A A A

In what way a posttest-only design with nonequivalent groups is better than a posttest-only design with no comparison group?