An Information Integration Analysis of How Trust and Expertise Combine to Influence Source Credibility and Persuasion

John C. Mowen, Oklahoma State University
Joshua L. Wien, Oklahoma State University
Shreekant Joag, Mankato State University
[ to cite ]:
John C. Mowen, Joshua L. Wien, and Shreekant Joag (1987) ,"An Information Integration Analysis of How Trust and Expertise Combine to Influence Source Credibility and Persuasion", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 564.

Advances in Consumer Research Volume 14, 1987      Page 564

AN INFORMATION INTEGRATION ANALYSIS OF HOW TRUST AND EXPERTISE COMBINE TO INFLUENCE SOURCE CREDIBILITY AND PERSUASION

John C. Mowen, Oklahoma State University

Joshua L. Wien, Oklahoma State University

Shreekant Joag, Mankato State University

The investigation of the construct of source credibility and its effect on persuasion has been ongoing for well over thirty years. See Sternthal, Phillips and Dholakia (1978) for a review of the field. One of the major problems in the research on source credibility is that the studies have for the most part hopelessly confounded its components of trust and expertise when the construct was manipulated. Thus, in most studies source credibility was manipulated by creating a highly attractive source (e.g., President Eisenhower) versus a highly negative source (eg., a Nazi leader). In such "gross" manipulations of credibility, the variables of expertise, likability, and attractiveness are all manipulated simultaneously. Therefore, it is impossible to parcel out the effects of each.

Recently, Wienar and Mowen (1986) performed a study that manipulated source expertise and trustworthiness independently while holding source attractiveness constant. In the study the measures of the persuasiveness of the messages from an auto mechanic were obtained. (No differences in attractiveness were found across the conditions formed by variations in expertise and trustworthiness.) The results revealed that increasing levels of both expertise and trustworthiness resulted in increased persuasion. Thus, trustworthiness is clearly an important source component and does not belong to the realm of "bubbapsychology," as had been suggested by McGuire (1968).

Given that both expertise and trustworthiness influence the persuasiveness of a source, an important question concerns how these dimensions of credibility are combined to create the construct. Quite different implications would exist for those who seek to persuade others if individuals tend to add trust and expertise to form the perception of credibility than if they multiplied trust and expertise.

In order to begin initially the process of evaluating how consumers may combine information on expertise and trust to form impressions of credibility and persuasiveness, a pilot study was developed to use information integration theory as an analysis technique. As developed by Norman Anderson and his colleagues, information integration theory is a potent tool for identifying the cognitive algebra used by individuals to combine information in order to form impressions. Among other things the approach allows the researcher to identify whether individuals are using multiplicative or additive rules. The approach is implemented by the development of factorial experiments in which information is presented to individuals in a within-subject experimental design. In the present study subjects role played that they encountered a professor in the university cafeteria who in the course of conversation made a strong argument for the state wide adoption of a peak load pricing program for electrical energy usage. In a 3 X 3 within-subjects design, the subjects received nine different descriptions of the professor. Within these descriptions the professor's expertise and trustworthiness were varied. Four different measures of the persuasiveness of the message were obtained. The analysis of variance was interpreted from an information integration perspective. It was hypothesized that subjects would reveal the use of a multiplicative model as demonstrated by a significant interaction between the two independent variables. In addition, it was expected that a plot of the results would reveal the typical "fan" shape found when multiplicative models are employed by respondents.

The results revealed a strong effect for the expertise variable (F=30.5, p<.01, df=2,40). As expertise levels increased, persuasion levels increased. No overall group effect was found for the manipulation of trust. Furthermore, no interaction occurred between trust and expertise, which would have been indicative of a multiplicative model.

The four measures of persuasion were treated as though replicated experiments had been performed allowing the investigators to analyze the data at an individual subjects level. (This analysis assumes that each measure is assessing validly the same construct.) The individual analysis of the data revealed that 13 of the 21 subjects showed a significant (p<.05) main effect for trust. Thus, different subjects appeared to be reacting in divergent ways to the manipulation of the external reasons for making the message. In addition, 10 of the 21 subjects revealed a trust by expertise interaction. (Of the ten subjects showing a trust by expertise interaction, eight also revealed a main effect for trust.) These individuals exhibiting the interaction may have been using a multiplicative model.

The study suggests that the question of how individuals integrate information on source expertise and trust should continue to be pursued. The question, however, now seems more complex than as originally proposed. Rather that one of simply identifying whether a multiplicative, additive, or other model is used, it now becomes one of also identifying the types of people who tend to use more complex versus more simplified integration models. Furthermore, one can also expect to find that the type of situation and perhaps the type of endorser may also influence the level of cognitive processing.

REFERENCES

McGuire, W. (1968), "The Nature of Attitudes and Attitude Change," in G. Lindzey and E. Aronson (eds.) Handbook of Social Psychology, Reading, Mass: Addison Wesley.

Sternthal, B., L. Phillips, and E. Dholakia (1978), "The Persuasive Effects of Source Credibility: A Situational Analysis," Public Opinion Quarterly, 285-314.

J. Wienar and J. Mowen (1986), "Source Credibility: On the Independent Effects of Trust and Expertise," Advances in Consumer Research, 13, Richard Lutz (ed.), 306-310.

----------------------------------------