Experimental Design

We replaced 3 subjects in our study. Two dropped out the day after
they started the study and gave us only baseline descriptive data.
They did not use any of the assessment methods.

The third subject was replaced after using 2 methods because we
found out he was not eligible. He was lost to follow-up after that
point. My question is, since he was replaced, should we remove him
from the analyses? We should have 18 subjects in our Latin Square
design (not 19). Alternatively, we could leave him in since we have
data on him.

Any ideas.

.