The Good, the Bad, and the Ugly:
Message Sidedness Effects on Resistance to Persuasion and the Moderating Role of Brand Love
Selena Kurtdere
Department of Communication Science, University of Amsterdam Master Thesis PersCom
Hans Feenstra June 26, 2020
Abstract
Because of inconsistencies in previous message sidedness research regarding the persuasive effectiveness of two-sided advertising and incomplete findings regarding the moderating role of consumers’ initial attitudes, this study aims to gain a better understanding of message sidedness effects (one-sided vs. two-sided advertisements) on consumers’ resistance to persuasion (i.e., counterarguing and attitude bolstering) and the moderating effect of brand love. It is believed that exposure to two-sided advertisements compared to one-sided advertisements leads to less
resistance to persuasion among all consumers, to less resistance to persuasion among consumers with a lower level of brand love, and to more resistance to persuasion among consumers with a higher level of brand love. An experiment was conducted where message sidedness was
manipulated using self-created Red Bull one-sided and two-sided advertisements, and consumers’ levels of counterarguing, attitude bolstering, and brand love were measured. Findings suggest that two-sided advertisements compared to one-sided advertisements reduce counterarguing among consumers with a lower level of brand love, enhance counterarguing among consumers with a higher level of brand love, and enhance attitude bolstering among all consumers. This means that two-sided advertising is generally less effective in reducing consumers’ resistance to persuasion than one-sided advertising. However, two-sided advertising can be effective in reducing
resistance to persuasion when presented to consumers with a lower level of brand love. Future research is needed on message sidedness in other contexts and message sidedness effects on other measures of persuasive effectiveness, such as consumers’ purchase intention.
Keywords: message sidedness, two-sided advertising, resistance to persuasion, counterarguing, attitude bolstering, brand love
Table of Contents Abstract ... 2 Introduction ... 5 Theoretical Framework ... 8 Resistance to Persuasion ... 8 Message Sidedness ... 10
Message Sidedness Effects on Resistance to Persuasion ... 11
Brand Love ... 14
The Moderating Role of Brand Love ... 15
Hypotheses ... 17 Method ... 18 Sample ... 18 Procedure ... 21 Measures ... 22 Message Sidedness ... 22 Resistance to Persuasion ... 23 Brand Love ... 25 Results ... 26 Discussion ... 30
Message Sidedness and Brand Love Effects on Attitude Bolstering ... 33
Limitations and Recommendations ... 35
Conclusion ... 37
References ... 38
Appendix A. Manipulation Conditions ... 42
Appendix B. Scales ... 44
Introduction
Heinz ketchup so thick that it does not easily pour out of the bottle, Buckley’s cough syrup so powerful that it tastes awful, and the Volkswagen Beetle so ugly that you get to love its shape. Every product possesses both positive and negative attributes, but having the negative attributes presented in an advertisement is not something one would expect and is rather counterintuitive for marketers (Eisend, 2006; Lang, Lee, & Zwick, 1999). Traditionally,
marketing communications attempt to influence consumer behavior by only presenting one side of a product, that is, its positive attributes (Crowley & Hoyer, 1994; Eisend, 2006). In fact,
marketers are reluctant to showcase the negative attributes of their products alongside the positive attributes because they reasonably do not want to reinforce consumers’ negative attitudes towards their products (Fransen, Verlegh, Kirmani, & Smit, 2015). However, these so-called two-sided advertisements are believed to be an effective advertising technique according to a variety of persuasive communication studies (e.g., Crowley & Hoyer, 1994; Fransen, Verlegh, Kirmani, & Smit, 2015).
Still, there are conflicting results on the complex relationship of message sidedness (i.e., one-sided vs. two-sided messages) and its persuasive effectiveness; whereas some studies state that two-sided messages are more persuasively effective than one-sided messages (e.g., Fransen, Verlegh, Kirmani, & Smit, 2015; Knowles & Linn, 2004a), other studies state the opposite (e.g., Allen, 1991; O’Keefe, 1999). One of the reasons for these conflicting results may be the different ways persuasive effectiveness is measured across these studies. Often-mentioned measures for this persuasive effectiveness of messagesin other studies are attitude to the product, attitude to the brand, and purchase intention (e.g., Crowley & Hoyer, 1994; Eisend, 2006), although the technique of two-sided advertising was introduced as a means to reduce consumers’ resistance to persuasion (Fransen, Verlegh, Kirmani, & Smit, 2015; Knowles & Linn, 2004a; McGuire, 1964).
This resistance to persuasion takes many forms, but when related to message sidedness, it mostly takes the form of counterarguing (i.e., contesting the content of the message) and attitude
bolstering (i.e., actively thinking of arguments supportive of one’s attitude). It is important to distinguish between different measures of persuasive effectiveness because a reduction of one measure (e.g., resistance to persuasion) does not automatically mean an enhancement of the other measure (e.g., purchase intention).
Moreover, an important reoccurring moderator for message sidedness effects on
consumers’ resistance to persuasion is consumers’ initial attitudes towards the advocated view of the message (e.g., Allen, 1991; Crowley & Hoyer, 1994). One finding that seems to prevail in message sidedness research is that two-sided messages are more effective than one-sided messages among consumers initially opposed to the message’s advocated view (Crowley & Hoyer, 1994; Eisend, 2006; O’Keefe, 1999; Zuwerink & Cameron, 2003). Yet, there are also some inconsistent findings regarding this moderating effect as sometimes, initial attitudes are found to not moderate message sidedness effects at all (Allen, 1991). One reason for this inconsistency could be there being relatively few studies in which consumers’ initial attitude is either positive or negative (O’Keefe, 1999). Another reason could be that in those studies, consumers initially opposed to the advocated view of an advertisement are typically
operationalized as users of competing brands (Crowley & Hoyer, 1994), so not as consumers having a necessarily negative initial attitude.
Taken together, it becomes apparent that there remains an unexplored side of message sidedness research, that is, a potentially moderating effect of positive initial attitudes. Crowley and Hoyer (1994) also advocate the need for more insights into message sidedness effects among those initially favorable to the advocated view of an advertisement, as there could be a negative persuasive effect among this group of consumers that is yet to be discovered. Knowles and Linn
(2004a) confirm this possibility by stating that strategies intended to reduce resistance to persuasion can enhance resistance to persuasion when they introduce resistance that was not initially there, for example through mentioning something negative about a beloved brand. A fairly recent concept that measures consumers’ initial attitudes and allows for positive initial attitudes to be measured—more adequately than “users of competing brands”—is the concept of brand love. Brand love refers to, among other things, the passion-driven behaviors and positive emotional connection that show one’s love for a brand, and predicts resistance to negative information about the brand (Batra, Ahuvia, & Bagozzi, 2012).
Besides the theoretical need for clarity regarding message sidedness effects on resistance to persuasion, and for greater research focus on the moderating effect of positive initial attitudes on message sidedness effects on resistance to persuasion, there is also a practical need for examining these effects. In today's marketing environment, marketers are constantly attacked by competitors and governmental authorities which are difficult communication situations in which two-sided advertisements have the potential to be persuasively effective (Eisend, 2006). By knowing the exact effects of two-sided advertisements compared to one-sided advertisements on consumers’ resistance to persuasion, marketers can apply message sidedness in their
advertisements more frequently and effectively than they do now. Besides, today’s marketing environment consists of not only consumers with negative initial attitudes but also a growing number of brand lovers (Batra et al., 2012). This makes research on the potentially moderating effect of consumers’ levels of brand love on message sidedness effects on consumers’ resistance to persuasion valuable to marketers as it shows under which levels of brand love, two-sided advertising is likely to be effective and ineffective.
This study aims to test message sidedness effects on consumers’ resistance to persuasion and the potentially moderating role of consumers’ levels of brand love on these effects by
answering the following research question: What are the effects of message sidedness on resistance to persuasion and are these effects moderated by brand love? An experimental design is employed in which message sidedness is manipulated and brand love and resistance to persuasion are measured. In the following chapters, the key concepts, central and new theories, and expectations are described as well as the methodology behind the experiment. Then, the results of the experiment are provided and discussed to establish concrete generalizations about message sidedness and brand love effects on resistance to persuasion, and with that, to contribute to the current message sidedness literature.
Theoretical Framework Resistance to Persuasion
Consumers are not always open to persuasive messages such as advertisements. In fact, they often arm themselves with several strategies to resist persuasion, which is a problem for marketers as they strive to influence consumer behavior with their marketing strategies. Fransen, Verlegh, Kirmani, and Smit (2015) subdivide these resistance strategies into avoidance,
contesting, and empowering strategies: avoidance strategies refer to strategies people adopt to avoid persuasion attempts (e.g., selective avoidance where one avoids content likely to contain information contradicting one’s own beliefs); contesting strategies refer to strategies people adopt to actively refute a message by challenging its content (i.e., counterarguing), the source (i.e., source derogation), or used persuasive tactics; and empowering strategies refer to strategies people adopt to strengthen the self or one’s existing attitudes (e.g., attitude bolstering where one actively thinks about arguments supportive of one’s existing attitudes without directly refuting the arguments presented in the message; Fransen, Smit, & Verlegh, 2015; Fransen, Verlegh, Kirmani, & Smit, 2015; Zuwerink & Cameron, 2003).
These strategies reflect a form of motivated resistance, that is, motivational states “in which people aim to reduce attitudinal or behavioral change and maintain their current attitude” (Fransen, Smit, & Verlegh, 2015, p. 1) —a conceptualization that echoes one of the earlier definitions of resistance to persuasion by McGuire: “the ability to withstand a persuasive attack” (1964, as cited in Knowles & Linn, 2004b, p. 4) —, rather than the more behavioral form of outcome resistance, that is, the lack of attitude change in response to a persuasive attack (Fransen, Smit, & Verlegh, 2015). This differentiation is important as it makes up for how
resistance is measured and ultimately, what is measured; whereas outcome resistance relies solely on the behavioral result of persuasion and therefore, does not provide insights in the oppositional forces persuasion may evoke when persuasion has no visible effect on the recipient, motivated resistance only captures those internal, oppositional forces which may not result in behavioral resistance (Knowles & Linn, 2004b). Because this study is limited in means to fully observe the behavioral outcomes of resistance, the focus of this study is on the motivational states evoked in consumers.
Furthermore, the source of resistance can be attributed either to traits of a person (i.e., resistance as a personality trait) or to another internal or external influence (Knowles & Linn, 2004b). Common external sources of resistance are threats to one’s freedom of choice or decisional autonomy (i.e., reactance; Brehm, 1966; Brehm & Brehm, 1981; Burgoon, Alvaro, Grandpre, & Voulodakis, 2002) because human beings are believed to have an innate desire for autonomy, and concerns of deception (i.e. skepticism) because people are believed to have a desire to hold a true and accurate belief system. A common internal source of resistance is a reluctance to change (i.e., inertia) which may happen for several reasons: for example, because people are believed to have a natural desire to stay the same which implies a desire for the known instead of the unknown or because people are believed to have a desire for consistency in their
existing beliefs, especially important ones (Fransen, Smit, & Verlegh, 2015; Knowles & Linn, 2004b).
Message Sidedness
In order to cope with this resistance, marketers have started applying strategies that aim to overcome consumer resistance to persuasion rather than strategies that aim to merely persuade consumers. Knowles and Linn (2004a) call these strategies omega strategies and alpha strategies respectively. Whereas alpha strategies focus on increasing approach forces, that is, building up consumers’ motivations to move towards buying behavior, omega strategies focus on decreasing avoidance forces, that is, reducing consumers’ motivations to move away from buying behavior, hence, omega strategies aim to neutralize resistance consumers may experience when exposed to an advertisement (Fransen, Verlegh, Kirmani, & Smit, 2015; Knowles & Linn, 2004a).
One of those omega strategies is the use of two-sided messages in communication efforts such as advertisements. Two-sided messages present both arguments in favor of a particular standpoint and opposing arguments but still favor one side, as opposed to one-sided messages that only present arguments in favor of a standpoint (Allen, 1991; Hovland, 1954). In
communication studies, this concept is referred to as message sidedness (Allen, 1991; Crowley & Hoyer, 1994; Eisend, 2006; Lang et al., 1999; O’keefe, 1993, 1999; Shen, 2013). Additionally, there are two types of two-sided messages: two-sided refutational and two-sided nonrefutational messages. Whereas two-sided refutational messages refer to messages that not only present the positive and negative arguments in a message but also refute the negative arguments in that message, two-sided nonrefutational messages refer to messages presenting both the positive and negative arguments in a message without refuting the negative arguments (Allen, 1991).
According to Allen (1991), two-sided refutational messages are more persuasive than one-sided messages but one-one-sided messages are more persuasive than two-one-sided nonrefutational
messages. However, O’Keefe (1993) criticized Allen’s (1991) findings by questioning his classification of the different sided message types and found evidence for both types of two-sided messages being more persuasive than one-two-sided messages. It is worth noting that such misclassification is a common reoccurrence in communication studies (see also Knowles & Linn, 2004a) since the positive arguments in a two-sided nonrefutational message are often mistaken for refutations while they do not directly refute the negative arguments in the message (O’Keefe, 1993, 1999).
Although the presence of refutation is the most common categorization feature of two-sided messages based on message structure, there are more message structure variables that influence the effect of two-sided messages such as the amount of negative information and the order of information presentation (Crowley & Hoyer, 1994; Eisend, 2006). Furthermore, message sidedness can be present in a non-advertising context as well as in an advertising context. Since this study focuses on two-sided messages in advertising context and such context is found to influence message sidedness effects (O'Keefe, 1999), hereafter will be referred to two-sided advertising. Two-sided advertisements feature positive and negative attributes of a product rather than arguments (Crowley & Hoyer, 1994).
Message Sidedness Effects on Resistance to Persuasion
In previous message sidedness literature, the direct persuasive effect of two-sided advertising has typically been explained by reference to inoculation theory (Cornelis, 2013; Crowley & Hoyer, 1994; Eisend, 2006; Etgar & Goodwin, 1982; McGuire, 1964; Shen & Bigsby, 2013), attribution theory (Crowley & Hoyer, 1994; Eisend, 2006), optimal arousal theory
(Crowley & Hoyer, 1994; Eisend, 2006), and the elaboration likelihood model (ELM; Cornelis, 2013; Crowley & Hoyer, 1994). As mentioned earlier, the persuasive effectiveness of two-sided advertising is measured differently across these studies, which could make up for the conflicting
results. For example, in line with attribution theory, two-sided advertising predicts consumers’ perceived credibility, but optimal arousal theory suggests that two-sided advertising predicts consumers’ attitudes towards the ad and brand. Also, some theories apply to one type (i.e., refutational vs. nonrefutational) of two-sided messages only, such as inoculation theory which emphasizes refutation. Due to the limited scope of this study and its focus on the effects of two-sided advertisements in their initial form, that is, with the mere inclusion of negative information, as opposed to positive-only one-sided advertisements, on consumers' resistance to persuasion, only the theories and frameworks that apply to two-sided nonrefutational advertisements and explain these exact effects are examined.
Two-sided advertisements are believed to derive their strength in two ways: by restoring freedom of choice and by eliminating concerns of deception (Fransen, Smit, & Verlegh, 2015). Firstly, two-sided advertisements owe their effectiveness to their consideration of consumers’ active participation in the processing of an advertisement (Cornelis, 2013); whereas one-sided advertisements trigger threats to freedom by imposing a certain claim (“the product is nothing but good”) and by ignoring possible counterarguments against the standpoint of an advertisement (Fransen, Smit, & Verlegh, 2015), two-sided advertisements emphasize freedom of choice by providing those counterarguments and by encouraging consumers to think for themselves (Tindale, 2004). Ultimately, one-sided advertisements are likely to lead to counterarguing as consumers tend to search for their own counterarguments if none are given. Two-sided advertisements, on the contrary, are likely to neutralize counterarguing as the given negative information reduces consumers’ tendency to come up with counterarguments themselves (Fransen, Smit, & Verlegh, 2015; Fransen, Verlegh, Kirmani, & Smit, 2015; Knowles & Linn, 2004a).
Secondly, while there are inconsistent findings regarding the overall persuasive
effectiveness of sided advertising, there seems to be consensus on the effectiveness of two-sided advertising on consumers’ perceived credibility of the advertisement; two-two-sided
advertisements are consistently found to enhance perceived credibility (Crowley & Hoyer, 1994; Eisend, 2006; Lang et al., 1999). This enhanced credibility can be explained with attribution theory, which describes consumers’ tendency to assign causes to events by asking “why?”
(Crowley & Hoyer, 1994). When applied to one-sided advertisements, attribution theory suggests a reduced perception of credibility since consumers are likely to attribute the positive-only information in a one-sided advertisement to the marketer’s desire to sell the product. This, in turn, might trigger consumers’ suspicions of ulterior motives and concerns of deception since such desire benefits the marketer more than the consumer.
Conversely, when applied to two-sided advertisements, attribution theory suggests an enhanced perception of credibility since the inclusion of negative information in a two-sided advertisement is not normative. Therefore, consumers are likely to attribute the positive and negative information in a two-sided advertisement to the actual characteristics of the product and conclude that the marketer has no manipulative intent and is telling the truth (Crowley & Hoyer, 1994; Eisend, 2006; Fransen, Smit, & Verlegh, 2015). O’Keefe (1999) confirms this with his finding of two-sided nonrefutational advertisements enjoying significantly greater credibility than one-sided advertisements—an advantage that was not found for two-sided messages in
nonadvertising context, probably because messages in nonadvertising context are not met with the same amount of initial skepticism as advertisements (i.e., advertising skepticism; see also Obermiller & Spangenberg, 1998; Fransen, Smit, & Verlegh, 2015; Fransen, Verlegh, Kirmani, & Smit, 2015; O’Keefe, 1999).
However, enhanced credibility does not automatically translate to enhanced overall persuasive effects (e.g., enhanced purchase intention) of two-sided advertisements as enhanced credibility of an advertisement applies to not only the positive attributes of a product but also the negative attributes. Consequently, enhanced credibility might reflect a “trade-off” with other measures of persuasiveness and ultimately, have a detrimental effect on those measures (Crowley & Hoyer, 1994; Eisend, 2006; O’Keefe, 1999).Yet, Crowley and Hoyer (1994) propose that enhanced credibility of two-sided advertisements does reduce counterarguing as consumers are less likely to counterargue messages they believe and trust. This proposition is supported by Fransen, Smit, and Verlegh (2015), who propose that when concerns of deception are absent, the likelihood of consumers contesting the content, source, or persuasive tactics of an advertisement decreases.
Brand Love
An often-mentioned moderator in message sidedness literature is the initial attitude of the consumer (e.g., Crowley & Hoyer, 1994; Eisend, 2006). Although the most reoccurring finding regarding the moderating role of initial attitude is that two-sided advertisements are more effective in reducing counterarguing than one-sided advertisements among consumers initially opposed to the advocated view of the advertisement (O’Keefe, 1999), there is much inconsistency regarding this moderating role. This study operationalizes consumers’ initial attitudes as
consumers’ levels of brand love as it is believed that brand love captures both negative and positive initial attitudes more adequately than another measure for initial attitudes (cf. Crowley & Hoyer, 1994).
The concept of brand love regained interest among both practitioners and theoretical researchers due to the recent article “Brand love” by Batra et al. (2012). Herein, rather than given a precise definition, brand love is characterized as a prototype, that is, a combination of features.
Those features are grouped into self-brand integration, passion-driven behaviors, positive emotional connection, long-term relationship, anticipated separation distress, attitude valence, and attitude strength. More importantly, positive attitudes play a central role in brand love; not only does brand love display positive attitude valence across a majority of its features, but it also echoes many characteristics of attitudinal strength such as greater attitude intensity (Batra et al., 2012). Besides, resistance to negative information is an element of the brand love prototype (i.e., one of the consequences) as it reflects a useful purpose of brand love and prototypes serve useful purposes (Batra et al., 2012).
The Moderating Role of Brand Love
As mentioned before, a reoccurring finding regarding the moderating role of initial attitudes is that two-sided advertisements are more effective than one-sided advertisements among those initially opposed to the advocated view of an advertisement. This finding finds support in social judgment theory, which explains that the receiver of a message is more likely to accept that message when the message falls into the receiver’s “latitude of acceptance” (Crowley & Hoyer, 1994). This happens when the information in the message matches the receiver’s initial attitude. This will then increase the probability of the message getting processed and decrease the probability of the message getting actively counterargued. So in the case of message sidedness, this would mean that the probability of a two-sided advertisement getting counterargued, decreases among consumers initially opposed to the advocated view of the advertisement as the negative information contained in two-sided advertisements falls into their latitude of acceptance (Allen, 1966; Crowley & Hoyer, 1994).
While social judgment theory is used to suggest a moderating effect of negative initial attitudes, it can also be applied to a potential moderating effect of positive initial attitudes, that is, of brand lovers. Whereas negative information contained in two-sided advertisements falls into
initially negative consumers’ latitude of acceptance, that same negative information might fall outside of initially positive consumers’ (i.e., brand lovers’) latitude of acceptance and thus, get counterargued. Conversely, one-sided advertisements that only contain positive information are likely to get accepted among brand lovers because that information falls into their latitude of acceptance.
Moreover, brand lovers tend to have strong positive attitudes and beliefs that are important to them—attitudes and beliefs based on careful elaboration as well as attitudes and beliefs tied to their self-view as beloved brands become integrated into brand lovers’ identity (Batra et al., 2012; Fransen, Smit, & Verlegh, 2015). When these strong attitudes are challenged, for instance by a two-sided advertisement, brand lovers are likely to experience cognitive
dissonance, that is, mental discomfort due to inconsistency within their belief system (Festinger, 1957). This happens because the negative information presented in two-sided advertisements is incongruent with brand lovers’ strong positive attitude. According to Festinger’s (1957) theory of cognitive dissonance, cognitive dissonance is a negative arousal state which people are motivated to reduce. In the formulation of his theory, Festinger (1957) emphasizes avoidance of inconsistent information as a means to reduce cognitive dissonance, hence, selective exposure became a topic of major interest and with a direct link to the theory of cognitive dissonance (Smith, Fabrigar, & Norris, 2008).
However, Fransen, Smit, and Verlegh (2015) argue that when experiencing such desire to stay consistent (i.e., reluctance to change), which may be even greater when confronted with inconsistent information to attitudes and beliefs important to oneself, people especially tend to employ the empowerment strategy of attitude bolstering. This is done because through attitude bolstering, people reinforce the particular attitude or belief that is challenged and subsequently, do not have to deal with cognitive dissonance. Furthermore, attitude-consistent information is
more pleasant and easier to process than attitude-inconsistent information (Festinger, 1957), as the latter is more cognitively effortful to process (Smith et al., 2008). Given the assumption that negative information about a brand in a two-sided advertisement is attitude-inconsistent for lovers of that brand, one could thus propose that brand lovers would employ the resistance strategy of attitude bolstering when presented a two-sided advertisement of their beloved brand. Hypotheses
To sum up, it is believed that exposure to two-sided advertisements, as opposed to exposure to one-sided advertisements, reduces consumers’ perceived threat to freedom and concerns of deception, which, in turn, reduces consumers’ resistance to persuasion (i.e., counterarguing). Therefore, the following hypothesis is proposed:
H1: Exposure to two-sided advertisements, as opposed to exposure to one-sided advertisements, will lead to less resistance to persuasion.
Furthermore, it is believed that the positive and negative information of two-sided
advertisements falls into the latitude of acceptance of those lower on brand love, but falls outside of the latitude of acceptance of those higher on brand love, which, in turn, reduces and enhances consumers’ resistance to persuasion (i.e., counterarguing) respectively. Also, it is believed that the positive and negative information of two-sided advertisements does not create cognitive dissonance among those lower on brand love, but does create cognitive dissonance among those higher on brand love, which, in turn, reduces and enhances consumers’ resistance to persuasion (i.e., attitude bolstering) respectively. Therefore, the following hypothesis is proposed:
H2: Exposure to two-sided advertisements, as opposed to exposure to one-sided
advertisements, will lead to less resistance to persuasion among those with a lower level of brand love, but to more resistance to persuasion among those with a higher level of brand love.
Figure 1
Conceptual Model
Method
As this study aims to test the effects of message sidedness on resistance to persuasion and the moderating effect of brand love, an online experiment is conducted in the form of a two-group between-subjects (Message sidedness: one-sided vs. two-sided)factorial design. A
quantitative, experimental approach was chosen because of this study’s focus on the independent variable (i.e., Message sidedness) and what it can predict, in other words, this study’s attempt to assess causal relationships.
Sample
The target population for this study includes all people over the age of 18 years (due to ethical convenience) who were familiar with the brand Red Bull (as a product by Red Bull was used in the manipulation). Because of this non-specific population and the need for
randomization in experimental studies, the sampling procedure was random cluster sampling using diverse Facebook survey sharing groups.
Out of 199 people that agreed to participate in the experimental study, thirty were
removed from the final dataset due to missing values. Analyses were, thus, conducted over a final sample of N = 169 participants in total. See Table 1 for the demographic characteristics of the sample.
Table 1
Frequencies and Percentages for Demographic Characteristics of Sample and P-Values for Fisher’s Exact Test
Variable Full sample One-sided
advertisement Two-sided advertisement p n % n % n % Gender .401 Male 61 36.1 34 39.5 27 32.5 Female 106 62.7 52 60.5 54 65.1 Other 2 1.2 0 0.0 2 2.4 Age .536 18–24 98 58.0 52 60.5 46 55.4 25–34 60 35.5 27 31.4 33 39.8 35–44 7 4.1 5 5.8 2 2.4 45–54 4 2.4 2 2.3 2 2.4 Nationality .313 China 5 3.0 2 2.3 3 3.6 Germany 11 6.5 7 8.1 4 4.8 India 5 3.0 3 3.5 2 2.4 The Netherlands 76 45.0 42 48.8 34 41.0 UK 12 7.1 8 9.3 4 4.8 USA 7 4.1 2 2.3 5 6.0 Othera 53 31.4 22 25.6 31 37.3
Table 1 (continued)
Frequencies and Percentages for Demographic Characteristics of Sample and P-Values for Fisher’s Exact Test
Variable Full sample One-sided
advertisement Two-sided advertisement p n % n % n % Education .048
Less than high school 1 0.6 1 1.2 0 0.0
High school 8 4.7 5 5.8 3 3.6 Associate degree 3 1.8 3 3.5 0 0.0 Bachelor’s degree 75 44.4 41 47.7 34 41.0 Master’s degree 79 46.7 33 38.4 46 55.4 Ph.D. or higher 3 1.8 3 3.5 0 0.0 Total 169 100.0 86 100.0 83 100.0
Note. N = 169. Using Fisher’s exact test, even distribution (randomization) over the one-sided and two-sided condition was checked on Gender, Age, Nationality, and Education.
a Other nationalities (Afghanistan, Australia, Azerbaijan, Bangladesh, Belgium, Brazil, Canada, Croatia, Czech Republic, Denmark, France, Greece, Honduras, Iceland, Indonesia, Ireland, Israel, Italy, Mexico, New Zealand, North Macedonia, Pakistan, Portugal, Romania, Russia, Sao Tome and Principe, Singapore, South Africa, South Korea, Switzerland, Turkey, Ukraine, and Vietnam) with n < 5.
In order to check for randomization regarding participants’ gender, age, nationality, and educational level over the message sidedness conditions, four Fisher’s exact tests were conducted
(see Table 1).1 There was no significant association between message sidedness and gender, age, and nationality respectively. Therefore, gender, age, and nationality were not added as control variables in the analyses for testing the hypotheses. There was a significant association between message sidedness and education (p = .048). Therefore, education was not yet dismissed as a potential control variable in the analyses for testing the hypotheses.
Procedure
Before the distribution of the online questionnaire, a pilot study was done to ensure the manipulations would be perceived as intended and the questions were comprehensible.
Suggestions from this pilot study for improvement of user-friendliness were implemented in the final questionnaire. Then, the final questionnaire was randomly distributed to anyone over the age of 18 and familiar with Red Bull. After the consent form ensuring voluntary and anonymous participation, and the filter questions making sure only participants over the age of 18 and familiar with Red Bull were able to continue, the Brand Love, reduced 6-item scale (Bagozzi et al., 2017) was presented. This was purposely done before the presentation of the advertisement in an attempt to minimalize short-term effects on participants’ levels of brand love. Then, one of the two Red Bull advertisements was randomly evenly presented using the randomizer tool in
Qualtrics, after which an adjusted version of the Bolster-Counterargue scale (BCS; Brinõl, Rucker, Tormala, & Petty, 2004) was presented. Lastly, the manipulation check items and demographic questions about the participants’ age, gender, nationality, and educational level were presented. After completion of the questionnaire, the participants were debriefed about the purpose of the study and the fact that the manipulations were fabricated for research purposes and not created by Red Bull.
1 Four Chi-square tests were conducted before that but because neither of those tests had all expected frequencies
Measures
Message Sidedness
Message sidedness was a dichotomous variable in this study and was manipulated using two self-created digital advertisements featuring high-selling brand Red Bull (Bjarnadotti, 2019; Higgins, Tuttle, & Higgins, 2010), more specifically its energy drink. Because brand love was to be measured, a made-up brand could not be used and a brand with a potentially almost equal division of brand lovers vs. “not brand lovers” was chosen. According to Sawyer (1973), high-selling brands “are able to isolate a segment which is negatively oriented toward that brand” (p. 31) and therefore, likely to benefit from two-sided advertisements. Given the prerequisites of initial attitudes towards the presented brand and of the inclusion of negative information about the presented product, Red Bull energy drink was chosen as the featured product in the
manipulation.
The two conditions of the manipulation were identical in terms of layout and message structure elements, except for the presented message in each condition. In one condition, the advertisement presented a positive-only message about a well-known positive attribute of Red Bull energy drink, that is, its energizing effect (better known as “it gives you wings”; Higgins, Tuttle, & Higgins, 2010; “Red Bull energy drink,” 2020). This one-sided advertisement read “[Red Bull] fuels the mind”. In the other condition, the advertisement presented the same message as the one-sided advertisement but with the addition of a negative message about a negative attribute of Red Bull energy drink, that is, its negative health effects (Bjarnadotti, 2019). This two-sided advertisement read “[Red Bull] toxicates the body, fuels the mind”. The
advertisements did not include elements of existing Red Bull advertisements other than a depiction of the energy drink can. See Appendix A for the actual advertisements.
In order to check whether the participants had perceived message sidedness as it was manipulated, two manipulation checks were conducted: one assessing whether the participants believed the presented advertisement conveyed a positive message only and another assessing whether participants believed the presented advertisement conveyed both a positive and negative message. An independent samples t-test indicated that the average participants’ belief that the advertisement presented a positive message only in the one-sided condition (M = 2.98, SD = 1.20) differed significantly from that in the two-sided condition (M = 2.16, SD = 1.18), t(167) = 4.48, p < .001, 95% CI [0.46, 1.18], d = 0.69.2 Another independent samples t-test indicated that the average participants’ belief that the advertisement presented both a positive and negative message in the one-sided condition (M = 2.34, SD = 1.25) also differed significantly from that in the two-sided condition (M = 3.57, SD = 1.23), t(167) = −6.43, p < .001, 95% CI [–1.61, –0.85], d = –1.00.3 Thus, it can be assumed that the manipulations were perceived as intended.
Resistance to Persuasion
Participants’ resistance to persuasion as a continuous dependent variable was assessed using an adjusted version of the Bolster-Counterargue, 12-item, 5-point Likert scale, ranging from 1 (not at all) to 5 (very much) (Brinõl et al., 2004). This scale consists of two subscales; the first six items measure participants’ attitude bolstering and the last six items measure
participants’ counterarguing. Given the broad concept of resistance to persuasion and this study’s focus on predicted resistance strategies that consumers may adopt when being exposed to
message sidedness in advertisements (i.e., attitude bolstering and counterarguing), a scale specifically measuring consumers’ adoption of these resistance strategies was chosen over one
2 Response categories to manipulation check item "How much do you believe the advertisement conveys a positive
message only?" as presented in the questionnaire: 1 (not at all) to 5 (very much).
3 Response categories to manipulation check item "How much do you believe the advertisement conveys both a
measuring another form of resistance. Furthermore, as the original items of the BCS are designed to measure preferences for resisting through attitude bolstering and counterarguing (Brinõl et al., 2004), those items were adjusted to measure actual adoption of attitude bolstering and
counterarguing when exposed to the advertisements of this study. Final example items are: “When I saw the advertisement, I reminded myself why my beliefs are important to me” and “When I saw the advertisement, I disputed what it had to say”. Items 10, 11, and 12 were reverse coded before analysis. See Appendix B for both the original and the adjusted BCS.
An exploratory factor analysis with Direct Oblimin rotation indicated that the scale was three-dimensional (see Appendix C, Table C1). 6-item subscale 1 (M = 2.51, SD = 1.14), 3-item subscale 2 (M = 2.55, SD = 1.18), and 3-item subscale 3 (M = 3.35, SD = 1.33) proved reliable as indicated by a Cronbach’s α of .93, .85, and .79 respectively. The items representing attitude bolstering and counterarguing loaded onto separate factors. This was expected given the original independence of the two subscales that make up the BCS (Brinõl et al., 2004).
However, the six items representing counterarguing also loaded onto different factors, which is remarkable since they do not represent different underlying traits but all represent the act of contesting the content of the advertisement. A possible explanation for the two derived
subscales for counterarguing is that items 10, 11, and 12 are the same statements as items 7, 8, and 9 but then negatively phrased (e.g., “I didn’t actively counterargue”) which could have confused the participants when selecting an appropriate response category. The remarkably higher mean score of subscale 3, as opposed to that of subscale 2, despite the subscales
measuring the same contextual dimension, and a significant, weak to moderate correlation (r = .38, p < .001) between subscales 2 and 3 support this explanation. Because of this and because of a too low Cronbach’s α of .53 for subscales 2 and 3 computed, subscale 3 was disregarded in the analyses for testing the hypotheses.
Moreover, subscales 1 and 2 were treated as independent resistance to persuasion
variables measuring the distinct dimensions of attitude bolstering and counterarguing, instead of as one single computed variable. Reasons for this were the original independence of the two subscales (Brinõl et al., 2004), a significant, weak correlation (r = .21, p = .009) between the subscales and a too low Cronbach’s α of .33 for the subscales computed. Therefore, hereafter subscale 1 will be referred to as Attitude bolstering and subscale 2 as Counterarguing.
Brand Love
Participants’ brand love as a continuous-made-dichotomous variable in the study was assessed by using the Brand Love, reduced 6-item, 5-point Likert scale, ranging from 1 (not at all) to 5 (very much), except for item 6 measuring participants’ overall attitude valence and ranging from 1 (extremely negative) to 5 (extremely positive) (Bagozzi et al., 2017). Example items are: “Please express the extent to which you feel emotionally connected to Red Bull” and “Please express the extent to which you believe that you will be drinking Red Bull for a long time”. See Appendix B for the Brand Love, reduced 6-item scale.
An exploratory factor analysis with Direct Oblimin rotation indicated that the scale was two-dimensional (see Appendix C, Table C2). Both 3-item subscale 1 (M = 2.30, SD = 0.96) and 3-item subscale 2 (M = 1.70, SD = 0.76) proved reliable as indicated by a Cronbach’s α of .83 and .71 respectively. There was a significant, moderate correlation (r = .50, p < .001 ) between the subscales and the subscales computed into one scale proved reasonably reliable as indicated by a Cronbach’s α of .66, hence, the two subscales were treated as one single computed variable.
Participants were divided into a lower vs. a higher relative group based on the median of brand love (Mdn = 1.83, IQR = 1.33–2.33) in the sample. An independent samples t-test indicated that the average brand love score in the lower group (M = 1.46, SD = 0.24) differed significantly from the average brand love score in the higher group (M = 2.69, SD = 0.58), t(94.11) = –17.23, p
< .001, 95% CI [–1.37, –1.09], d = –3.55.4 Thus, it can be assumed that the relative groups for brand love represent lower and higher brand love scores.
Results
It was predicted that message sidedness in advertising (i.e., one-sided vs. two-sided advertisements) affects consumers’ resistance to persuasion (i.e., attitude bolstering and counterarguing; H1) and that consumers’ levels of brand love moderate this effect (H2). Before these predicted effects were analyzed, two one-way ANOVAs were conducted to check whether participants' educational level needed to be controlled for in the analyses for testing the
hypotheses. The ANOVAs had Education as the independent variable, and Counterarguing and Attitude bolstering respectively as dependent variables. The ANOVAs showed that neither participants’ mean scores of counterarguing, F(5, 163) = 1.18, p = .321, nor participants’ mean scores of attitude bolstering, F(5, 163) = 0.39, p = .853, were significantly different across education categories. Therefore, education was not added as a control variable in the analyses for testing the hypotheses.
In order to test H1 and H2, two two-way ANOVAs were conducted with Message
sidedness (one-sided vs. two-sided) as the independent variable, Brand love (lower vs. higher) as the moderator, and either Counterarguing or Attitude bolstering as the dependent variable. The ANOVA testing effects on counterarguing indicated that neither message sidedness, F(1, 165) = 0.01, p = .938, nor brand love, F(1, 165) = 0.63, p = .429, had a significant effect but the
interaction of message sidedness and brand love, F(1, 165) = 13.46, p < .001, ηp2 = .08, had a significant, medium effect on counterarguing. Counterarguing was significantly higher in the two-sided condition than in the one-sided condition, but only in the higher brand love group; and
4 The assumption of equal variances in the population was violated, Levene’s F(167) = 50.69, p < .001, so degrees of
counterarguing was significantly lower in the two-sided condition than in the one-sided condition, but only in the lower brand love group (see Table 2 and Figure 2).
The ANOVA testing effects on attitude bolstering indicated that message sidedness, F(1, 165) = 6.49, p = .012, ηp2 = .04, had a significant, small effect; brand love, F(1, 165) = 3.53, p = .062, ηp2 = .02, had a marginally significant, small effect; and the interaction of message
sidedness and brand love, F(1, 165) = 1.89, p = .171, did not have a significant effect on attitude bolstering. Attitude bolstering was significantly higher in the two-sided condition than in the one-sided condition and was marginally significantly higher among those higher on brand love than among those lower on brand love (see Table 2 and Figure 3).
These findings suggest that exposure to two-sided advertisements compared to exposure to one-sided advertisements does not lead to less counterarguing, but does lead to more attitude bolstering. Thus, H1 is rejected. Also, these findings suggest that exposure to two-sided
advertisements compared to exposure to one-sided advertisements leads to less counterarguing among those with a lower level of brand love, but to more counterarguing among those with a higher level of brand love; and that there is no such interaction effect on attitude bolstering. Thus, H2 is partially accepted.
Table 2
Means and Standard Deviations for Resistance to Persuasion (i.e., Counterarguing and Attitude Bolstering) as a Function of a 2 (Message Sidedness) x 2 (Brand Love) Design
Message sidedness
Brand love
Lower Higher Total
M SD M SD M SD Counterarguing One-sided 2.94ab 1.21 2.15ab 1.03 2.59 1.20 Two-sided 2.28ab 1.10 2.78ab 1.19 2.50 1.17 Total 2.61 1.20 2.46 1.15 2.55 1.18 Attitude bolstering One-sided 2.03 0.94 2.59 1.09 2.28a 1.04 Two-sided 2.71 1.29 2.79 1.07 2.75a 1.19 Total 2.36c 1.17 2.69c 1.08 2.51 1.18
a Means between Message sidedness conditions differ significantly at p < .05 level. b Means between Brand love relative groups differ significantly at p < .05 level.
Figure 2
Counterarguing as a Function of Message Sidedness and Brand Love
Figure 3
Discussion
This study aims to answer the research question “What are the effects of message
sidedness on resistance to persuasion and are these effects moderated by brand love?”. Resistance to persuasion was measured using the Bolster-Counterargue scale (BCS; Brinõl et al., 2004), which measures the resistance strategies of counterarguing and attitude bolstering independently. The findings suggest that exposure to two-sided advertisements does not affect counterarguing, but does lead to more attitude bolstering than exposure to one-sided advertisements. With that, H1, proposing that exposure to two-sided advertisements, as opposed to exposure to one-sided advertisements, would lead to less resistance to persuasion, is rejected. Moreover, the findings suggest that brand love moderates message sidedness effects on counterarguing, as exposure to two-sided advertisements compared to exposure to one-sided advertisements leads to less
counterarguing among those with a lower level of brand love, but to more counterarguing among those with a higher level of brand love. However, the findings suggest that brand love does not moderate message sidedness effects on attitude bolstering. Therefore, H2, proposing that exposure to two-sided advertisements, as opposed to exposure to one-sided advertisements, would lead to less resistance to persuasion among consumers with a lower level of brand love, but to more resistance to persuasion among consumers with a higher level of brand love, is partially accepted.
Message Sidedness and Brand Love Effects on Counterarguing
The findings suggest a moderating effect of brand love on message sidedness effects on counterarguing. As expected, these findings correspond to social judgment theory, which states that people are less likely to counterargue a message that falls into their latitude of acceptance (Allen, 1966; Crowley & Hoyer, 1994). As the negative information in a two-sided advertisement likely falls into the latitude of acceptance of those lower on brand love and the positive-only
information in a one-sided advertisement likely falls into the latitude of acceptance of those higher on brand love, social judgment theory explains the moderating effect of brand love well.
Although it was believed that there would also be a direct effect of message sidedness on counterarguing regardless of consumers’ levels of brand love (i.e., consumers would generally counterargue the contents of a two-sided advertisement less than those of a one-sided
advertisement), the findings suggest that this is not the case. This could have several reasons. Firstly, two-sided advertisements, as opposed to one-sided advertisements, might not reduce consumers’ perceived threat to freedom or concerns of deception as was initially expected. In other words, the mere inclusion of negative information in a two-sided advertisement does not necessarily restore consumers’ freedom of choice, nor does it necessarily enhance consumers’ perceived credibility of the advertisement. If that is the case, the findings of this study contradict a large number of previous message sidedness studies that mention a restoration of freedom of choice (Cornelis, 2013; Fransen, Smit, & Verlegh, 2015; Fransen, Verlegh, Kirmani, & Smit, 2015; Knowles & Linn, 2004a; Tindale, 2004) and elimination of concerns of deception (Crowley & Hoyer, 1994; Eisend, 2006; Fransen, Smit, & Verlegh, 2015; Lang et al., 1999; O’Keefe, 1999) as two key forces behind two-sided advertisements.
Secondly, two-sided advertisements might enhance consumers’ perceived credibility of the advertisement but this enhanced perceived credibility might not translate to a reduction in counterarguing. This could happen when not only the positive information but also the negative information in a two-sided advertisement is believed more, which makes a reduction of
counterarguing unlikely. This "trade-off" of enhanced credibility and negative persuasive effects finds support in previous message sidedness research (Crowley & Hoyer, 1994; Eisend, 2006; O’Keefe, 1999) but is often met with inconsistent findings. Although this reason explains why two-sided advertisements, as opposed to one-sided advertisements, are not found to reduce all
consumers’ counterarguing among all consumers, it does not explain why two-sided
advertisements, as opposed to one-sided advertisements, are found to reduce counterargument among those lower on brand love.
Thirdly, it is most probable that the interaction effects of message sidedness and brand love on counterarguing cancel each other out, resulting in no found direct effect of message sidedness on counterarguing. This means that the reasoning (i.e., perceived threat to freedom and concerns of deception) behind a possible direct effect of message sidedness on counterarguing might still apply. A perceived threat to freedom and concerns of deception might be reduced by two-sided advertisements, as opposed to one-sided advertisements, but just among those with a lower level of brand love. This is not surprising considering that those lower on brand love might experience a greater threat to freedom and have more concerns of deception when seeing a one-sided advertisement than those higher on brand love because they are initially less positive towards the presented brand than those higher on brand love. Conversely, those higher on brand love might like the presented brand so much, that they do not experience their freedom of choice being limited and suspicions of ulterior motives being activated by a one-sided advertisement, and subsequently, do not show a similar reduction in counterarguing when seeing a two-sided advertisement.
Altogether, this study’s findings regarding counterarguing are mostly in line with previous message sidedness studies that include consumers' initial attitudes as a potential moderator of message sidedness effects on counterarguing. The moderating effect of negative initial attitudes on message sidedness effects has been widely examined and two-sided
advertisements were predominantly believed to reduce counterarguing among initially negative consumers. This study complements this belief by finding a significant, moderating effect of brand love—which also represents positive initial attitudes—and with that, suggests that brand
love is an antagonistic moderator of message sidedness effects on counterarguing. This implies that although two-sided advertising is seen as an advertising strategy aiming to reduce resistance to persuasion (i.e., omega strategy) in persuasion literature, it can also enhance resistance to persuasion when exposed to brand lovers.
Message Sidedness and Brand Love Effects on Attitude Bolstering
The findings suggest that two-sided advertisements, as opposed to one-sided
advertisements, enhance attitude bolstering, but that this effect is not moderated by brand love. So two-sided advertisements, as opposed to one-sided advertisements, do not reduce attitude bolstering among those lower on brand love and do not enhance attitude bolstering among those higher on brand love. This could mean that the theory of cognitive dissonance stating that people are motivated to reduce cognitive dissonance (Festinger, 1957), is not applicable to message sidedness as consumers lower and higher on brand love might not be motivated to reduce the attitude-inconsistent information in one-sided and two-sided advertisements respectively. However, it is more presumable that these consumers are motivated to reduce experience
cognitive dissonance by the attitude-inconsistent information in these advertisements but that this cognitive dissonance leads to other resistance strategies than attitude bolstering. This is because attitude bolstering is but one strategy to reduce cognitive dissonance and cognitive dissonance is a common occurrence when one encounters attitude-inconsistent information (Festinger, 1957; Fransen, Smit, & Verlegh, 2015; Smith, Fabrigar, & Norris, 2008).
It could also mean that those higher on brand love do adopt attitude bolstering to reduce experienced cognitive dissonance from the negative information in a two-sided advertisement, which is still in line with Festinger’s theory of cognitive dissonance (1957); but those lower on brand love do so as well, for another reason. It is unlikely that the inclusion of negative
dissonance which they aim to reduce, as that negative information should not be so attitude-inconsistent for them. Yet, negative information in advertisements is not normative and therefore, represents a discrepancy from the adaptation level, that is, “some kind of stimulation that the organism has been receiving in the recent past or is for some other reason expecting” (Berlyne, 1971, as cited in Crowley & Hoyer, 1994, p. 563). Optimal arousal theory posits that such novel stimuli are preferred over normative stimuli, which results in consumers being more motivated to process messages containing novel stimuli (Crowley & Hoyer, 1994). Perhaps attitude bolstering reflects not only one's resistance to persuasion but also one's processing of persuasion attempts, which could explain why attitude bolstering as a reaction to two-sided advertisements compared to one-sided advertisements is enhanced among both those higher and lower on brand love.
The findings also suggest a marginally significant direct effect of brand love on attitude bolstering. Attitude bolstering is marginally significantly higher among those higher on brand love than among those lower on brand love, regardless of exposure to different message sidedness types. Albeit marginally significant, this finding lends support to the possibility of attitude bolstering representing one’s processing of persuasion attempts rather than one’s resistance to persuasion. When disregarding this message sidedness, one is left with both brand love groups being exposed to Red Bull advertisements whose contents might be generally more attitude-consistent for those higher on brand love than for those lower on brand love. Given that attitude-consistent information is more pleasant and easier to process than attitude-inconsistent information (Festinger, 1957), it could be that those with a higher level of brand love process the persuasion attempts of a Red Bull advertisement more than those with a lower level of brand love, hence, the found direct effect of brand love on attitude bolstering.
Altogether, the findings regarding attitude bolstering suggest that two-sided
brand love level. If attitude bolstering as measured in this study reflects resistance to persuasion well, this finding illustrates a serious persuasive disadvantage of two-sided advertisements compared to one-sided advertisements, and with that, opposes a majority of the message
sidedness studies advocating the resistance-reducing ability of two-sided advertising. However, attitude bolstering as measured in this study could also reflect one’s processing of persuasion attempts rather than one’s resistance to persuasion. If that is the case, this finding implies that consumers process two-sided advertisements more than one-sided advertisements. This could be due to a higher perceived novelty of two-sided advertisements, as opposed to one-sided
advertisements, since the inclusion of negative information in two-sided advertisements is not normative.
Limitations and Recommendations
This study contributes to the current message sidedness literature by its attempt to provide clarity on message sidedness and brand love effects on resistance to persuasion, but there are some limitations. Firstly, as this study aims to examine the effects of one-sided vs. two-sided nonrefutational messages in an advertising context and operationalized these concepts
accordingly, the findings are generalizable to other contexts only to some extent. This is because two-sided nonrefutational advertisements have a different format than two-sided refutational advertisements and two-sided messages in nonadvertising context. Although the one element of message sidedness that should cause the greatest effect on persuasive effectiveness is the inclusion of negative information in the message, it remains unclear to what extent that effect is generalizable to message sidedness effects of sided refutational advertisements and two-sided messages in nonadvertising context. Therefore, future message two-sidedness studies should put greater research focus on the inclusion of refutation and messages in a non-advertising context.
Secondly, this study examined message sidedness effects on resistance to persuasion, and not on other measures of persuasive effectiveness, such as consumers’ purchase intention. This makes the findings of this study limitedly generalizable to other, more practical contexts. This is due to the possible “trade-off” effect of enhanced perceived credibility caused by two-sided advertisements with other measures of persuasiveness. For example, when a two-sided
advertisement reduces consumers’ resistance to persuasion by reducing their enhanced perceived credibility of the advertisement, it might also reduce consumers’ purchase intention as consumers might not only believe the positive information contained in the two-sided advertisement more but also the negative information (Crowley & Hoyer, 1994; Eisend, 2006; O’Keefe, 1999). Although marketers might benefit from this study’s findings on message sidedness effects on consumers’ resistance to persuasion, future message sidedness research is needed on other measures of persuasive effectiveness that are more visible and thus practical for marketers.
Finally, it should be noted that the way resistance to persuasion was measured in this study is not common in persuasion literature. The BCS was chosen to measure counterarguing and attitude bolstering but contains closed-ended questions. Although closed-ended question measures allow for comparability across studies and show higher levels of reliability, they are unable to determine the types of thoughts that participants generate when being exposed to a stimulus. During analyses, this does not provide as much depth or understanding of the given data in comparison to open-ended responses or generated thoughts listed during a thought-listing process (Ivanov, Parker, & Dillingham, 2013). Such deeper understanding is valuable for the future direction of message sidedness research because it shows what exactly in the
advertisement caused people to adopt the resistance strategies of counterarguing or attitude bolstering. With that information, marketers will be better able to design persuasively effective advertisements. Therefore, future message sidedness studies should either include more
open-ended questions or employ a different method aimed at gaining a deeper understanding of the types of thoughts generated by the participants.
Conclusion
Because of inconsistencies in previous message sidedness research regarding the persuasive effectiveness of two-sided advertising and incomplete findings regarding the
moderating role of consumers’ initial attitudes, this study aims to gain a better understanding of message sidedness effects (one-sided vs. two-sided advertisements) on consumers’ resistance to persuasion (i.e., counterarguing and attitude bolstering) and the moderating effect of brand love. It can be concluded that two-sided advertisements compared to one-sided advertisements reduce counterarguing among consumers with a lower level of brand love, enhance counterarguing among consumers with a higher level of brand love, and enhance attitude bolstering among all consumers. This means that, contrary to what some previous message sidedness studies claim, two-sided advertising is not always effective in reducing consumers’ resistance to persuasion, and one-sided advertising might be better to employ in general. However, in line with a majority of previous message sidedness studies, two-sided advertising can be effective in reducing resistance to persuasion when presented to consumers with a lower level of brand love. Future research is needed on two-sided refutational advertisements; message sidedness in nonadvertising context; and message sidedness effects on other, more practical measures of persuasive effectiveness, such as consumers’ purchase intention. For now, it might be wise for marketers to only praise their products’ thick consistency, curing effect, or lovely shape, without mentioning their slow pouring, awful taste, or ugly look.
References
Allen, M. (1991). Meta‐analysis comparing the persuasiveness of one‐sided and two‐sided messages. Western Journal of Speech Communication, 55(4), 390–404.
doi:10.1080/10570319109374395
Allen, V. L. (1966). [Review of the book Attitude and attitude change: The social judgment-involvement approach, by C. W. Sherif, M. Sherif & R. E. Nebergall]. American Sociological Review, 31(2), 283–284. doi:10.2307/2090931
Bagozzi, R. P., Batra, R., & Ahuvia, A. (2017). Brand love: Development and validation of a practical scale. Marketing Letters, 28(1), 1–14. doi:10.1007/s11002-016-9406-1
Batra, R., Ahuvia, A., & Bagozzi, R. P. (2012). Brand love. Journal of Marketing, 76(2), 1–16. doi:10.1509/jm.09.0339
Bjarnadotti, A. (2019, September 16). What are the side effects of drinking Red Bull? [Online article]. Retrieved from https://www.healthline.com/nutrition/red-bull-side-effects Brehm, J. W. (1966). A theory of psychological reactance. New York. NY: Academic Press. Brehm, S. S., & Brehm, J. W. (1981). Psychological reactance: A theory of freedom and control.
New York, NY: Academic Press.
Brinõl, P., Rucker, D. D., Tormala, Z. L., & Petty, R. E. (2004). Individual differences in resistance to persuasion: The role of beliefs and meta-beliefs. In E. S. Knowles & J. A. Linn (Eds.), Resistance and persuasion (pp. 83–104). Mahwah, NJ: Lawrence Erlbaum. Burgoon, M., Alvaro, E., Grandpre, J., & Voulodakis, M. (2002). Revisiting the theory of
psychological reactance: Communicating threats to attitudinal freedom. In J. P. Dillard & M. Pfau (Eds.), The persuasion handbook: Developments in theory and practice (pp. 213– 232). Thousand Oaks, CA: Sage Publications, Inc.
Cornelis, E. (2013). Persuasive effects of two-sided social marketing messages (Doctoral dissertation). Ghent University, Ghent, Belgium.
Crowley, A., & Hoyer, W. (1994). An integrative framework for understanding two-sided persuasion. Journal of Consumer Research, 20(4), 561–574. doi:10.1086/209370
Eisend, M. (2006). Two-sided advertising: A meta-analysis. International Journal of Research in Marketing, 23, 187–198. doi:10.1016/j.ijresmar.2005.11.001
Etgar, M., & Goodwin, S. A. (1982). One-sided versus two-sided comparative message appeals for new brand introductions. Journal Of Consumer Research, 8(4), 460–465.
doi:10.1086/208888
Festinger, L. (1957). A Theory of Cognitive Dissonance. Stanford, CA: Stanford University Press. Fransen, M. L., Smit, E. G., & Verlegh, P. W. J. (2015). Strategies and motives for resistance to
persuasion: An integrative framework. Frontiers in Psychology, 6, 1–12. doi:10.3389/fpsyg.2015.01201
Fransen, M. L., Verlegh, P. W. J., Kirmani, A., & Smit, E. G. (2015). A typology of consumer strategies for resisting advertising, and a review of mechanisms for countering them. International Journal of Advertising, 34(1), 6–16. doi:10.1080/02650487.2014.995284 Higgins, J. P., Tuttle, T. D., & Higgins, C. L. (2010). Energy beverages: Content and safety.
Mayo Clinic Proceedings, 85(11), 1033–1041. doi:10.4065/mcp.2010.0381 Hovland, C. I. (1954). Effects of the mass media of communication. In G. Lindzey (Ed.),
Handbook of Social Psychology (327–335). Cambridge, MA: Addison-Wesley Publishing Company.
Ivanov, B., Parker, K. A., & Dillingham, L. L. (2013). Measuring counterargument: A review and critique of the most popular techniques. The International Journal of Interdisciplinary Studies in Communication, 7(3), 59–74. doi:10.18848/2324-7320/CGP/v07i03/53578
Knowles, E., & Linn, J. A. (2004a). Approach-avoidance model of persuasion: Alpha and omega strategies for change. In E. Knowles & J.A. Linn (Eds.), Resistance and persuasion (pp. 117–148). Mahwah, NJ: Lawrence Erlbaum.
Knowles, E., & Linn, J. A. (2004b). The importance of resistance to persuasion. In E. Knowles & J.A. Linn (Eds.), Resistance and persuasion (pp. 3–9). Mahwah, NJ: Lawrence Erlbaum. Lang, B., Lee, C. K.-C., & Zwick, R. (1999). Message sidedness at the brand and product form
levels: Overcoming the shortcomings of two-sided messages? Advances in Consumer Research, 26(1), 485–490.
McGuire, W. J. (1964). Inducing resistance to persuasion: Some contemporary approaches. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 1, pp. 191–229). New York, NY: Academic Press.
Obermiller, C., & Spangenberg, E. R. (1998). Development of a scale to measure consumer skepticism toward advertising. Journal of Consumer Psychology, 7(2), 159–186. doi:10.1207/s15327663jcp0702_03
O’Keefe, D. J. (1999). How to handle opposing arguments in persuasive messages: A meta-analytic review of the effects of one-sided and two-sided messages. Communication Yearbook, 22(1), 209–249. doi:10.1080/23808985.1999.11678963
O’Keefe, D. J. (1993). The persuasive effects of message sidedness variations: A cautionary note concerning Allen's (1991) meta‐analysis. Western Journal of Communication, 57(1), 87– 97. doi:10.1080/10570319309374432
Red Bull energy drink. (2020). Retrieved April 22, 2020, from https://www.redbull.com/us-en/energydrink
Sawyer, A. G. (1973). The effects of repetition of refutational and supportive advertising appeals. Journal of Marketing Research, 10(1), 23–33. doi:10.2307/3149405
Shen, L., & Bigsby, E. (2013). The effects of message features: Content, structure, and style. In J. P. Dillard & L. Shen (Eds.), The SAGE handbook of persuasion: Developments in theory and practice (p. 20–35). Thousand Oaks, CA: Sage Publications, Inc.
Smith, S. M., Fabrigar, L. R., & Norris, M. E. (2008). Reflecting on six decades of selective exposure research: Progress, challenges, and opportunities. Social and Personality Psychology Compass, 2(1), 464–493. doi:10.1111/j.1751-9004.2007.00060.x
Tindale, C. W. (2004). Rhetorical argumentation: Principles of theory and practice. Thousand Oaks, CA: Sage Publications, Inc.
Zuwerink Jacks, J., & Cameron, K. A. (2003), Strategies for resisting persuasion. Basic and Applied Social Psychology, 25(2), 145–161.
Appendix A Manipulation Conditions Figure A1
Figure A2
