STAT 2040 Lecture Notes - Lecture 13: Confidence Interval, John Tukey, Bonferroni Correction
20 Apr 2018
School
Department
Course
Professor
Document Summary
Chapter 14: one-way analysis of variance (anova: 14. 1: introduction. If this is the case, how to proceed depends on the outcome of the anova f test. These methods control the family-wise significance level or confidence level in different ways, and they each have their pros and cons. Fisher"s least significant difference (lsd) method: the bonferroni correction, tukey"s honest significant different (hsd) method. We typically proceed with the lsd multiple comparison procedure only if we have already rejected the null hypothesis in one-way anova. If the number of groups is not very large, then the lsd method is reasonable. First, the estimate of the within-group variance is based on all groups in the study, and not just the two groups under consideration. Anova: this method controls the family-wise significance level by controlling the probability of finding the largest difference among sample means statistically significant.
