Special Session Summary New Directions in Research on Self-Referencing

Douglas M. Stayman, Cornell University
H. Rao Unnava, Ohio State University
[ to cite ]:
Douglas M. Stayman and H. Rao Unnava (1997) ,"Special Session Summary New Directions in Research on Self-Referencing", in NA - Advances in Consumer Research Volume 24, eds. Merrie Brucks and Deborah J. MacInnis, Provo, UT : Association for Consumer Research, Pages: 73-74.

Advances in Consumer Research Volume 24, 1997     Pages 73-74



Douglas M. Stayman, Cornell University

H. Rao Unnava, Ohio State University

Research has suggested that the extent of self-referencing done by viewers of persuasive messages can influence the persuasion process and thus the impact of messages. However, different explanations have been proposed and different persuasion results found. According to the elaboration explanation, self-referencing affects the elaboration of message information which causes messages with strong arguments to be more persuasive. The affect transfer explanation, on the other hand, argues that when subjects access positive experiences from memory, the affect that is linked to them is transferred to the product whose ad is being processed. Positive attitudes resulting from self-referencing are attributed to this affect transfer mechanism. The purpose of this special session was to bring together self-referencing researchers working from different perspectives to gain insight on how to reconcile and expand on the varying explanations and results in the literature.

In the first paper, "Self-referencing, Persuasion and the Role of Affect," Lien and Stayman used affect as a moderator to explore the influence of self-referencing on message elaboration and the role of affect in influencing persuasion outcomes. Since positive affect, like self-referencing, has been found to induce a large positive set of associations, Lien and Stayman argued that the effects of affect and self-referencing may be interdependent. They proposed that the interdependence may account for effects in which self-referencing could lead to increased or decreased elaboration, since the influence of self-referencing may depend on the level of affect.

Lien and Stayman discussed five studies. The first three studies demonstrated that affect and self-referencing could be manipulated and measured independently and thus their interdependence could be studied experimentally. The fourth and fifth studies manipulated both affect and self-referencing to assess their interdependence as well as manipulating argument strength to assess the influence of the self-referencing/affect interdependence on message elaboration and persuasion. The results presented suggest that affect and self-referencing do interact in their influence on persuasion as expected from an elaboration framework, with increased elaboration due to self-referencing in neutral, but not positive affect conditions.

The second paper, "How Self-Referencing Affects Information Processing: Attention vs., Elaboration," by Unnava, Burnkrant and Jewell contrasted the elaboration explanation and the affect transfer explanations of the self-referencing effect. The authors argued that if elaboration is NOT the cause of the self-referencing effect, then affecting the extent of elaboration that a subject can perform on an advertiing message should not have any effect on the self-referencing effect on attitudes. Subjects were assigned to one of three conditions of elaboration: a) high elaboration condition in which they had as much time as they needed to read the target ad, b) medium elaboration condition in which exposure to ads was limited at 75 seconds each, and, c) low elaboration condition in which subjects were limited to 75 seconds of reading per ad and were also required to do a concurrent task of writing an unrelated digit sequence on a separate sheet of paper. These three conditions were crossed with self-referencing such that half the subjects in each elaboration condition received either the high or the low self-referencing copy. Self-referencing was manipulated in the traditional way by reminding subjects of their experiences with their calculator and how the target product, another brand of calculator, could solve their problems.

The results were supportive of the elaboration explanation of self-referencing effects. When elaboration was high, the classic self-referencing effect emerged on recall and attitude with high self-referencing subjects showing greater recall and more positive attitudes than low self-referencing subjects. As the elaboration was reduced, the advantage of self-referencing disappeared. Under low levels of elaboration, the self-referencing effect actually reversed with the low self-referencing subjects recalling more than the high self-referencing subjects. This was attributed due to the fact that the high self-referencing copy could not be processed in its entirety due to competing demands of the concurrent task as well as the elaboration demanded by the ad copy.

The third paper, "Variations in Self-Referencing: An Empirical Investigation," by Krishnamurthy and Sujan, investigated self-referencing as a mode of ad processing using the current advances in understanding the self as a complex multi-dimensional knowledge structure. Krishnamurthy and Sujan presented a framework for systematically organizing the variations in self-referencing that may be relevant in a consumer behavior context. More specifically, they investigated whether referencing past versus future selves related to outcome variables like attitude toward the ad and brand, and behavioral intention. Krishnamurthy and Sujan presented results from an initial study in which they found that when consumers describe what makes advertisements personally relevant, thoughts relating the ad to their past are more likely to be about specific episodes and contain contextual detail about the episode, whereas thoughts relating the ad to their future are likely to be more general and lack detail. In their second study Krishnamurthy and Sujan built on this and manipulated the temporal orientation of self-referencing and the amount of contextual information in the ad. They found that, when consumers engage in retrospective (past-based) self-referencing, providing more contextual information in the ad reduces ad effectiveness because consumers tend to contrast the information in the ad with the detail associated with their experiences, whereas providing minimal contextual information encourages consumers to assimilate the ad into their own experiences and thus enhances ad effectiveness. On the other hand, when consumers engage in anticipatory (future-based) self-referencing, Krishnamurthy and Sujan find that contextual information in the ad provides the resources for generating episodes of the future, thus, contrary to the retrospective self-referencing condition, a contextually detailed ad enhances ad effectiveness whereas a contextually impoverished ad reduces ad effectiveness. Their mediation tests indicate that the effects of temporal orientation in self-referencing and contextual detail in the ad on ad effectiveness may be mediated by the extent to which consumers could form brand-related consumption visions.

The discussion leader, Rich Yalch, did triple duty during the session. Due to extenuating circumstances the authors of the second paper could not present. Therefore, Rich ended the session by presenting a distillation of the second paper , providing integrative insights and then leading a wide rnging discussion on self-referencing. Several points in the discussion indicated the need for continuing work on how self-referencing influences persuasion, particularly using process measures or other methods to examine more closely the process(es) underlying the varying effects reported in the session and the literature.