Preliminary information suggests that the effect of (face-to-face) counseling is larger for migraine sufferers who are women than men. Knowing this, the scientist would probably use a: A) completely randomized design. B) block design, with counseling and no counseling as the blocks. C) block design, with women and men as the blocks. D) stratified random sample.