On humor and humor in advertising

Abstract

The generalizability of previous research on humor in advertising is challenged because of conceptual and operational problems involving HUMOR and HUMOROUS MESSAGES. To remedy this, the author proposes (1) a multidimensional model of humor response, (2) an elaboration likelihood model of humor's communication effects, and (3) a semiotic taxonomy of humorous message forms. An application of this taxonomy to previous research and to primetime advertising demonstrates that the type of humorous message that is generally studied by researchers (message-dominant) differs significantly from the type of humorous message typically used by television advertisers (humor-dominant). An experiment was designed (1) to study these humor-dominant message forms, (2) to determine the incidence of general humor effects (G-effects), effects related to particular humor types (HT-effects) and effects related to underlying humor dimensions (D-effects), and (3) to demonstrate the value of the humorous message taxonomy. Each of 182 subjects viewed ten ads in an incomplete but balanced factorial design. The design included five humor-dominant message forms (HTs) and a nonhumorous control (NHC). Five examples of each message type were included (one from each of five product categories). Thirty 30-second ads were used to operationalize the 6 x 5 design. In all, 1424 observations were analyzed. Dependent measures included attention, comprehension, attitude to the source, attitude to the ad, attitude to the product, and attitude toward product purchase. ANOVA and ANCOVA results suggest (1) that humor significantly affects all measures of communication effectiveness, (2) that humor effects are generally but not universally positive, (3) that the five HTs exhibit five distinct patterns of communication effects, (4) that they evidence a combination of G- and HT-effects, and (5) that results based on an analysis of humor-dominant ads differ in several ways from research that was based on message-dominant ads. These findings suggest that it is very important for researchers and practitioners to distinguish among humorous message types. Apparently, different humor types are more likely to enhance (or diminish) specific effects. If an advertiser desires to make one communication effect paramount, he should use the humorous message type that is most likely to produce that effect. In short, specific advertising objectives should dictate the kind of humor employed. This research provides some initial understanding of the differential benefits (and risks) associated with the five humorous message types most frequently used on television.