Samuelsen, Bendik ”Contextual priming and attitude change processes: Advertising context, elaboration, and attitude strength”, Dissertation.
Abstract:
The advertisements consumers are bombarded with every day normally occur in a context. Editorial stories in print media, TV shows, and other advertisements or commercials are examples of such contexts. The context in which advertisements are embedded can vary in congruity with the content of the advertising messages. This dissertation first asked if attitudes formed under different levels of congruence between content of advertising context and the content of the advertisement arise from different levels of elaboration. If level of congruence gave different levels of elaboration, the second question was: do attitudes formed under different levels of congruence between content of advertising context and the content of the advertisement differ in attitude strength? To guide the search in answers for these theoretical questions, this dissertation focused on an integration of the literatures of on the concept of priming and the theory and methods from research on attitude change.
The basic hypothesis derived from the multiple roles of persuasion variables postulate of the ELM (Petty and Cacioppo, 1986) was that congruence between editorial context and advertisement content could lead to increased elaboration. In conditions with higher congruence, primed memory content would be more applicable to advertising content, thus increasing elaboration likelihood compared to conditions with lower congruence. The following hypotheses were suggested: at different levels of congruence we expected: a) brand attitude extremity should differ, qualified by sensitivity to argument quality, b) content of cognitive responses would differ, qualified by sensitivity to argument quality, c) correlation between cognitive responses and brand attitude would differ, and d) correlation between brand attitude and purchase intention would differ. These hypotheses were tested in a 2 (prime congruence: congruent vs. incongruent) ´ 2 (argument quality: strong vs. weak) between subjects factorial design (n = 137). Congruence was manipulated by crossing a functionally positioned advertisement with a prime story with either a functionally or experientially content. Participants were undergraduate students of Ohio State University.
Results showed that congruently primed participants were significantly more sensitive to the argument quality manipulation (F(1,133) = 9.8, p = .002), that correlations between the cognitive response index and brand attitudes were higher for congruently primed participants (Fisher z = 2.48, p < .01), and that correlations between brand attitudes and purchase intentions were higher for congruently primed participants (Fisher z = 1.81, p < .04). The predicted interaction between prime congruence and argument quality on the cognitive response index failed to reach significance.
In response to the second question, the next assumption based on the ELM (Petty and Cacioppo, 1986; Petty, Haugtvedt, and Smith, 1995) was that if congruently primed attitudinal responses resulted from higher message scrutiny than incongruently primed attitudinal responses, congruently primed attitudes should be stronger. Consequently, it was hypothesized that both brand attitude, and confidence in that brand attitude would resist counter attack messages better in congruent compared to incongruent prime conditions.
The second study kept and crossed the strong argument quality version of the advertisement with the same editorial stories as in study 1. Additionally, a counter argument message in two versions: strong and weak was developed. Brand attitudes and brand attitude confidence were measured pre and post attack, thus the design of study 2 was a 2 (counter argument: strong vs. weak) ´ 2 (prime congruence: congruent, incongruent) ´ 2 (time of attitude measurement: post advertisement, post counterattack) mixed factorial design with the first two treatments as between subjects factors, and the last a repeated measurement factor within subjects. Participants were 181 undergraduate students of Ohio State University. First, the results showed that the hypothesized prime congruence ´ time of measurement interaction on attitude resistance was significant (F(1,177) = 8.73, p < .005). Attitudes of congruently primed participants resisted counter attacks better than attitudes of incongruently primed participants even though the brand attitudes in the two conditions were not significantly different pre attack. Secondly, the hypothesized prime congruence ´ time of measurement interaction on attitude confidence resistance was significant (F(1,177) = 9.26, p < .005). The confidence attached to the attitudinal response resisted counter attacks better for participants in the congruent prime condition, compared with the incongruent prime condition.
In combination, the results form the two studies indicate that congruence between advertising context and advertising content can increase elaboration likelihood, and consequently also drive attitude resistance and attitude confidence. These questions have not been addressed by previous priming research, or by previous persuasion research. Particularly, priming researchers have normally not utilized manipulations necessary to assess elaboration, and persuasion researchers have not investigated the elaborative capacity of contextual priming. This dissertation contributes to both priming and persuasion research by examining the elaborative capacity of contextual primes in the form of editorial context of print advertisements. It also provides new knowledge about the strength of brand attitudes formed or changed in the presence of different contextual primes.
