This study tested in the context of branded products a set of hypotheses de
rived from theory in political science concerning negative advertising mess
ages: those that unfavorably compare Brand B with Brand A. We expected and
found that the most favorably-rated message contains negative elements, and
that any negative comparison is perceived differently by users of Brand B,
users of A, users of a third brand and users of multiple brands. Two other
expectations were also supported. Responses are more favorable to a messag
e simply denigrating product features of a rival brand vs, a message smeari
ng a corporation. Also, all ratings do not differ in the same way when one
contrasts responses to negative comparative messages with responses to mess
ages that are simply brand comparisons.