I have two group of animals (transgenic (tg) and nontransgenic (wt)) tested before and after a treatment (A and B type). I have grouped the animals as TG-A, TG-B, WT-A, WT-B.
I have genotype, treatment and time effects since measurements performed before and after the treatment. My research question is whether the mean of each group changes after the operation and whether there is any difference between the groups after the operation.
I have created following fixed effects: genotype, treatment, time, treatment * genotype, treatment *time, and treatment*time* genotype. In jump I have loaded my data as follows:
ID #, Group (TG-A, TG-B, WT-A, WT-B), Genotype, Operation, Time, output.
Can I construct my model by writing: group, time, group x time or should I write it as genotype, treatment, time, treatment * genotype, treatment *time, and treatment*time* genotype.
The second one is hard to interpret: if genotype x operation is significant interaction which groups should I just perform one way or two way ANOVA for groups before and after the treatment. If genotype x time is significant then should I only compare each group with its postoperation group such as preoperation TG-A vs postoperation TG-A.
Thanks
Data format is important and the analysis setup is a little complex too.
I go back to this example when I'm working this type of problem.
ANALYZING REPEATED- MEASURES DATA
It looks like the example was done on a mid-90's vintage mac, despite the differences in the way the interface looks, the setup still works the same way, and the example table, dogs.jmp, is still in the sample data.
@Duane_Hayes , the author. Thanks for this great
