In the context of a t-test when using a pretest-posttest research design, µD < 0 would be an example of a possible alternative hypothesis.