Hello,
We completed a study where schizophrenia/schizoaffective patients (n=10/10)
and controls (n=15) carried out a cognitive paradigm. We would normally
have 2 groups, "patient" and "control" in the design.
However, we may be interested in asking e.g. whether schizophrenia patients
show differences from schizoaffectives or controls. Creating 3 groups may
not be advisable since we have no a priori reason to think that
schizophrenia and schizoaffective patient groups would have different
variances associated with them and these would normally be analyzed in a
single group.
What would be the best way to approach this issue, maintaining flexibility
to look at subgroups while also analyzing main groups of patient vs.
control?
Thank you,
Dost
|