Dear Stats Folk,

I have a one Control group and one patient group. I have measurements on Controls at one time point, while I have a single measurement on the patient group at 3 different time points. However, not all patients have measurements at all the three time points.

I think a Dunnett's test would be ideal for this data set where I want to test for the difference in means of the Controls and the Patients (at three different time points).

However, I understand that the Patient data should not be correlated, as is the case above.

I wonder if there is a correction that I can apply in this case OR perhaps a simple t-test for each session between be the Controls and Patients is better suited.

Any input/leads would be highly welcome.

Thank you in advance for your time,

Jim.