Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.


David
Posts:
4
Registered:
6/15/12


degrees of freedom for repeated measures ANOVA
Posted:
Jun 15, 2012 8:42 AM


Hi,
I am trying to work out manually the degrees of freedom for my experiment. I have a balanced mixed factorial experiment. I have treatment with two levels (A and B), time with two levels (1 and 2) and Disease subgroup with three levels (Ctl, G1 and G2). Each patient has been studied in the two time points but has only been given one treatment, so if I am not wrong, Patient is nested in Treatment and Disease Subgroup. I have 8 replications at each level of the factors, so in total I have 96 observations. I am interested in two interactions: Treat*Time and Treat*Disease. I want to do this manually because a specific program I am using within my field complains that I don't have enough degrees of freedom to do contrasts on my interaction terms. I hope I am explaining myself since I am biologist with some experience in statistics
So to build my ANOVA table, I have
Treatment 1df Time 1df Disease 2df Treat*Time 1df Treat*Diseas 2df
But I am not sure how to follow. How do I account for Patients and how to calculate my error? I imagine my total is 96 1
Thanks for your help
Dave



