Persuasion Theories in Marketing, Advertising, and Consumer Behavior Contexts
* * *
Theory of Reasoned Action
Although perhaps not a strict persuasion theory, the theory of reasoned action is a model of behavioral intentions developed by Fishbein and Ajzen (1975; also see chapter 8 in this volume). The model incorporates both attitudes and subjective norms that people hold in predicting their future behavior. Formally, the theory of reasoned action is:
B ˜ BI = Aact(w1) + SN (w2)
where
B = a particular behavior
BI = intention to engage in the particular behavior
Aact = attitude toward engaging in the behavior
SN = subjective norm pertaining to what others think.
The theory posits that the most proximal input into a behavior is a person’s intention to engage in that behavior. (Although seemingly obvious, this assumption is important because it implies that behavior is intentional.) In turn, behavioral intentions are determined by one’s attitude toward performing the behavior or act (Aact) and one’s beliefs about what important others think about one performing the behavior (SN). The weights for each component (w1, w2) indicate that the relative weights for each component of behavioral intention will vary across people and situations. For example, some behavioral intentions may be overly influenced by subjective norms (wearing a particular brand of clothing), whereas other intentions are more heavily influenced by personal attitudes (choice of chewing gum).
Fishbein and Ajzen further specified that each component of intention, attitudes, and subjective norms were themselves determined by specific beliefs about each. Using an expectancy-value approach, they quantified attitude toward the behavior as a cross-product of the subjective likelihood that performing a particular behavior (b) would lead to a specified outcome (i) and their evaluation of that outcome (e):
where n represents the number of different consequences that come to mind. Similarly, Fishbein and Azjen quantified the subjective norm component as the cross-product of the belief that an important other (j) thinks one should perform a particular behavior (b) and one’s own motivation to comply with that important other (m):
Tests of the Model
Given that marketers are particularly keen on being able to predict the behavior of their customers, the theory of reasoned action was put to the test in a number of consumer situations, and formed the basis of a number of doctoral dissertations (cf. Lutz, 1973b; Ryan, 1975). For example, the Fishbein model (or variations of it) has been shown to be predictive of the purchase of a specific brand of grape drink (Bonfield, 1974), toothpaste (Wilson, Mathews, & Harvey, 1975), generic prescription drugs (Brinberg & Cummings, 1984), football tickets (Lutz, 1973a), model of automobile (Raju, Bhagat, & Sheth, 1975), and even the purchase of term papers (Weddle & Bettman, 1973). In a comprehensive meta-analysis, Sheppard, Hartwick, and Warshaw (1988) found very strong evidence of the predictive validity of both the relation between intentions and behavior and between the combination of attitudes and subjective norms and behavioral intentions. Interestingly, and of particular interest to marketers, the meta-analysis found that the predictive validity of intentions on behavior was substantially stronger when the criterion variable represented a choice among alternatives (r = .77) than when it did not (r = .47).
More recent studies have shown that the Fishbein and Ajzen model is predictive of consumer behavior across cultures. For example, Bagozzi, Wong, Abe, and Bergami (2000) found that the theory of reasoned action predicted fast food restaurant patronage in samples from the U.S., Italy, Japan, and the People’s Republic of China. However, they also found that the effects were generally stronger in the U.S. sample, and that the relative influence of attitudes and subjective norms differed across samples. Specifically, those in Western cultures, which are more individualistic (independent) and emphasize internal aspects of the self in decision-making (e.g., be oneself), showed much stronger correlations between attitudes and intentions than those in Eastern cultures. In contrast, those in Eastern cultures, which are more collectivistic (dependent) and emphasize how personal actions influence the group, showed much stronger correlations between subjective norms and intentions than those in Western cultures.
In general, the theory of reasoned action model is attractive to marketers because it affords them the ability to determine what components of intentions to target. For example, because the theory is quantifiable, it is relatively easy for marketers to determine all of the components of the model through surveys. This allows them to determine which beliefs about the consequences of a purchase decision are salient, how these beliefs are evaluated, which others’ opinions about the behavior are most salient, and how motivated consumers are to comply with what those others think. Just as important, for any particular product or service, regression analyses can reveal the relative weights that the attitudes and subjective norms represent. Consequently, each variable in the entire equation represents a marketing opportunity. Advertisements and collateral marketing can be created to change the belief about the behavior (a Volvo results in greater safety), the evaluation of the belief (how important safety is), what important others (e.g., parents, spouse) would want you to do, and how much you care about what they think.
In sum, the theory of reasoned action has had substantial impact on marketing research, both academic and in marketing practice. It was one of the first models that provided a clear articulation and quantification of the inputs into behavioral intentions, and did so in a way that was easily measurable, and thus easy for marketers to implement. However, one limitation of the model is that it restricts itself to volitional behaviors. Thus, the model has little to say about behaviors performed outside of awareness and specific intention. In fact, the model specifies that thoughts must mediate actions, and thus does not easily allow for the possibility of spontaneous, impulsive types of behaviors or other influences outside of conscious volition (mood, anger, etc., Eagly & Chaiken, 1993; for a spirited debate of these issues, see Fishbein & Middlestadt, 1995, 1997; Haugtvedt & The Consumer Psychology Seminar, 1997; Herr, 1995; Miniard & Barone, 1997; Schwarz, 1997).
To address the possibility that there may be multiple routes to persuasion (and the formation of attitudes) in general (for a review, see Chaiken & Trope, 1999), dual process models of attitude formation and cognitive processing were developed. In the next section, we review the elaboration likelihood model (ELM) of persuasion (Petty & Cacioppo, 1986), which has been the dominant persuasion model in consumer research over the last three decades.
Elaboration Likelihood Model
The ELM is a model of persuasion that proposes two distinct routes to persuasion, the central route and the peripheral route (for a more in-depth discussion of the ELM, see chapter 9 of this volume), which refer to attitude changes that occur through different levels of evaluative processing. In the central route, attitudes are formed through an extensive, effortful process that scrutinizes a message for the quality of its arguments. In many respects, this highly effortful central route to persuasion resembles the highly effortful process of attitude formation described by the theory of reasoned action. In contrast, the peripheral route refers to attitude formation that is based on nonargument cues, such as mood, source attractiveness (when not relevant to the argument quality), and heuristics (e.g., number of arguments, source expertise, message length).
The ELM provides an integrative model that addresses some of the perplexing inconsistencies in earlier attitude research. As Petty and Wegener (1999) note, attitude research in the late 1970s was in a remarkable state of disarray. Commonly accepted variables of attitude change, such as the mood of the receiver and the credibility of the message source, often produced conflicting effects (cf. Kelman & Hovland, 1953; Sternthal, Dholakia, & Leavitt, 1978; Zanna, Kiesler, & Pilkonis, 1970). The ELM was intended to provide a unifying framework that could explain how the classic inputs into persuasion (source, message, recipient, conte
xt) could have different impacts, depending on the particular route to persuasion. Thus, either the central route or the peripheral route can be evoked in various situations involving different message types, individual differences among receivers, and environmental (or situational) factors. Persuasion can be effective in both routes, although the strength, durability, and resistance of attitudes formed via the two routes may differ (Haugtvedt & Kasmer, 2008; Petty & Wegener, 1998).
The underlying mechanism of the ELM is indicated in its name: elaboration likelihood. The model posits that when people have both the motivation and the ability to process the information presented in a persuasive communication, the likelihood of message elaboration is high, and people will take the central route. In contrast, when the likelihood of message elaboration is low as a result of either lack of motivation or ability to process information, people tend to take the peripheral route to persuasion. Which route is taken has a number of important implications. First, it determines which components of a persuasive communication will be the most effective, either central cues (message quality) or peripheral cues (mood, expertise, source attractiveness). This helps explain the rather counterintuitive finding that quality of the message may have little effect on persuasion in some situations, such as when motivation or ability to process the arguments is low, but other seemingly comparatively trivial variables (liking for background music in an ad) may have strong effects.
A second important implication of which route to persuasion is taken pertains to the qualities of the attitudes formed. The two routes may yield attitudes that are of equal valence and extremity. However, other important qualities of the attitudes will differ as a function of the two routes. Attitudes formed through the central route tend to be more highly accessible, held with more confidence, more predictive of behavior, more resistant to change, and persist longer over time, compared to attitudes formed through the peripheral route (Petty & Krosnick, 1995). Attitudes formed through the central route result from active information processing and a well-integrated cognitive structure, whereas attitudes formed through the peripheral route are led by passive acceptance or rejection of simple cues and are weaker, particularly over the long term.
Tests of the Model
Given that the classic persuasion inputs (source, message, recipient, context) that the ELM addresses are all critical components of advertising, the ELM has had substantial influence on persuasion research in marketing and advertising. One of the first studies in consumer behavior to employ the ELM investigated the role of product involvement (Petty, Cacioppo, & Schumann, 1983). Petty et al. manipulated three factors: motivation to process the information in the ad, central cues, and peripheral cues. Motivation was manipulated through product involvement (personally relevant or irrelevant), the central cue was manipulated through argument quality (strong or weak arguments), and the peripheral cue was manipulated via the source (celebrity or noncelebrity endorser). Supporting the ELM, argument quality had a greater effect on attitudes under high than low involvement conditions, whereas the celebrity endorser had a greater effect under low than high involvement conditions.
As noted, even though attitudes formed via both the central and peripheral routes may produce apparently equivalent attitudes (reflected by attitude scores), the two routes produce attitudes that differ on other important qualities. In a series of studies, Haugtvedt and colleagues tested the elaboration-persistence and elaboration-resistance hypotheses, which state that the more extensive elaboration that occurs through the central route produces more persistent attitudes and attitudes more resistant to change (Haugtvedt & Petty, 1992; Haugtvedt, Schumann, Schneier, & Warren, 1994; Haugtvedt & Wegener, 1994). These studies showed that whether degree of elaboration was operationalized via individual differences variables (e.g., need for cognition, Cacioppo & Petty, 1982) or situational manipulations of personal relevance, central route attitudes persisted longer over time and changed less after exposure to an opposing message than did peripheral route attitudes.
One aspect of the ELM that has caused some confusion is precisely what makes a cue central or peripheral. Consider the example of source attractiveness. This variable is typically referred to as a peripheral cue, which it often is. Examples might include automobile and beer ads, which often employ attractive endorsers. Clearly, the attractiveness of the endorser has little relation to the message in these instances. However, that is not always the case. In some instances, endorser attractiveness may be perceived to be very relevant to the message. For example, the attractiveness of endorsers or models may be considered particularly relevant for certain products but not others. Consistent with this notion, research shows that an endorser’s physical attractiveness serves as a central cue for beauty products, such as shampoo (Petty & Cacioppo, 1980) and razors (Kahle & Homer, 1985), yet other aspects of the source, such as their celebrity status (with attractiveness held constant) serve as a peripheral cue for the same product (razors; Petty et al., 1983; see also Kang & Herr, 2006).
Shavitt, Swan, Lowrey, and Wänke (1994) directly examined the relevance hypothesis by testing the proposition that the influence of endorser attractiveness as a peripheral or central cue depends on the message processing goals that receivers have at the time of exposure. Participants were exposed to ads for a fictitious restaurant that was ostensibly supposed to open soon. Processing route was manipulated through personal involvement (to open in a local or distant area), and endorser attractiveness was also varied. The third factor, motive for processing, was manipulated via a priming task intended to make either sensory or image attributes salient. Participants were primed with either a sensory cue, in which they rated 20 sensory experiences (e.g., smelling fresh air, feeling sore muscles) on how good or bad they made them feel), or an image cue in which they rated 20 image events (wearing a Rolex, losing a job) on how much they would make an impression on others. The results from both attitude ratings and cognitive responses showed that under image prime conditions, endorser attractiveness served as a central cue (influenced attitudes under high but not low involvement conditions), but under sensory prime conditions, the opposite effects occurred.
Extensions of the Model
The general notion of dual routes to persuasion has led to advances in other aspects of attitudes and persuasion research that have had a strong input on marketing and advertising research (Haugtvedt & Kasmer, 2008). One example is the notion of metacognitive processing, or “thinking about thinking.” This line of research looks at the extent to which thoughts about thought processes involved in attitude formation affect attitudes, particularly for attitude qualities such as attitude confidence, resistance, and certainty (for a review, see Petty, 2006). For example, attitude certainty tends to increase when people perceive themselves as resisting a persuasive communication (Tormala & Petty, 2002). In a follow-up to that set of studies, Tormala and Petty (2004) demonstrated that this effect depends on source credibility. Participants’ product attitudes were more certain and predicted behavior better when they perceived themselves to be resisting the persuasive communication than when they did not, but only when the communication came from an expert source.
In sum, the ELM has proved to be a robust model for predicting the effects of advertising and marketing messages on consumer attitudes and behavior. It provides a clear theoretical framework for understanding the conditions under which typical executional variables will have an effect, thereby providing both a guide for how marketers can maximize the persuasiveness of their ads and how consumers can maximize their resistance to those ads. More recently, other models of persuasion have been developed that provide a somewhat different focus on how consumers process persuasive communications, and in particular the thoughts consumers have about motives underlying messages. In the next section, we analyze the persuasion knowledge model.
Persuasion Knowledge Model
Unlike the theory of reasoned action and elaboration likelihood model, both of which originated in the field of so
cial psychology, the persuasion knowledge model (Friestad & Wright, 1994) is uniquely marketing-focused. Although the model could likely be applied to persuasion in other contexts, to date its focus has been on the interaction between marketers and consumers. Because the model is relatively new and has had little exposure outside of the marketing literature, we begin with a more thorough description of the model compared to the previous ones we discussed, and then proceed to discuss recent tests of the model.
The persuasion knowledge model was formally introduced in 1994 as the first model to explain how knowledge of marketers’ persuasion tactics affects consumers’ responses to such tactics (Friestad & Wright, 1994). The model asserts that over time, consumers develop knowledge of marketers’ persuasion tactics and, in doing so, become better able to adapt and respond to such attempts in order to achieve their own personal goals. Figure 19.1. provides a depiction of the model and its components (for a review, see Campbell & Kirmani, 2008).
Friestad and Wright (1994) decompose the persuasion process into two primary elements: the target and the agent. The target refers to the intended recipient of the persuasion attempt (the consumer), whereas the agent represents whomever the target identifies as the creator of the persuasion attempt (the marketer). The persuasion attempt encompasses not only the message of the agent, which itself is influenced by the agent’s knowledge of the topic, target, and the effectiveness and applicability of different persuasion tactics, but also the target’s perception of the agent’s persuasion strategy.
The SAGE Handbook of Persuasion Page 64