I'm trying to analyze the effects of a covariate on my data for a repeated measures anova, and I am having trouble figuring how how to enter the data into statistica. Is this the correct format? Any other advice is appreciated - thank you!
Here is an example of my table. As you can see I have:
What I'm really looking for, is a way to see how:
the Dependent Variable for Factor A: 1, Factor B: 1 is affected by the covariate for Factor A: 1, Factor B: 1
and likewise Dependent Variable for Factor A: 1, Factor B: 2, is affected by covariate for Factor A: 1, Factor B: 2
and so on..
I have found this article helpful when working with repeated measures data: http://www.statsoft.com/support/blog/entryid/223/with-great-power-comes-great-responsibility/
So if you have 5 subjects, perhaps these are people or animimals, locations... then each should have it's own row in the data. then perhaps you took measurements across time for one factor and under different conditions for a second repeated factor. if that is the case, then your data are set up correctly.
the problem is that you have too many factors for the number of experimental units. when you look at the anova output, all the DF values are 0, so no significance tests can be performed. are all of the covariates necissary in the analysis? perhaps you can compute the difference in the largest and smallest one and then just have one covariate in the analysis.
Thank you! I really appreciate the response, and the article is helpful too.
I think I've finally identified the problem I'm having. My research questions is: How does my "covariate" (or 2nd measured value) changes with the factors?, rather than "How does my covariate change for each subject?" By adding these data as a column, the inherent assumption is that I want to understand the relationship between these values and the subjects.
How can I measure the relationship between Dependent Variable 1 and Dependent Variable 2 (my "covariate") and account for this in the analysis?