Biology 2244A/B Chapter Notes -F-Distribution, F-Test, Null Hypothesis
11 Dec 2013
School
Department
Course
Professor
Document Summary
What is it: method for testing equality of three or more populations by studying sample variances, better than using two samples at a time because doing this multiple times will reduce the confidence level (cl^trials) Properties: f distribution not symmetric; right skewed, values for f greater or equal to 0, there is different f distribution for each pair of df for numerator and denominator. One-way distribution: because sample data separated into groups according to one characteristic, ex. Each treatment group is separated based on treatment. Things to consider: small p-value will reject null hypothesis, know sum of squares and mean square and their role in f statistic. Method: one-way analysis of variance: also called treatment (not factor) Factor is the explanatory variable being studied. Variation: two ways to estimate common value of variance. Variance between samples (variance b/c of treatment) Estimate of common population variance based on variation among sample means. Variance within samples (variance b/c of error)
