman, J
摘要：A recent misstatement of R. A. Fisher''s regarding Karl Pearson''s knowledge of the standard deviation of a binomial has led the author to examine some early work of Pearson''s of 1895 dealing with a suggested method of fitting a binomial series to a frequency distribution.
-
作者：Nussey, AM
摘要：A piebald family, the 4th to be published in England, is recorded. A typical representative of the family is described and the tree covering 5 generations (3 of which are complete) is analyzed.
-
作者：Yule, GU
摘要：This article contains a summary, taken from notebooks, of the subjects dealt with in Karl Pearson''s lectures at University College, London, during the sessions 1894-96, on the Theory of Statistics. The author was one of 3 students who attended what was, perhaps, the first university course ever given on the modern theory of statistics. The summarising of the lectures is preceded by some general comments and recollections.
-
作者：Lawley, DN
摘要：A quantity Z, which is a generalization of Fisher''s z for multivariate populations, is defined and its sampling distribution is found. It is shown how this provides a suitable statistical test for certain generalized analysis of variance problems, and a numerical example is given. The application of the test to the problem of regression of a set of dependent variates on another set of independent variates is also discussed.
-
作者：Finney, DJ
摘要：The distribution of this ratio has been investigated by Bose (1935) by a method dependent on term-by-term integration of infinite series. The simplicity of his result suggested the possibility of the more direct approach given below, which is followed by the evaluation of the probability integral for the distribution. It is then shown how, by a simple transformation of existing tables, a test of significance may be applied when the population correlation coefficient is known, and how the test ...
-
作者：Haldane, JBS
摘要：By a change of the variable it is possible to make a number of probability distributions approximate very closely to normality. The transformations may be used where the integral of the distribution is not tabulated.
-
作者：Neyman, J; Pearson, ES
摘要：This paper gives some fuller explanation of certain points in Student''s paper which, it was known, he had intended to revise.
-
作者：Pitman, EJG
摘要：Given m batches or blocks, each consisting of n individuals, to which are allocated at random n treatments, a test of the null hypothesis that the treatments are equivalent in their effects may be made without any assumption as to the population form by finding the distribution of W, the ratio of treatment to treatment plus error sums of squares, in the population generated by the chance allocation of the different treatments to the individuals in the batch. The author obtains the first 4 mome...
