The Estimation of Sample Size in Experiments

Publisher Website
Google Scholar
Abstract
In order to estimate the size of samples necessary for assertion of statistical significance, the investigator must have information as to what contemplated result is considered important, and possess a certain amt. of preliminary data. Illustrations are presented using preliminary data in which comparisons of arithmetic avgs. are studied, for both independent and correlated samples. By the use of formulas based upon the t-curve and normal curve probability values, and upon the use of a table from the statistical literature, methods are presented for estimating the size of samples necessary for a difference in avgs. to be found at different significance levels and powers of the test.

This publication has 2 references indexed in Scilit: