Holm’s sequential Bonferroni post hoc test is a less strict correction for multiple comparisons. The ordinary Bonferroni method is sometimes viewed as too conservative. In other words, it overcorrects for Type I errors. This is due to several reasons, including the fact that Type II error rates are high for each test. The Bonferroni does suffer from a loss of power. For example, if you are running 20 simultaneous tests at α = 0.05, the correction would be 0.0025. This post hoc test sets the significance cut off at α/n. The problem with running many simultaneous tests is that the probability of a significant result increases with each test run. This multiple-comparison post hoc correction is used when you are performing many independent or dependent statistical tests at the same time. Tukey’s Test (see also: Studentized Range Distribution)īonferroni Procedure (Bonferonni Correction).Fisher’s Least Significant Difference (LSD).
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |