284: Sampling the Reference Set.
Year: 1961
Type: article
Abstract: proportionality in a two by two table, have frequently been cited to show that such a method of ' 'repeated sampling from the same population is erroneous, and irrelevant to the test of significance which it is proposed to verify. The recognition of the appropriate reference set is an essential first step to understanding a test of significance, and, therefore, to setting up an appropriate process of possible verification. We may exemplify such a process using the long-disputed test of significance for the difference between the means of two hypothetical populations, the variances of which are not in any known ratio, when a random sample of each is available. 2. Inferences from two normal samples The means of the two samples will be represented by xx and x2, where (n1+l)x1 = S(x1) (n2+l)%2 = S(x2) the estimated variances of the means by s and s2, where
Author: Sir Ronald A. Fisher
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2479174190
