7 Ease and persuasion Multiple processes , meanings , and effects

This chapter describes the mechanisms through which the experi­ ence of processing ease, or fluency, can influence attitudes and persuasion. In particular, we argue that ease can impact attitude change not only by serving as a peripheral cue (e.g., being experienced as positive affect or as input to an availability heuristic), but also by affecting the thoughts people generate and the confidence with which those thoughts are held. Of impor­ tance, the conditions necessary for each of these processes to operate are specified in this review. Because the different mecha­ nisms operate in different contexts, appreciation of the mUltiple roles for ease can shed new light on situations in which ease effects should be more or less likely to emerge, and more or less likely to persist. This chapter describes the mechanisms by which the experience of processing ease, or fluency, can influence attitudes and persuasion, specifying the conditions under which several distinct processes are likely to operate. In particular, we argue that ease can impact attitude change not only by serving as a peripheral cue (e.g., being experienced as positive affect or as input to an availability heuristic), but also by affecting the thoughts people generate and the confidence with which those thoughts are held. We begin by providing a brief description of some of the persuasion paradigms in which ease has been studied. We focus on paradigms in which people persuade themselves through their own thoughts, either in response to persuasive messages from external sources or in response to instructions to produce their own messages. In each case, the ease with which those thoughts come to mind plays a critical role in persuasion. The next section describes work on ease of retrieval conducted in the domain of attitude change, focusing on the multiple processes by which ease of retrieval effects can operate and examining the moderating conditions for each of those processes. Finally, we provide an overview of persuasion research suggesting that ease can be associated with different meanings. 102 Pablo Brinol, Zakary L. Tormala and Richard E. Petty Persuasion as a function of thoughts Classic and contemporary research on persuasion suggests that persuasive messages can influence people's attitudes through both thoughtful and non­ thoughtful routes (e.g., Petty & Cacioppo, 1986). When persuasion is thoughtful, attitudes depend on the thoughts people generate in response to messages or message topics. The idea that persuasion depends on the extent to which indi­ viduals articulate and rehearse their own idiosyncratic thoughts to external messages was first outlined by Greenwald (1968) in what he called a cognitive response theory of attitude change (for a comprehensive review, see Petty, Ostrom, & Brock, 1981). This view essentially argues that people are persuaded (or resist persuasion) by virtue of their own thoughts rather than by learning the message per se, as had been argued by earlier learning theories (Hovland, Janis, & Kelley, 1953). Persuasive appeals that elicit thoughts that are primarily favora­ ble toward a particular recommendation (e.g., "if that new laundry detergent makes my clothes smell fresh, I'll be more popular"), produce agreement whereas appeals that elicit unfavorable thoughts toward the recommendation do not (e.g., "I'm no better offjust because that new laundry detergent comes in an attractive box"), regardless of whether the message content can be learned and recalled. Although most work on persuasion focuses on messages that come from other people (e.g., advertisers), messages that people generate themselves can also be quite effective in producing attitude change. The persuasive effect of completely self-generated messages was shown in early research on role-playing. This litera­ ture shows that individuals who generate arguments (e.g., following instructions to convince a friend to quit smoking) are more persuaded than those who receive the same information passively (e.g., Janis & King, 1954). In this research, people were typically asked to generate messages on certain topics (e.g., the dangers of smoking), and their subsequent attitudes were compared with those in a control group who had either passively listened to the communication or who had received no message. Generally speaking, active generation ofa message was shown to be a successful strategy for producing attitude change (Watts, 1967; Huesmann, Eron, Klein, Brice, & Fischer, 1983). Just as the act of generating a communication on a topic has been found to influence one's position on that topic, so too has the mere anticipation of performing such an act, by affecting the thoughts that people generate about the topic (e.g., Cialdini & Petty, 1979). In addition to generating and anticipating messages, research has shown that people can persuade themsel ves when they try to remember past behaviors, imag· ine future behaviors, explain some behavior, or merely think about an event. For example, people who are asked to imagine hypothetical events come to believe that those events have a higher likelihood of occurring (e.g., Anderson, 1983; Anderson, Lepper, & Ross, 1983; Sherman, Cialdini, Schwartzman, & Reynolds, 1985). In another line of research, Tesser and colleagues showed that merely thinking about an attitude object, without any external information presented or I requested, can lead to attitude change. For example, spontaneously thinking I . about a person who did something nice leads to more favorable evaluations of Ease and persuasion 103 that person (compared to when distracted from thinking), whereas thinking about a person who was insulting leads to more negative evaluations (see Tesser, Martin, & Mendolia, 1995, for a review). Thus, when work on cognitive responses is considered along with the research just described on role-playing and mere thought, it suggests that virtually all attitude change stems from self-persuasion. That is, at least when people are thinking, attitude change is based on the extent 10 which people generate favorable rather than unfavorable thoughts, on their own or in response to a persuasive message. Ease of thought generation Traditional approaches to persuasion have focused on primary thoughts-that is, the thoughts individuals have about attitude objects-whether self-generated or in response to a message. Interestingly, though, recent research suggests that in addition to primary thoughts, people can have secondary thoughts that is, thoughts about their primary thoughts or metacognitions (Brinol & DeMarree, 2012; Petty, Brinol, Tormala, & Wegener, 2007). According to this metacogni­ tive view, generating favorable or unfavorable thoughts in response to a persua­ sive message is an important factor in producing attitude change, but what people think about their thoughts is critically important as well (e.g., Petty, Brinol, & Tormala, 2002). Of particular relevance to the current chapter is the perceived ease with which people's thoughts come to mind. One of the earliest demonstrations of the effect of ease of thought generation on judgment came from Schwarz, Bless, Strack, Klumpp, Rittenauer-Schatka, and Simons (1991) now classic ease of retrieval studies. Schwarz et al. asked participants to rate their own assertiveness after generating six versus twelve examples of their own assertive behavior. They found that people viewed them­ selves as more assertive after retrieving six rather than 12 examples. This result was initially surprising because a straightforward application of accessibility and self-persuasion principles would have suggested that people generating 12 instances of assertiveness would have judged themselves to be more assertive than those generating six instances. The fact that the opposite was observed suggested that something beyond the mere number and direction of thoughts generated must have played a role. Schwarz and colleagues concluded that people also considered the ease with which their thoughts could be retrieved or generated. Since this initial demonstration, the ease of retrieval effect has been observed in numerous domains andacross diverse topics and measures (see Schwarz, 1998, 2004, Sanna & Lunberg, 2012, for reviews). In an example from our own research that is particularly relevant to persuasion, Tormala, Petty, and Brinol (2002) asked undergraduates to generate either two or ten arguments in support of a new campus policy. Results indicated that generating two favorable thoughts led to more favorable attitudes than did generating ten favorable thoughts. Thus, thinking of fewer arguments was more persuasive than thinking of many, because of the ease of generating those arguments when just a few were requested. 104 Pablo Brinol, Zakary L. Tormala and Richard E. Petty "" , ,f" " When it was easier to think of favorable arguments, those arguments carried more weight. Thus, the subjective experience of ease can play an important role in self­ persuasion. Next, we discuss why ease matters. Multiple processes driving ease effects How does ease influence persuasion? Perspectives have varied with respect to the mechanisms driving ease of retrieval effects. Understanding these mechanisms is critical for a number of reasons, however, not the least of which is that it has implications for the immediate and long-term consequences of persuasion stem­ ming from ease. Fot example, the more (less) thoughtful the mechanism that is involved, the more (less) the persuasion it creates is expected to be durable, resistant, and impactful over time (Petty, Haugtvedt, & Smith, 1995). Consistent with the Elaboration Likelihood Model of persuasion (ELM; Petty & Cacioppo, 1986; Petty & Brinol, 2012), we suggest that the psychological processes mediat­ ing the effect of ease on attitude change can be organized into a finite set that operate at different points along an elaboration continuum. Under low thinking conditions, ease-like other variables-can influence attitudes by operating as a judgment cue or heuristic. This would typically produce an effect consistent with its valence which is generally positive (e.g., see Winkielman & Cacioppo, 2001). When the likelihood of thinking is relatively high, the same experience of ease can impact persuasion by affecting the direction of the thoughts that come to mind, or by serving as a piece of evidence (i.e., an argument) to be scrutinized. When elaboration is not constrained to be very low or high, ease can influence attitudes by affecting the amount of thinking that occurs. Thus, the ELM describes several processes of primary cognition through which variables such as ease can affect persuasion: by serving as a simple cue, by affecting either the amount or direction of thinking, and by functioning as an argument. In addition to these four possibilities, we have recently proposed that any vari­ able (including ease) can also impact whether or not people use their thoughts by influencing what people think about their thoughts. This idea is referred to as the self-validation hypothesis (Petty, Brinol, & Tormala, 2002). The key tenet is that generating or having thoughts is not sufficient for these thoughts to impact judg­ ment. Rather, people must also have confidence in the thoughts. When people perceive their thoughts to be valid, they have confidence in them and rely on them in forming their judgments. When people have doubt about their thoughts or perceive them to be invalid, they do not use them as a basis for judgment. Thus, self-validation provides a fifth mechanism by which variables such as ease can influence attitudes-by affecting thought confidence. Unlike previous mecha­ nisms of attitude change that focus on primary or first-order cognition, this new process emphasizes secondary or meta-cognition. As an illustration of the integrative power of this conceptual framework, consider the effect of another subjective experience-one's incidental emotions­ on evaluative judgments. Consistent with the ELM, prior research has shown that a person's emotions can influence attitudes through multiple processes (see Petty, Ease and persuasion 105 Fabrigar, & Wegener, 2003). First, when thinking is constrained to be low, emotions tend to serve as simple associative cues and produce evaluations consist­ ent with their valence (e.g., happiness leads to more persuasion than sadness; Petty, Schumann, Richman, & Strathman, 1993). When the likelihood of thinking is not constrained to be high or low by other variables, emotions can affect the extent of thinking. For example, people may think about messages more when in a sad than happy state either because sadness signals a problem to be solved (Schwarz, Bless, & Bohner, 1991) or conveys a sense of uncertainty (Tiedens & Linton, 2001). When thinking is high, emotions can bias one's ongoing thoughts (Petty et aI., 1993). For example, positive consequences seem more likely when people are in a happy rather than sad state (DeSteno, Petty, Wegener, & Rucker, 2000). Finally, the self-validation hypothesis suggests that emotions can also affect thought confidence (e.g., happy people have more confidence in their thoughts than do sad people). Consistent with this possibility, Briiiol, Petty, and Barden (2007) found that when placed in a happy (versus sad) state following a persuasive message, participants relied more on their valenced thoughts as a basis for their attitudes (e.g., forming favorable attitudes when their thoughts were favorable). We postulate that ease, like emotions, can influence persuasion through multi­ ple mechanisms. First, when thinking is low, ease should act as a simple cue to persuasion by invoking positive affect or a simple heuristic. In fact, in the very first report of the ease of retrieval effect, the explanation was based on a heuristic account. Specifically, Schwarz, Bless, Strack, Klumpp, Rittenauer-Schatka, and Simons (1991; see also Schwarz, 1998) argued that the effect is driven by an availability heuristic (Tversky & Kahneman, 1974), whereby the easier it is to generate information in favor of something, the more supportive information people assume there must be. Conversely, having difficulty induces the percep­ tion that there is little support available. When it is difficult to generate a list of positive thoughts about a policy, for instance, people are assumed to infer that there must not be many positive things about it. When it is easy to generate posi­ tive thoughts, on the other hand, people are assumed to infer that there must be many positive things about the policy. These simple inferences could provide simple cues to guide persuasion when one's motivation or ability to think is rela­ tively low (Rothman & Schwarz, 1998). Indeed, Kiihnen (2010) recently provided evidence that ease can influence judgment by working as a simple cue (at least when ease is salient) when motivation and ability to think are low. Also consistent with the notion that ease can operate through low thinking processes, ease has been known to provide a simple associative cue that produces judgments consistent with its valence. Specifically, ease has been shown to be associated with, and even actively produce, positive affect (Moons, Mackie, & Garcia-Marques, 2009; Winkielman & Cacioppo, 2001; Winkielman, Schwarz, Reber, & Fazendeiro, 2003). This feeling can become attached to or associated with a persuasive advocacy, and thus produce more favorable attitudes following that advocacy, perhaps via misattribution effects or classical conditioning. In sum, under low thinking conditions, ease of retrieving or generating argu­ ments can influence attitudes by operating as a cue implying that the arguments 106 Pablo Brinal, Zakary L. Tarmala and Richard E. Petty