Statistical Sciences 2244A/B Chapter Notes - Chapter 24: Null Hypothesis, Sampling Distribution, F-Distribution
18 views4 pages
25 Dec 2016
School
Department
Professor
Document Summary
When comparing multiple means we can use a t-test several times but this will result in multiple p-values that will not tell us how spread apart the samples are. Multiple comparison tests have to steps: overall test to see if there is evidence of difference among parameters, follow-up analysis to find which parameters differ and by how much. Analysis of variance compares variation due to specific sources with variation among individuals who are expected to be similar. Anova tests whether several populations have same mean by comparing how far apart sample means are with variance within the samples. Two-way anova is organizing data to represent two factors. Analysis of variance f statistic for testing equality of several means uses formula: F= (variation among the sample means/variation among individuals in the same sample) The numerator is a variance calculated from the sample means. Denominator is an average of the sample variances. Test statistic=n(variance of group means/mean of group variances)