When selecting a comparison group in a quasi-experimental design, one should:
a. select a group unlike the experimental group:
b. select a group as similar as possible to the experimental group
c. use random assignment.
d. avoid the use of switching replications.