Post Hoc
I am trying to decide on my post hoc and have a couple questions. I am using the chart on p. 133 of our ANOVA text book. For my type of comparison, what factors lead me to decide between a complex and pairwise comparison? Also, does "equal n" mean exactly equal? If I decide to do a pairwise comparison and don't have exactly equal "n" or homogeneity of variance, it tells me to use Dunnet C when the error df is 75 or more. When I try to do so on SPSS using Analyze, General Linear Model, Univariate, Post Hocs, it will not let me select Dunnet C... am I doing something incorrectly?
posted by bfullmer @ 11:15 PM
1 Comments:
At 9:01 AM,
David said…
I'm not sure why SPSS won't let you run certain post-hocs. You guys didn't learn about post-hocs extensively in the class, and usually there is more than one appropriate post-hoc you can use.
Pick the one that makes the most sense to you from what you know and you should be OK.
Post a Comment
<< Home