When comparing two samples of scores we try to avoid ______.
conflating evidence
dependent samples
confounding variables
independent populations