Factorial ANOVA Practice
For the Factorial ANOVA Practice, I have two questions:
1.) I am trying to figure out which post hoc comparison to run. In the chart on page 133 of our book, it asks for the type of comparison... complex or pairwise. What is the difference again? From here, if we are doing a complex and decide we should use a modified Fs, how do we select this with SPSS? I know it means modified Scheffe, but SPSS does not list that option. If we were to go with Pairwise and choose Tukey-Kramer, I do not see this option either in SPSS. What do you recommend?
2.) If we have unequal cell sizes, is there a way to specify to SPSS to use Type I or Type II Sum of Squares?
Thank you for your help!
Brooke
posted by bfullmer @ 10:53 PM
2 Comments:
At 11:10 PM,
Mari Clements said…
All of the automated post hocs in spss are pairwise. Complex comparisons would include contrasting groups 1 and 2 with 3, for instance, and you cannot do this in an automated spss analysis.
Neither Tukey nor Scheffe, as implemented by spss, are adjusted for unequal cell size, so alpha levels are not guaranteed (as the printout would tell you if you used them).
Adjusting the sums of squares is done under the model menu in spss.
As a general note, I would strongly encourage you to try out some of these things and see what you get/how they work. Your learning will be both deeper and more long lasting if you do some of the exploration of the analyses and the ideas. Nothing wrong with asking questions, of course, but you won't be truly comfortable with the analyses until you have seen how they work for yourself.
At 5:50 PM,
Anonymous said…
I thought according to the assignment we were specifically told to you LSD and Bonferroni?
Post a Comment
<< Home