And yet, we also see ways the basic model is stretched or even challenged in applied scholarship. Sometimes inoculation pretreatments enhance threat and counterarguing but fail to confer resistance (e.g., Pfau et al., 2008). Some political inoculation studies found inoculation protects attitudes against attacks, but also derogates the sources of the attack messages (e.g., Pfau et al., 1990), an effect not clearly connected to the analogy. To further explore how such findings fit within a larger conversation of inoculation in laboratory and field studies, we review key issues in contemporary inoculation theory next.
Issues in Contemporary Inoculation Theory
* * *
Inoculation is a classic persuasion theory—“the grandparent theory of resistance to attitude change” (Eagly & Chaiken, 1993, p. 561)—and it continues to have contemporary relevance. Some contemporary inoculation issues reflect components of the original model, while others stretch the model into new domains and mechanisms.
Consider, for example, how contemporary work with the issue of timing has supported and challenged assumptions of the analogy. Timing involves two separate yet related issues: What is the optimum amount of time between pretreatment and attack, and how long do inoculation effects last? These are issues of timing and decay, respectively (Compton & Pfau, 2005). As McGuire (1964) first reasoned, those inoculated need some time between pretreatment and attack for resistance to strengthen, (e.g., through continued counterarguing). The connection to the medical inoculation is tight: A body needs time to produce antibodies, for example, after it is alerted to an offending agent. But, with time, those inoculated likely forget and lose motivation to counterargue (Insko, 1967).
Much of the early inoculation research used short delays: a matter of seconds (e.g., McGuire, 1961a) or days (e.g., McGuire, 1961b). But more recent research has extended the delay between pretreatment and attack to weeks (e.g., Ivanov et al., in press) or even up to 21 months (Pfau & Van Bockern, 1994). Now, a common delay between pretreatment and attack is about two weeks (e.g., Ivanov et al., in press). Banas and Rains’s (2010) meta-analysis suggested that the ideal length of the delay between pretreatment and attack does not line up with how McGuire (1964) and others (e.g., Compton & Pfau, 2005) have theorized. Instead of finding that resistance gradually increases after the pretreatment before eventually dropping off, Banas and Rains found no difference between immediate and moderate delays, with resistance diminishing after about two weeks.
If inoculation effects eventually wear off, wouldn’t booster sessions be a way to prolong resistance? It was an idea first proposed by McGuire (1964), and it is continued in more recent scholarship; but to date, effects of boosters on resistance are mild, at best, from early work (e.g., McGuire, 1961) through more recent study (e.g., Pfau et al., 1997). Have we lost an analogical link? Then again, perhaps boosters can enhance inoculation, but we haven’t yet discovered optimal timing of when to boost (Compton & Pfau, 2005).
While some scholars have focused on dynamics of inoculation in contemporary studies, such as timing, others turn to mechanisms of resistance. Banas and Rains’s (2010) meta-analysis of inoculation theory scholarship, which analyzed decades of laboratory studies, supported the basic model in some ways, but failed to find support for other key predictions. Among their findings, results supported inoculation’s comparative advantage over supportive pretreatments and no pretreatment (control), as well as inoculation’s ability to confer resistance to refutational-same and refutational-different attacks. However, in regard to the basic model, results failed to find impacts of elicited threat on resistance. (The researchers point out that power was low for their test of threat.)
Results of Banas and Rains’s (2010) meta-analysis and other studies that confront the basic model warrant another look at the analogy. Has the analogy outlived its usefulness? Or do we need to fine-tune the way we are referencing the analogy in our theorizing and scholarship? A review of issues in contemporary inoculation theory, including an overview of what we know of inoculation theory’s moderators, mediators, and outcomes, helps to offer some answers to these and other questions.
Moderators: Pre-attitude, Involvement, Self-Efficacy
Pre-attitude
Pre-attitude is an especially important moderator in the process of inoculation. Because inoculation is a preventative strategy, the target attitude must be in place before inoculation (Compton & Pfau, 2005; McGuire, 1964). When the intended attitude is not in place, inoculation messages can have a persuasive effect, but such an application of inoculation messages is not consistent with the domain of the theory (Wood, 2007). Pre-attitude, then, is a moderator of unique importance, serving as both a moderator during inoculation and as a necessary condition for inoculation.
Most contemporary inoculation scholarship assesses pre-attitude and then accounts for pre-attitude in analysis (e.g., treating it as a covariate, e.g., Pfau et al., 2009). Taking an even more precise look at pre-attitude, Ivanov, Pfau, and Parker (2009b) examined attitude bases (affective or cognitive) and subsequent effects on pretreatment message strategies (affective or cognitive), finding that matching the base of an attitude to message strategy is most effective (e.g., an affective-based inoculation pretreatment message is more effective with an affective-based attitude).
Perceived Involvement
When inoculation scholars study involvement, they are usually studying people’s perceived involvement with an issue. Pfau and his colleagues (1997) once reasoned: “[I]nvolvement holds the key to inoculation’s terrain” (p. 210), theorizing that when involvement levels are too high, people will not experience more threat. On the other hand, if involvement levels are too low, people will not care enough about the issue to experience threat. Consequently, because threat is a requisite for inoculation, low and high involvement levels should counter inoculation. Moderate involvement levels, they concluded, are most susceptible to inoculation pretreatments—those moderately involved care enough, but not too much, to experience threat. In a departure from the basic model, results of their study suggested involvement, and not threat, directly led to counterarguing, and with low- and high-involvement issues, they found a direct path from involvement to resistance (Pfau et al., 1997). Banas and Raines’s (2010) meta-analysis failed to find support for involvement levels enhancing resistance, although power was low (.12).
Later, Pfau and colleagues found that preinoculation involvement levels lead to resistance by working through anger and not traditional mechanisms of threat or counterarguing (Pfau et al., 2001). Even when looking at specific types of initial involvement, most involvement types “bypassed the mechanisms of threat and counterarguing, and instead, exert[ed] a direct impact on elicited anger, attitude strength, and resistance to persuasive attacks” (Pfau et al., 2010, p. 12).
Self-Efficacy
One’s perception of self-efficacy can affect inoculation. Higher levels of pre-inoculation self-efficacy boost resistance when inoculation messages use anger appeals, but when inoculation messages use rational appeals, a moderate level of self-efficacy is optimal; when self-efficacy is low, inoculation messages that use happiness appeals are most effective (Pfau et al., 2001). Extant theory does not offer much clarity as to why these differences emerged.
Of these moderators, pre-attitude fits the medical analogy most clearly. Just as people cannot be inoculated against a disease they already have, people cannot be inoculated against a position they already hold. Identifying the presence and valence of an attitude is a crucial first step in successful inoculation. The connection is less clear with involvement and self-efficacy as moderators of inoculation. However, treating the two variables as mediators of inoculation may offer a better fit. This is a possibility we explore next, in addition to the mediators suggested by the basic model.
Mediators in the Basic Model: Threat and Counterarguing
Threat
We have argued before that threat is required for inoculation (Compton & Pfau, 2005), something first proposed by McGuire
in his original explanations for how inoculation works (McGuire, 1964). Without threat, resistance is not inoculation-conferred resistance—whether implicit (from exposure to counterattitudinal content in the pretreatment message) and/or explicit (a product of a forewarning). To date, research has failed to differentiate how much independent threat is caused by each (Wood, 2007).
Recently, scholars have turned attention to constructing more focused manipulations of threat (see Banas & Rains, 2010) to generate more threat. In most inoculation research, threat levels fail to exceed moderate ranges (Compton & Pfau, 2005). To elicit more threat, researchers have varied language intensity in pretreatment messages (Pfau et al., 2010), constructed affective-negative pretreatments (Pfau et al., 2009), and exploited the force of psychological reactance (Ivanov et al., 2011). From such research, we have learned that stronger language does not necessarily enhance threat (Pfau et al., 2010), affective-negative messages can generate more threat than affective-positive messages (Pfau et al., 2009), and boosting reactance during inoculation boosts resistance (Ivanov et al., 2011). Studies like these are illuminating not only the role of threat, but also, the larger processes of inoculation itself.
Counterarguing
Scholars began regularly measuring counterarguing output in the 1990s, and more recently, have taken an even closer look at this process of inoculation-conferred resistance. Results indicate most post-inoculation refutational preemption is cognitive (Pfau, et al., 2009; Wigley & Pfau, 2010a), but when affective counterarguments are generated, they are strong (Wigley & Pfau, 2010a).
Since the earliest research, scholars assumed that counterargument output grows in the days following inoculation pretreatments, and then begins to decay (McGuire, 1964). As Insko (1967) pointed out, those inoculated need time to generate “belief-bolstering material” (p. 316), but their motivation to bolster the belief also diminishes with time. Yet these assumptions about counterarguing have not received consistent empirical support. Instead, inoculation messages appear to immediately boost counterarguing (Pfau et al., 2006; Pfau et al., 2009). One study found counterarguing lasts for at least 44 days, and then plays an active role during resistance to the attack message (Pfau et al., 2004, 2006), while another study found that counterarguing output dissipates after about two weeks (Pfau et al., 2009). Some research shows that inoculation elicits more counterarguing, but counterarguing doesn’t directly lead to resistance (e.g., Pfau et al., 2008).
Moderators Beyond the Basic Model: Involvement, Self-Efficacy, Affect, and Talk
Perceived Involvement
Earlier research assessed initial involvement levels on resistance—that is, how involved people were with the issue prior to the inoculation pretreatment. Beginning in 2004, researchers began assessing perceived involvement as a product of inoculation pretreatments—a dynamic variable elicited by the inoculation message (e.g., Compton & Pfau, 2004). Researchers wondered if the active process of issue consideration (e.g., counterarguing) motivated by inoculation treatment messages leads to changes that go beyond attitude strength. And that is what they found. A single ino culation pretreatment message can enhance perceived involvement with a target issue (e.g., Compton & Pfau, 2004, 2008; Pfau et al., 2004, 2005, 2009). Pfau and colleagues (2004) found elicited involvement to boost resistance through traditional mechanisms: pretreatments enhanced threat, which then enhanced involvement, which then lead to resistance through counterarguing and attitude accessibility.
Self-Efficacy
As with issue involvement, the first inoculation studies to explore self-efficacy treated it as an independent variable only—perceptions of self-efficacy held prior to inoculation (Pfau et al., 2001). Later, researchers assessed self-efficacy as a dependent variable, finding that pretreatments enhance perceptions of self-efficacy (Pfau et al., 2009). We can assume that perceived self-efficacy enhances resistance, but to date, its function as a moderator can only be inferred. It is a product of inoculation pretreatments, but not yet clearly linked to the process of inoculation.
Affect
For much of inoculation’s development, and despite threat’s importance as a motivator, inoculation-conferred resistance was viewed primarily through a cognitive lens—raising and refuting counterarguments to strengthen positions (McGuire, 1964). But beginning in the 2000s, researchers began taking closer looks at affect.
The earliest affect inoculation research (Pfau et al., 2001) found anger boosts, while happiness weakens, resistance. We know from more recent research that inoculation is optimized “when the refutational preemption component of inoculation messages feature arguments supported by hard evidence in addition to the use of affect triggers which signal that goals may be thwarted” (Pfau et al., 2009, p. 93). Additionally, inoculation makes people less fearful immediately after the pretreatment, and as the process continues, those inoculated get angrier and less happy (Pfau et al., 2009).
Precise effects of affect—and of different types of affect—during inoculation are difficult to pin down. We are in early stages of understanding affect in inoculation. And yet, some consistencies are beginning to emerge. Consider, for example, findings that point to unique and important roles of anger during (e.g., Ivanov et al., 2011; Pfau et al., 2001)—and even after—(Pfau et al., 2009) the inoculation process. Ivanov and Miller and colleagues (2011) approached anger with an innovative twist, turning to another classic theory of persuasion: reactance (see Brehm, 1966). Reactance describes how persuasion efforts can fail when message recipients interpret persuasive attempts as threats to their freedoms. Reactance, then, can be an obstacle to persuasion. But building on recent findings by Dillard and Shen (2005) that conceptualized reactance as anger and negative cognitions, Ivanov and Miller and colleagues exploited reactance instead of trying to avoid it. They designed inoculation messages specifically to elicit anger by framing the future persuasive attack as a threat to freedom, and this, in conjunction with the negative cognitions consistent with how inoculation confers resistance (counterarguing), resulted in enhanced resistance. Evidence is accumulating that anger plays an important role in inoculation-conferred resistance, consistent with a growing body of research that suggests, in a more general sense, inoculation pretreatments are triggering emotions and feelings, and these affect dimensions are affecting resistance.
Post-inoculation Talk
Usually, scholars assume that counterarguing in inoculation is “an internal process” (Pfau et al., 2006, p. 144), or a “silent dialogue” (Wigley & Pfau, 2010a, p. 218). Yet recent evidence suggests inoculation also triggers external dialogue. Post-inoculation, people talk about the issue with their friends and family (Compton & Pfau, 2004; Ivanov et al., in press), even about beliefs that they perceive to be minority positions (Lin & Pfau, 2007). Building off of these findings, scholars built a case for why inoculation increases talk, surmising when people are shaken by the threat process of inoculation and buoyed by the preemptive refutation, this causes them to turn to their friends for support (to alleviate threat) or advocacy (to pass along the newly acquired refutational content; Compton & Pfau, 2009). Compton and Pfau (2009) also suggested that post-inoculation talk plays a role in how inoculation confers resistance—a conclusion that has since been empirically supported (Ivanov et al., in press). Talking about the issue seems to strengthen attitudes about the issue.
Does the analogy allow for new moderators such as affect and talk? Yes, it does, although connections may not be as immediately clear as those between refutations and antibodies, viruses, and counterarguments. For example, with affect, the best analog may be psychoneuroimmunology (PNI), a field of immunity research that explores dynamic relationships between emotions and immunity responses, including the influence of emotions on vaccination efficacy (see Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002). Affect, in both cases, determines immunity. Considering new inoculation moderators can be consistent with the analogy when we maintain a focus on the two main components—threat and counterarguing—and adopt a more
nuanced view of biological inoculation.
Outcomes: Attitudes, Self-Efficacy, Behavioral Intentions
Attitudes
For many inoculation scholars, the most important benchmark for inoculation’s efficacy is whether a target attitude, postattack, demonstrates resistance to persuasion. The simple question is: Was the attack message more persuasive to those who did not receive an inoculation pretreatment compared to those who did? Beginning in the 2000s, researchers also began to assess inoculation’s impacts on attitude dimensions beyond valence, exploring attitude strength (e.g., Pfau et al., 2003), certainty (e.g., Pfau et al., 2004), and confidence (e.g., Compton & Pfau, 2004). These attitudinal dimensions have been used—in conjunction with attitudes toward an issue—as measures of conferred resistance.
Behavioral Intentions
Inoculation scholars have also assessed inoculation’s impacts on behavioral intentions, particularly in applied scholarship. As examples: inoculating against credit card marketing message affects credit behaviors like efforts to increase credit card debt (Compton & Pfau, 2004), and inoculation in a political context can affect voting intentions (e.g., Pfau, Park, Holbert, & Cho, 2001).
Alternative and Supplementary Explanations for Resistance
Source Derogation
Tannenbaum’s congruity research in the 1960s examined how source derogation impacts resistance (e.g., Tannenbaum & Norris, 1965), and source derogation is sometimes considered a competing explanation for resistance to influence. But instead of treating it as an alternative resistance process, some inoculation scholars have considered source derogation as an outcome of inoculation (e.g., Ivanov et al., 2011), although successful inoculation does not always derogate sources of attack messages (e.g., Pfau et al., 2007). One early inoculation study even attempted to link source derogation to the medical analogy: “Extending the biological ‘inoculation’ analogy with some cautions, we are then asking the question: ‘Does the prestige and status of the physician who administers a drug alter the effectiveness of the drug in combating disease?” (Anderson, 1967, p. 351). We may find that inoculation-elicited counterarguing does just that by not only raising refutations of issue-specific counterarguments, but also, refutations of the source’s credibility.
The SAGE Handbook of Persuasion Page 46