PSY202H1 Lecture Notes - Lecture 2: Repeated Measures Design, Type I And Type Ii Errors, John Tukey
Document Summary
When the effect is significant: we will run a post hoc test to see where the significant difference exists. (which pair of means are significantly different) The main test is called the omnibus test. (this will tell us if there is main effect) The post hoc tests (or post-tests) is the follow up test that tell us where the effect exists. This is only conducted when there is a main effect exist. Use this value to compare each pair of groups mean. If the difference between means exceed the hsd then conclude there is a significant difference. If the difference between means not exceed the hsd then conclude that this is not a significant difference. (we only use the tukey"s hsd when the n is the same for each group) When you did an omnibus anova (the main one) and it is significant. You are sure that the biggest mean difference will be significant.