It did not mean anything (about me)

Abstract

Perpetrating romantic infidelity is discrepant with how most individuals see themselves and theoretically should produce cognitive dissonance. Accordingly, perpetrators of infidelity should experience symptoms of dissonance (e.g. self-concept discrepancy, psychological discomfort, poor affect) and employ tactics that reduce these symptoms (e.g. trivialization). These hypotheses were tested in four experiments. In each experiment, participants were given bogus feedback indicating that they had acted either faithfully or unfaithfully during a prior romantic relationship (this manipulation was evaluated in experiment 1). Participants who received unfaithful feedback reported higher levels of self-concept discrepancy, psychological discomfort, and poor affect (experiments 2 and 4) and trivialized to a greater extent the importance of their ostensive infidelities (experiments 3 and 4). Experiment 4 further showed that trivialization significantly reduced self-concept discrepancy and psychological discomfort but not poor affect. These results are generally consistent with the view that infidelity is a dissonance arousing behavior and that perpetrators of infidelity respond in ways that reduce cognitive dissonance.

Keywords

Close relationships cognitive dissonance theory infidelity perpetrators self-concept social psychology

Most people see themselves as being very loyal and faithful (Foster, Shrira, & Campbell, 2005) and consider romantic infidelity to be a morally unacceptable behavior (e.g. Gallup Poll, May 8–11, 2008). Logically speaking then, infidelity should be rare, but instead it is common (Hansen, 1987; Kinsey, Pomeroy, & Martin, 1948; Kinsey, Pomeroy, Martin, & Gebhard, 1953; Lawson & Samson, 1988; Thompson, 1983). Indeed, infidelity is so common that many of the same individuals who view it with disdain and consider themselves faithful must also contribute to its frequency.

Accordingly, many if not most perpetrators of infidelity should (at least sporadically) experience simultaneous conflicting cognitions surrounding their self-concepts and their behavior. That is, they consider themselves faithful and they despise infidelity, yet their behavior suggests just the opposite. These conflicting cognitions should theoretically give rise to the experience of cognitive dissonance (Festinger, 1957; see Harmon-Jones & Mills, 1999, for an overview). If true, then perpetrators of infidelity should report psychological experiences that are symptomatic of cognitive dissonance, such as self-concept discrepancy (Aronson, 1999; Aronson & Carlsmith, 1962), psychological discomfort, and generally poor affect – i.e. low positive affect and high negative affect (Devine, Monteith, Zuwerink, & Elliot, 1991; Devine, Tauer, Barron, Elliot, & Vance, 1999; Elliot & Devine, 1994; Harmon-Jones, 2000, 2001). They should also employ tactics that reduce cognitive dissonance, such as trivializing the importance of their discrepant behaviors (Simon, Greenberg, & Brehm, 1995).

In short, the ways that perpetrators of infidelity think and behave may be conceptualized in terms of a dissonance process: their infidelities conflict with their self-concepts causing cognitive dissonance, which they then reduce using dissonance reduction tactics. Investigating romantic infidelity using a cognitive dissonance framework has several practical and theoretical implications. Practically speaking, it will not be surprising to learn that perpetrator accounts of interpersonal transgressions, including infidelity, can influence the functioning of relationships (Schonbach, 1980). Romantic partners who, for example, minimize the importance of their infidelities to reduce cognitive dissonance may make themselves feel better, but at the same time make it impossible for their partners to forgive them, thus severely reducing the chances of reconciliation. Establishing whether cognitive dissonance plays a role in perpetrator responses to infidelity may encourage clinicians to develop methods that suppress the motivation to respond in ways that undermine the healing process.

In terms of our understanding of infidelity, this research is fairly unique in that is focuses on how infidelity affects perpetrators rather than victims. Exclusive focus on victims of infidelity in problematic for several reasons, including the fact that many relationships that experience infidelity endure. Therefore, it is important to understand how infidelity affects the thoughts and feelings of both partners. Only through research that examines how perpetrators are affected by infidelity will we attain a complete understanding of the impact that infidelity has on romantic partners and their relationships.

More generally, it is important to investigate the extent to which cognitive dissonance theory applies to real-world phenomena. As Festinger stated in his last public remarks on cognitive dissonance theory (as documented in Harmon-Jones & Mills, 1999), this type of research is needed to explore the extent to which dissonance affects the cognitions of people on an everyday basis. Although the research presented in this article will employ laboratory-based experimental methods, the phenomenon that they are being used to investigate (i.e. infidelity) is one that occurs frequently in the real world outside of the laboratory. Thus, we think our research contributes to Festinger’s mission of establishing the real-world implications of one of social psychology’s most well established, powerful, and relevant theories.

Connecting cognitive dissonance and infidelity

Individuals who see themselves as loyal and faithful, but who commit infidelity are likely to experience discrepancy involving the self-concept. This discrepancy has been theorized to be the root of cognitive dissonance (Aronson, 1999; Aronson & Carlsmith, 1962). That is, individuals experience cognitive dissonance because their cognitions targeting their behavior conflict with their cognitions targeting the self. Importantly, cognitive dissonance is often experienced as a state of psychological discomfort and poor affect (Devine et al., 1991; 1999; Elliot & Devine, 1994; Harmon-Jones, 2000, 2001). We should take a moment here to clarify that while Elliot and Devine’s (1994) seminal study found that cognitive dissonance only increased psychological discomfort and not poor affect, later research suggests that dissonance can also result in poor affect (Harmon-Jones, 2000; 2001). Thus, individuals who act in ways that conflict with their self-concepts are theoretically disposed to experiencing psychological discomfort and poor affect, each stemming from discrepancy involving the self-concept. Likewise, perpetrators of infidelity should similarly be disposed to feeling psychologically uncomfortable and affectively poor because of the self-concept discrepancy that arises from their behavior.

Fortunately (for those who commit infidelity), there are many ways to alleviate cognitive dissonance. For example, perpetrators may subjectively minimize the importance of their infidelities. In the present research, we focused on this particular strategy termed trivialization. Trivialization is an effective dissonance reduction mechanism in general (Simon et al., 1995) and may be an especially frequent and potent way to reduce cognitive dissonance stemming from infidelity. Various alternative mechanisms may not be available to perpetrators of infidelity. For example, it seems unlikely that individuals would change their beliefs regarding the importance of fidelity or how faithful they consider themselves to be. Although doing either may lessen cognitive dissonance, both would require perpetrators to change potentially long-standing and central beliefs, which research suggests are resistant to change even in the face of cognitive dissonance (Cooper & Mackie, 1983; Sherman & Gorkin, 1980; Simon et al., 1995). Trivializing the importance of specific infidelities, however, requires less fundamental change, if any, to self-relevant beliefs. Therefore, perpetrators may be especially inclined to trivialize their infidelities, and doing so may alleviate cognitive dissonance stemming from their infidelities.

In summary, we propose that cognitive dissonance plays a role in the thoughts and feelings of perpetrators of infidelity. Prior acts of infidelity are likely to promote inconsistencies involving the self-concept. We propose that these inconsistencies manifest as symptoms of cognitive dissonance including self-concept discrepancy, psychological discomfort, and poor affect and should promote the use of dissonance reduction tactics, such as trivialization. We test these specific predictions in the present research. First, however, we present two lines of evidence from the literature supporting the general link between infidelity and cognitive dissonance.

Evidence of infidelity–dissonance link from social transgression literature

Baumeister and colleagues’ (Baumeister, Stillwell, & Wotman, 1990; Stillwell & Baumeister, 1997) research on victim/perpetrator accounts of social transgressions may provide insight into the motivations that guide perpetrator reactions to infidelity. Perpetrators often describe their behaviors as justifiable, out of their control, and nondeliberate. They also tend to ascribe blame to their victims. Victims tend to describe transgressions in decidedly opposite terms. Other researchers have reported similar findings (Cameron, Ross, & Holmes, 2002; Kowalski, Walker, Wilkinson, Queen, & Sharpe, 2003). In a particularly relevant twist, Mikula, Athenstaedt, Heschgl, and Heimgartner (1998) asked spouses to describe an identical conflict. They observed that spouses who were more responsible for the conflicts tended to describe the conflicts as less serious and more justifiable. An even more relevant study collected narrative accounts of romantic infidelity. Perpetrators tended to describe their infidelities as less consequential than did victims (Shrira & Foster, 2005). This suggests that perpetrators of infidelity recall events similarly to perpetrators of other types of transgressions.

There are several possible explanations for these findings. We think it is likely, however, that desire for intrapersonal consistency influences perpetrator accounts of transgressions. Individuals tend to possess positive self-views and employ a variety of strategies to maintain them (Sedikides, Gaertner, & Toguchi, 2003). For most people, even perpetrators of social transgressions, a positive self includes being nice to others, or at least not harming others. For example, narcissistic individuals are more likely than others to commit infidelity (Buss & Shackelford, 1997), but nevertheless tend to describe themselves similarly to less narcissistic people in terms of communal traits (Bradlee & Emmons, 1992). In short, harming others for no good reason is likely to be inconsistent with most individuals’ self-concepts – even individuals who perpetrate social transgressions, such as infidelity. Deflecting blame and distorting history may thus serve to promote intrapersonal harmony between perpetrators’ self-concepts and their prior behaviors. To the extent that intrapersonal harmony is inversely associated with cognitive dissonance, distorted perpetrator accounts of past transgressions may be viewed as attempts to reduce cognitive dissonance.

Evidence of an infidelity–dissonance link from infidelity literature

When researchers (or clinicians or romantic partners) ask participants (or patients or romantic partners) why they committed infidelity, their responses are usually post hoc in nature. This has been a point of criticism (e.g. Drigotas, Safstrom, & Gentilia, 1999) because post hoc rationales for infidelity are likely influenced by self-serving motives. Individuals often have little insight into their behaviors and frequently rely upon socially acceptable heuristics to explain them (Nisbett & Wilson, 1977). It is therefore unreasonable to expect perpetrators to provide perfectly accurate explanations of their infidelities. Rather, their explanations will be influenced by social–cognitive factors including the motivation to maintain consistency between how they view themselves and what their behaviors suggest (i.e. minimize cognitive dissonance). That is, perpetrators with positive self-concepts should give reasons for their infidelities that reflect positively on the self.

Exemplifying this, Buunk (1987) discussed the “pushes” and “pulls” of infidelity. Pushes describe motivations to exit primary relationships, whereas pulls describe motivations to enter extra-pair relationships. Commonly cited pushes are low relationship satisfaction (Buss & Shackelford, 1997; Feldman & Cauffman, 1999a; Glass & Wright, 1977) and past partner betrayals (Buss & Shackelford, 1997; Feldman & Cauffman, 1999a; Greene, Lee, & Lustig, 1974; Mongeau, Hale, & Alles, 1994). Importantly, infidelities that stem from these pushes are generally looked upon more sympathetically by others (Feldman & Cauffman, 1999b). Citing these motives may therefore alleviate inconsistencies surrounding personal beliefs and behavior or at least justify them. This is not to suggest that infidelity never or even rarely occurs because of these pushes. We think it is likely, however, that some perpetrators are motivated to cite these pushes to alleviate conflict between their behaviors and their self-concepts.

A critical element of cognitive dissonance theory is that individuals must freely choose to act for dissonance to arise (Harmon-Jones & Mills, 1999). A commonly used experimental manipulation of cognitive dissonance involves conditions where participants are either required to do something counterattitudinal (e.g. writing an essay in favor of a position they disagree with) or given the choice to perform the counterattitudinal behavior (e.g. Cooper, Zanna, & Taves, 1978). Perceptions of free choice as well as cognitive dissonance effects are generally not observed in the condition where participants have no choice whether to perform the counterattitudinal behavior. In short, when individuals perceive their behaviors to have been out of their control, they are less likely to experience cognitive dissonance.

If cognitive dissonance is experienced by perpetrators of infidelity, one way for them to reduce this uncomfortable experience would be to recall the events surrounding their infidelities in such a way that subjectively minimizes the control they exerted over their behaviors. Indeed, research suggests that perpetrators frequently cite these types of causal factors, such as being intoxicated or otherwise lacking control over their behavior (Feldman & Cauffman, 1999a; Mongeau et al., 1994). Although it is certain that factors that undermine volition contribute to infidelity, we suspect that in some cases, perpetrators place more emphasis on these reasons because they remove personal responsibility and consequently reduce cognitive dissonance.

To summarize, situational factors that reflect positively on the self or diminish personal responsibility may be exaggerated to some degree by perpetrators of infidelity to alleviate negative intrapersonal consequences associated with infidelity, such as cognitive dissonance. Although these situational factors are not directly examined in the present research, they reflect the primary argument driving this research. That is, perpetrators of infidelity are likely to experience cognitive dissonance and employ cognitive dissonance reduction tactics.

Overview of the present research

The general purpose of the present research was to conduct an initial test of the hypothesis that perpetrating infidelity makes one susceptible to cognitive dissonance. Of course, we could not directly observe cognitive dissonance, so instead we assessed cognitive and affective responses that have been theoretically and empirically linked to cognitive dissonance. To this end, there were three specific hypotheses that guided this research. First, it was hypothesized that perpetrators of infidelity would experience a variety of empirically established symptoms of cognitive dissonance, including self-concept discrepancy (Aronson, 1999), psychological discomfort, and poor affect (i.e. low positive affect and high negative affect; Devine et al., 1999; Harmon-Jones, 2000, 2001). Second, perpetrators of infidelity were hypothesized to trivialize the importance of their infidelities, a tactic that has been empirically shown to alleviate symptoms of cognitive dissonance (Simon et al., 1995). Finally, consistent with the dissonance-reduction role of trivialization, it was hypothesized that perpetrators who trivialized their infidelities would effectively diminish their symptoms of cognitive dissonance. Each of these hypotheses was tested using a novel manipulation of perceived infidelity in which some participants were led to believe that they had been relatively unfaithful during a prior romantic relationship. This manipulation was evaluated in the first experiment and then used to test our hypotheses in three additional experiments.

Experiment 1

The first thing we needed to do was develop an effective manipulation of infidelity perceptions. One might question our decision to experimentally manipulate infidelity perceptions rather than assess actual infidelity either cross-sectionally or longitudinally. The experience of cognitive dissonance is uncomfortable, and individuals are motivated to alleviate its symptoms as quickly as possible. This fact makes cross-sectional and longitudinal research difficult because perpetrators of infidelity are likely to have taken measures to reduce dissonance prior to being measured. Therefore, if perpetrators of infidelity report no symptoms of dissonance, it is impossible to know whether they never experienced them or merely eliminated them.

In light of this, we created an experimental manipulation in which participants assigned to what we refer to as the “unfaithful condition” were given feedback suggesting that they had been unfaithful in a previous relationship. Participants assigned to the “faithful condition” were given feedback indicating the opposite. Readers may question why we did not manipulate infidelity perceptions during ongoing relationships. We wanted our manipulation to change perceptions of infidelity and remain uncontaminated by extraneous effects such as relationship maintenance motivations. The only way to do this was to remove the current relationship from the equation, and thus we had participants focus on a prior relationship. We think that this strategy results in a purer test of our cognitive dissonance hypotheses.

Participants in later experiments were asked questions designed to assess dissonance processes stemming from the experimental feedback. These questions would likely have biased manipulation check questions that assessed whether participants in the unfaithful condition actually felt as though they had been less faithful than participants in the faithful condition. We therefore conducted this first experiment as essentially a check of our experimental manipulation. Specifically, participants were asked to recall how faithful or unfaithful they felt after receiving the experimental feedback.

Method

Participants

Eighty undergraduates who all reported having at least one prior romantic relationship that lasted at least 3 months and was “at least somewhat serious” participated in this study. Eight participants were removed from this original sample because their responses to a survey component of the infidelity manipulation caused feedback to be given that was opposite of the condition to which they were randomly assigned (this will be explained in greater detail in the Results and Discussion section). Thus, the final sample included 72 participants (75% female; M age = 19.53 years, SD = 1.19 years).

Procedure

Participants were first reminded about how faithful they considered themselves to be (“infidelity reminder”) and then they were given bogus feedback suggesting that they had been either relatively faithful or unfaithful during a prior romantic relationship (“infidelity manipulation”). The dependent variables assessed after the infidelity manipulation varied by experiment. In the present experiment, the dependent variable was a measure of how faithful participants perceived themselves to be after receiving the experimental feedback (“faithfulness perception”)

Several filler surveys (e.g. political attitudes) were also intermixed throughout the procedure to disguise the focus of the study. At the conclusion of each of the four experiments, we debriefed participants and asked them to guess the purpose of the study. Many thought it had something to do with relationships and infidelity, but nobody reported suspicion that we were attempting to manipulate their fidelity perceptions. That is, nobody recognized the infidelity manipulation for what it was.

Materials

Fidelity reminder

Participants rated the extent to which eight words related to fidelity (i.e. faithful, loyal, truthful, honest, genuine, sincere, trustworthy, and dependable) described them (1 = not at all; 5 = completely). Responses to these descriptors were averaged to create a fidelity reminder score. Average scores were near the ceiling and are reported in Table 1.

Table 1.

Means comparison showing effect of infidelity manipulation on perceptions of faithfulness (experiment 1).

Experiment phase	Measure	Faithful	Unfaithful	t
Premanipulation	Fidelity reminder (.78)
	M	4.31	4.39	.91
	SD	.41	.37
Manipulation	Infidelity score (.84)
	M	10.16	11.88	1.68
	SD	4.52	4.15
Postmanipulation	Faithfulness perception
	M	7.32	3.91	9.51^***
	SD	1.53	1.51

Notes. ***p < .001; numbers in parentheses are Cronbach’s alphas; postmanipulation comparison remained significant after controlling for fidelity reminder and infidelity scores.

Infidelity manipulation

Participants thought about their most recent prior romantic relationship that lasted at least 3 months and was “at least somewhat serious” and recalled the person or persons whom they were most attracted to other than their primary romantic partners. They reported (a) the amount of time they spent thinking about these individuals, (b) the amount of flirting that occurred between them, and (c) how often they did “couple” things together, such as talking on the phone (1 = none; 7 = a great deal). These items were selected from a more extensive measure of infidelity developed by Drigotas et al. (1999) and were appropriate for the present manipulation because the severity of the infidelities assessed was ambiguous. Participants next summed their scores for the three items and wrote this score at the top of the next page of the questionnaire (average scores are reported in Table 1). This page contained a bogus description of the scale, indicating that it was designed by a distinguished researcher from a prestigious university.¹ The scale was purported to be a valid assessment of common forms of infidelity that had been administered to more than 50,000 students across the United States.

Approximately half of the participants (unfaithful condition, n = 34) were next informed that 83% of past participants scored lower than 6 on the scale. The remaining participants (faithful condition, n = 38) were told that 83% of past participants scored higher than 18. In fact, earlier pilot testing indicated that the vast majority of participants would report infidelity scores between 6 and 18. Therefore, most participants in the unfaithful condition should have perceived their infidelity scores to be higher than the vast majority of the population, with the opposite being true for participants in the faithful condition.

We should clarify here why we gave participants in what was effectively the control condition (i.e. faithful condition) feedback suggesting that they had scored much lower than average on the infidelity scale, as opposed to scoring at the average. Earlier pilot testing revealed that nearly all participants would come into the experiment seeing themselves as extremely faithful and loyal. This was confirmed in this experiment by the exceptionally high scores on the fidelity reminder. In light of this, we were concerned that feedback suggesting that participants were merely average in terms of prior infidelities would actually make participants feel unfaithful. In short, we gave participants in the faithful condition feedback suggesting that they had been particularly faithful in order to hopefully match their preexisting self-perceptions.

Finally, participants were told that a computer would be used to scan their responses into a database. To allegedly help a computer scanning device divide participants into groups depending on their scores on the scale, participants were asked to print very neatly and in large letters the words “UNFAITHFUL” or “FAITHFUL” in a box depending on whether their scores were above or below what the majority of past participants reported.

Faithfulness perception

Following the infidelity manipulation, participants completed a series of filler surveys (e.g. political attitudes) and a single-item assessment of how faithful they felt after they completed the infidelity manipulation. At this point, participants were still under the impression that the feedback they received was valid. They were specifically asked to recall how they felt immediately after learning what their scores meant on the measure of infidelity. They responded on a 9-point Likert-type scale (1 = extremely unfaithful; 9 = extremely faithful).

Results and discussion

Exclusion of participants

As noted in the Participants section, eight participants were excluded from the original sample. Six participants were assigned to the unfaithful condition but reported infidelity scores lower than six causing them to receive feedback indicating that they had been faithful. Two participants were assigned to the faithful condition but reported infidelity scores higher than 18 causing them to receive feedback indicating that they had been unfaithful. Even with these exclusions, the two experimental conditions did not differ significantly in terms of their average fidelity reminder or infidelity scores (see Table 1). In the four experiments, only once was a difference observed after excluding these nonconforming participants.² In experiment 3, participants in the unfaithful condition reported very slightly but significantly lower average scores on the fidelity reminder (see Table 3 and note that both groups scored very high on the measure). Out of an abundance of caution, we reran all analyses after controlling for infidelity scores and scores on the fidelity reminder. Statistically controlling for these variables did not alter any of the results reported in this article, and thus we report the results attained without the use of statistical controls.

Table 2.

Means comparisons showing effect of infidelity manipulation on self-concept discrepancy, psychological discomfort, and affect (experiment 2).

Experiment phase	Measure	Faithful	Unfaithful	t
Premanipulation	Fidelity reminder (.82)
	M	4.38	4.22	1.88
	SD	.46	.42
Manipulation	Infidelity score (.81)
	M	10.17	11.55	1.67
	SD	4.10	3.89
Postmanipulation	Self-concept discrepancy
	M	2.85	3.77	2.82**
	SD	1.49	1.65
	Psychological discomfort (.81)
	M	2.05	2.83	2.80**
	SD	1.10	1.54
	Negative affect (.62)
	M	2.22	3.01	2.90**
	SD	1.24	1.40
	Positive affect (.90)
	M	5.04	4.26	2.78**
	SD	1.27	1.46

Notes. **p < .01; numbers in parentheses are Cronbach’s alphas; all postmanipulation comparisons remained significant after controlling for fidelity reminder and infidelity scores.

Table 3.

Means comparison showing effect of infidelity manipulation on importance ratings of ostensive infidelities (experiment 3).

Experiment phase	Measure	Faithful	Unfaithful	t
Premanipulation	Fidelity reminder (.82)
	M	4.50	4.27	2.11*
	SD	.41	.43
Manipulation	Infidelity score (.90)
	M	11.65	10.46	.94
	SD	5.02	4.37
Postmanipulation	Importance (.94)
	M	5.42	2.90	7.33***
	SD	1.33	1.25

Notes. *p < .05, ***p < .001; numbers in parentheses are Cronbach’s alphas; postmanipulation comparison remained significant after controlling for fidelity reminder and infidelity scores.

Hypothesis testing

The results from this experiment are reported in Table 1. Participants in the unfaithful condition reported feeling significantly less faithful than participants in the faithful condition. This difference was very large by conventional standards (d = 2.24). Furthermore, the mean score for the unfaithful condition was nearly a full standard deviation lower than the scale midpoint, whereas the mean score for the faithful condition was more than a full standard deviation above the scale midpoint. We also examined participants’ gender and found that it did not moderate the effect of the manipulation.

A reviewer of this article noted that participants in this experiment might have simply acquiesced to the demands of the experiment. We cannot completely rule this out; however, we think there are at least three reasons to doubt that demand characteristics fully explained the effect. First, participants did not report during debriefing sessions that they thought the purpose of the study was to change self-perceptions regarding infidelity. Therefore, participants did not appear to perceive experimental demands. Second, our interactions with participants during debriefing sessions suggest that most thought the experimental feedback was legitimate. Indeed, many participants in the unfaithful condition expressed significant relief when informed that the feedback was bogus. Again, participants did not appear to perceive that the feedback was designed to change their self-perceptions. Finally, although demand characteristics, to the extent that they existed, would have biased the results in the direction of supporting the validity of our experimental manipulation, opposing psychological forces, such as reactance (Brehm, 1966), would have done the opposite. In short, some participants in the faithful condition might have reported feeling less faithful because they thought this is what was expected from them, however, other more reactive participants might have done just the opposite and reported that they felt more faithful. In light of all this, while we cannot rule out the possibility that some participants were responding according to the demands of the experiment, we are doubtful that this effect accounts for the experimental manipulation’s very large effect.

In sum, participants who received the experimental feedback indicating that they had been unfaithful reported that they felt unfaithful in both a relative and absolute sense. We used this experimental manipulation in the next three experiments to investigate whether perceptions of fidelity caused symptoms and responses consistent with cognitive dissonance.

Experiment 2

The purpose of this experiment was to test whether thinking about prior acts of infidelity promotes symptoms associated with cognitive dissonance. The most commonly used marker of cognitive dissonance is attitude change in the direction of discrepant behavior. This does not occur, however, when attitudes are very important or central to the self-concept (Devine, Froning, & Elliot, 1995; Devine et al., 1999; Sherman & Gorkin, 1980). Specific attitudes toward infidelity are connected to more general and basic human values, such as loyalty and honesty, which cross-cultural research suggests are universally important (Schwartz, 1994; Schwartz & Bilsky, 1990; Schwartz et al., 2012). Because attitudes toward infidelity are anchored in these longstanding basic values, we were doubtful that cognitive dissonance induced by a laboratory manipulation would cause them to noticeably change. To be clear, we do not think it is impossible that attitudes toward infidelity might shift in response to cognitive dissonance. Indeed, we suspect that this happens in the real world. However, we did not think assessing attitude change would provide a sensitive enough test of our prevailing hypotheses.

Fortunately, there are a variety of other markers of cognitive dissonance that do not involve attitude change. In the present study, we assessed three markers of cognitive dissonance: psychological discomfort, poor affect (i.e. high negative and low positive affect), and self-concept discrepancy. A number of studies have shown that cognitive dissonance manipulations cause aversive psychological states in participants (Devine et al., 1991, 1999; Elliot & Devine, 1994; Harmon-Jones, 2000, 2001). For example, participants who behave counterattitudinally generally report heightened states of psychological discomfort and poor affect. Furthermore, Aronson (1999) suggests that cognitive dissonance results from intrapsychic dilemmas involving behavior (e.g. infidelity) and the self-concept. Counterattitudinal behaviors are aversive because they suggest something about the self that is at odds with the self-concept. Therefore, individuals who commit counterattitudinal behaviors should report heightened levels of self-concept discrepancy, psychological discomfort, and poor affect. We tested these predictions in the present study by manipulating perceptions of prior infidelity and measuring psychological discomfort, affect, and self-concept discrepancy.

Method

Participants

After excluding participants for reasons explained in experiment 1, data were analyzed from 93 undergraduates who all reported having at least one prior romantic relationship that lasted at least 3 months and was “at least somewhat serious” (72% female; M age = 19.02 years, SD = 1.04 years).

Materials and procedure

Participants completed the fidelity reminder and the infidelity manipulation described in experiment 1 (Ns = 47 and 46 for unfaithful and faithful conditions, respectively; see Table 2 for average fidelity reminder and infidelity scores) and then completed measures of self-concept discrepancy, psychological discomfort, and affect (negative and positive). Each of these measures is described below.

Self-concept discrepancy

This measure was created by the authors and modeled after pictorial assessments that use overlapping circles to depict similarity or closeness (e.g. Inclusion of Other in the Self Scale; Aron, Aron, & Smollan, 1992). Participants were asked to consider how they felt about themselves typically versus at the moment. They were presented with seven pairs of overlapping ovals. The two ovals in each pair were labeled “typical” and “right now.” The ovals at the top of the scale did not overlap at all (scored as a 7), whereas the ovals at the bottom of the scale completely overlapped (scored as a 1). The sets of ovals in between represented varying degrees of overlap. Participants placed a mark next to the pair of ovals that best described how they felt at that moment. Larger differences between the “typical” and “right now” ovals were operationally defined as greater discrepancy involving the self-concept. Although technically, higher self-concept discrepancy is just that, we did observe that scores on this measure were significantly correlated with higher negative affect (r = .58, p < .001), psychological discomfort (r = .42, p < .001), and lower positive affect (r = −.47, p < .001; these measures are described below). Therefore, it appears that most participants who reported higher self-concept discrepancy were feeling worse about themselves than usual.

Psychological discomfort and affect

Participants completed Elliot and Devine’s (1994) measures of psychological discomfort, negative affect, and positive affect. Psychological discomfort was assessed by reporting how uncomfortable, uneasy, and bothered participants felt at the moment (1 = does not apply at all; 7 = applies very much). Positive affect was assessed using the words good, happy, optimistic, and friendly. Negative affect was assessed using the words disappointed, guilty, and annoyed. Responses to individual items were averaged for each scale.

Results and discussion

We predicted that participants in the unfaithful condition would report higher self-concept discrepancy, psychological discomfort, and poorer affect than participants in the faithful condition. All of these predictions were supported and are shown in Table 2. In short, participants who were led to believe that they had been unfaithful reported symptoms associated with cognitive dissonance. We also tested whether gender moderated the effect of the infidelity manipulation on any of the dependent variables. We did find evidence of one significant interaction. Whereas the infidelity manipulation had a strong effect on female reports of negative affect (i.e. they reported feeling more negative in the unfaithful condition), it did not increase negative affect in males. Although we do not wish to minimize the importance of this specific gender interaction, it is worth stressing that men and women were in general similarly affected by the infidelity manipulation.

Experiment 3

Experiment 2’s results were consistent with the proposition that perpetrators of infidelity experience cognitive dissonance. One tactic used to reduce dissonance is called trivialization: minimizing the subjective importance of a discrepant behavior (Simon et al., 1995). Trivialization is especially useful in situations where attitude/belief change is not feasible. Personal beliefs regarding infidelity are likely to be resistant to change because they are important and central to most people. So although perpetrators are unlikely to alter their self-beliefs regarding infidelity, they may nevertheless trivialize their infidelities to reduce consequent dissonance. In the present experiment, we experimentally manipulated perceptions of infidelity and then gave participants the opportunity to trivialize their ostensive infidelities. We specifically predicted that participants who were led to believe that they had acted less faithfully in a previous relationship would trivialize to a greater extent the importance of their ostensive infidelities.

Method

Participants

After excluding participants for reasons explained in experiment 1, data were analyzed from 57 undergraduates who all reported having at least one prior romantic relationship that lasted at least 3 months and was “at least somewhat serious” (82% female; M age = 18.96 years, SD = 1.00 years).

Materials and procedure

Participants completed the fidelity reminder and infidelity manipulation described in experiment 1 (Ns = 26 and 31 for unfaithful and faithful conditions, respectively; average fidelity reminder and infidelity scores are reported in Table 3) and then completed a measure of trivialization. We assessed trivialization using a measure that was similar to that used by Simon et al. (1995). We changed their measure to reflect what participants did in the present experiment (in Simon et al.’s study, participants wrote counterattitudinal essays). Specifically, participants were asked (1) “how important is the extent to which you committed these behaviors in terms of describing you,” (2) “how meaningful is the extent to which you committed these behaviors in terms of describing you,” (3) how much can one infer about you from the extent to which you committed these behaviors,” and (4) “how much does the extent to which you committed these behaviors suggest about you.” Participants responded on Likert scales (1 = not at all; 7 = a whole lot). Each participant received a single score representing their average responses to the four questions. Lower scores indicated more extensive trivialization.

Results and discussion

We predicted that participants in the unfaithful condition would report that their behaviors were less important (i.e. higher trivialization) than participants in the faithful condition. The results offered unambiguous support for this hypothesis (see Table 3). When participants received feedback that they had been unfaithful, they responded by downplaying the importance of their behaviors. This is consistent with prior research on cognitive dissonance and trivialization (Simon et al., 1995) and suggests that perpetrators of infidelity may respond to their belief-discrepant behaviors by minimizing the importance of them. We also tested whether gender moderated the effect of the infidelity manipulation on trivialization but observed no evidence of this.

Experiment 4

Altering perceptions of infidelity produces a variety of experiences and responses associated with cognitive dissonance, including self-concept discrepancy, psychological discomfort, poor affect (experiment 2) and trivialization (experiment 3). With regard to trivialization, it may be assumed that individuals who trivialize their ostensive infidelities experience relief from the symptoms of dissonance, but this has not as of yet been demonstrated empirically. Simon et al. (1995) showed that individuals who trivialize counterattitudinal behaviors are less likely to change their attitudes in the direction of a counterattitudinal position, suggesting that trivialization operates as a dissonance-reduction mechanism. If so, then participants in experiment 3, who were given feedback suggesting that they had acted unfaithfully, may have reduced the severity of their dissonance symptoms by trivializing their ostensive infidelities. The purpose of the present experiment was to test this empirically. Perceptions of infidelity were again manipulated, but this time participants in the unfaithful condition were further randomly assigned to one of two conditions where they were given the opportunity to trivialize their ostensive infidelities either before or after the dissonance symptoms were assessed. If trivialization serves a dissonance reduction function in the context of infidelity, then “unfaithful” participants allowed to trivialize their ostensive infidelities before dissonance assessment should report less self-concept discrepancy, less psychological discomfort, and improved affect relative to participants only allowed to trivialize their ostensive infidelities after dissonance assessment.

Method

Participants

After excluding participants for reasons explained in experiment 1, data were analyzed from 121 undergraduates who all reported having at least one prior romantic relationship that lasted at least 3 months and was “at least somewhat serious” (55% female; M age = 19.36 years, SD = 1.47 years).

Materials and procedure

After completing the fidelity reminder and infidelity manipulation (average fidelity reminder and infidelity scores are reported in Table 4), participants in the faithful condition (n = 39) were next administered the dissonance measures (i.e. self-concept discrepancy, psychological discomfort, affect) and then finished with the trivialization measure. Participants in the unfaithful condition (n = 82) were randomly assigned to two additional conditions. Participants assigned to the “unfaithful–no trivialization” condition (n = 40) followed the same procedure as participants in the faithful condition (i.e. they completed the trivialization measure after the dissonance measures). Participants in the “unfaithful–trivialization” condition (n = 42), however, completed the trivialization measure prior to the dissonance measures. This procedure made it possible to internally replicate experiments 2 and 3 as well as test whether trivialization alleviated dissonance symptoms.

Table 4.

Means comparisons showing effects of infidelity manipulation and opportunity to trivialize on self-concept discrepancy, psychological discomfort, and affect (experiment 4).

			Unfaithful
Experiment phase	Measure	Faithful	No trivialization	Trivialization	F _omnibus
Premanipulation	Fidelity reminder (.82)
	M	4.22	4.25	4.31	.37
	SD	.46	.44	.53
Manipulation	Infidelity score (.85)
	M	11.56	13.00	12.07	1.11
	SD	4.45	4.36	4.28
Postmanipulation	Self-concept discrepancy
	M	2.33_a	4.20_b	3.31_c	14.20***
	SD	1.18	1.79	1.63
	Psychological discomfort (.83)
	M	2.22_a	3.37_b	2.64_a	6.07**
	SD	1.27	1.67	1.47
	Negative affect (.67)
	M	2.07_a	3.37_b	3.05_b	8.72***
	SD	1.12	1.64	1.50
	Positive affect (.89)
	M	5.09_a	4.24_b	4.05_b	6.08**
	SD	.92	1.61	1.60

Notes. **p < .01, ***p < .001; numbers in parentheses are Cronbach’s alphas; within row means with different subscripts differ significantly (p < .05) when tested with t tests; All postmanipulation comparisons remained significant after controlling for fidelity reminder and infidelity scores.

Results and discussion

Replications of experiments 2 and 3

Table 4 shows that the results from experiment 2 were successfully replicated in the present experiment. Participants in the unfaithful–no trivialization condition reported higher levels of self-concept discrepancy, psychological discomfort, and poorer affect compared to participants in the faithful condition. The results from experiment 3 were also successfully replicated. Participants in the unfaithful–no trivialization condition reported that their ostensive infidelities were less important (M = 3.01, SD = 1.31) compared to participants in the faithful condition (M = 4.17, SD = 1.35), t(77) = 3.90, p < .001.

Did trivialization alleviate the symptoms of dissonance?

We should first note that participants in the unfaithful–trivialization condition reported that their ostensive infidelities were equally unimportant (M = 3.11, SD = 1.32) when compared to participants in the unfaithful–no trivialization condition, t < 1. So the results were not confounded by the extent to which participants in the two unfaithful conditions utilized trivialization.

As reported in Table 4, we observed mixed support for the hypothesis that trivialization alleviated dissonance symptoms. Specifically, “unfaithful” participants who practiced trivialization before completing the dissonance measures reported significantly lower levels of self-concept discrepancy and psychological discomfort compared to “unfaithful” participants who were not given this opportunity to trivialize their ostensive infidelities. Indeed, psychological discomfort following trivialization fell to levels equivalent to that reported by participants in the faithful condition. Self-concept discrepancy remained elevated in participants who trivialized compared to participants in the faithful condition. This suggests that, at least in the present experiment, trivialization did not fully eliminate self-concept discrepancy. We consider the most important comparison, in terms of determining whether trivialization reduced dissonance, to be between the two unfaithful conditions. Consider if this had been a medical experiment and we observed that participants with a disease who received an experimental drug experienced fewer disease symptoms compared to participants with the disease who received a placebo but more symptoms than a healthy control group of participants. Even though the experimental drug did not completely eliminate the disease symptoms, there is nevertheless evidence that it reduced them. Similarly, although trivialization did not completely eliminate self-concept discrepancy, there is evidence that it reduced it. We consider this result to be supportive of our contention that trivialization reduced cognitive dissonance in participants led to believe that they had acted unfaithfully.

Interestingly, trivialization did not alleviate negative affect nor did it improve positive affect. This may suggest that trivialization targeted the discrepancy and discomfort components of cognitive dissonance but not the affective components. In some respects this finding is consistent with early research showing that cognitive dissonance alters discomfort but not affect (Elliot & Devine, 1994). While it is, in our opinions, highly unlikely that perpetrating infidelity would increase psychological discomfort without also creating poor affect, it is theoretically possible that dissonance reduction tactics, such as trivialization, might specifically counter the resulting discomfort, which is theoretically brought about by self-concept discrepancy. The dissociability of psychological discomfort and affect as they occur during the cognitive dissonance process is an area that in general needs further empirical testing and refinement.

Alternatively, it is possible that infidelity promotes changes in affect that are independent from cognitive dissonance. For example, individuals who perpetrate infidelity may experience cognitive dissonance stemming from discrepancies between their behavior (i.e. infidelities) and self-concepts. However, these same individuals may also experience poor affect, for example, stemming from the fact that their partners were upset about what had happened. Although trivialization might alleviate perceived discrepancies involving the self-concept, the fact that their partners were upset remains. To illustrate this using a more extreme thought experiment, imagine that a man named John gets into a car crash that kills an entire family. Will John experience cognitive dissonance as a result of this terrible tragedy? If John caused the accident because he was speeding and sending a text message, then he will probably experience significant cognitive dissonance because the inference that can be drawn from his behavior (i.e. he is a thoughtless and careless person) most likely conflicts with how he normally views himself. On the other hand, if the person driving the other car was the one responsible for the accident, John will probably experience very little cognitive dissonance. So the situational manipulation in this thought experiment has clear implications for the experience of cognitive dissonance. With regard to affect, however, we would expect to see very poor affect in either situation. An entire family was killed and this is a great tragedy and will likely traumatize John regardless of who caused the accident.

Of course, committing infidelity does not compare to killing an entire family, but the point remains that behaviors and events that have the power to create cognitive dissonance can also have the power to create poor affect, and these two consequences are capable of being quite independent of one another. As this relates to the present research, trivializing the importance of infidelities might alleviate self-concept discrepancy but have no effect on affective states triggered by non-dissonant cognitions (e.g. thoughts about pain experienced by partners, thoughts about infidelity in general). This could create a situation similar to what was observed in the present study. We should acknowledge, however, that our account is highly speculative and will require further testing before any conclusions can be drawn.

Gender effects

We tested whether gender moderated the effect of the infidelity manipulation on any of the dependent variables, including whether it moderated the effect of trivialization on self-concept discrepancy or psychological discomfort. We only uncovered two significant gender interactions. Female participants relative to males were more strongly affected by the infidelity manipulation in terms of negative affect and psychological discomfort. Thus, negative affect was significantly moderated by gender in both experiments 2 and 4. It appears that females feel affective worse than do males when made to feel unfaithful. Although this gender interaction was not specifically predicted, it appears to be reliable. Like we stated in experiment 2, however, we think the overall similarity between men and women in these studies outweighs the very few differences that were observed. Nevertheless, we would encourage research that examines possible roles that gender plays in these processes.

General discussion

The purpose of this research was to conduct an initial test of the hypothesis that perpetrators of infidelity are susceptible to cognitive dissonance. To test this hypothesis, we employed a novel experimental manipulation that was designed to make participants think that they had acted either faithfully or unfaithfully during a prior relationship. Consistent with our hypothesis, participants who were manipulated into thinking that they had acted unfaithfully during a prior relationships reported higher levels of self-concept discrepancy, psychological discomfort, and poor affect. These “unfaithful” participants responded by trivializing the importance of their ostensive infidelities, and this resulted in decreased self-concept discrepancy and psychological discomfort. Although we are not in a position to declare with absoluteness that cognitive dissonance is the only or best explanation for our results, we think it is likely that we were observing cognitive dissonance in action.

The studies reported in this article were entirely experimentally based. This is fairly unique within the field of relationship research, which mostly relies upon correlational methods. Correlational studies are certainly useful (most of our published research is correlational in nature), but they lack the ability to establish causal mechanisms. Thus, a real benefit of the present research is that, not only did it present evidence of whether infidelity is related to cognitive dissonance, it presented evidence that perceptions of infidelity cause cognitive dissonance. We hope that our experimental approach to this research will inspire other relationship researchers to similarly adopt experimental methods to test their research hypotheses.

Although we think our experimental approach offers more in the way of benefits than costs, an important question that remains unanswered is whether what we observed using an experimental manipulation of infidelity perceptions reflects what happens when individuals perpetrate actual infidelity. As we noted in experiment 1, we thought experimentally manipulating infidelity perceptions gave us the best opportunity to observe cognitive dissonance in the laboratory. Longitudinal and cross-sectional investigations of actual infidelity were, in our opinions, likely to be confounded by perpetrators reducing the symptoms of dissonance before they could be measured. Nevertheless, we would certainly encourage others to develop methods that permit valid testing of this hypothesized phenomenon in the real world.

One interesting possible future study would be to measure trivialization longitudinally and assess whether it changes post-infidelity. We have examined trivialization following infidelity cross-sectionally and found that participants who report having committed more extensive infidelities generally report that their infidelities are less important (Foster, 2005). This finding is consistent with our dissonance hypothesis, but it is possibly confounded by preexisting views of infidelity. That is, individuals who view infidelity as a trivial matter are probably more likely to commit infidelity. This confound could possibly be remedied using longitudinal tests.

In experiments 2 and 3, we noted that we specifically selected dependent variables that were most likely to be affected by dissonance stemming from infidelity and, more specifically, a laboratory-based manipulation of infidelity perceptions. In short, we tried to maximize the probability that if perpetrating infidelity triggers dissonance we would observe it. Thus, it is probable that the effect sizes we observed in these experiments are larger than what occurs in the real world (at least with the types of infidelities assessed). Additionally, we do not think the effects of perpetrating infidelity are limited to the dependent variables measured in these experiments. For example, as discussed in the introduction, according to cognitive dissonance theory, individuals must view their behaviors as freely chosen for dissonance to occur. Therefore, cognitions that reduce perceptions of volition may diminish cognitive dissonance in the context of infidelity. Perpetrators may thus exaggerate the influence of external forces (e.g. consumption of intoxicants) that essentially put their infidelities out of their control. To be clear, we recognize that external forces do play roles in dissonant behaviors, including infidelity. Thus, we are not suggesting, for example, that perpetrators who claim to have been too drunk to make effective decisions are always lying. Nevertheless, it is conceivable that perpetrators are motivated by cognitive dissonance processes to cite external causes of their discrepant behavior.

In a related vein, our infidelity manipulation assessed relatively mild forms of infidelity. This probably played a role in the exceptionally large effect of the manipulation on trivialization. If we had assessed more serious forms of infidelity (e.g. sexual infidelity), it is questionable whether participants would have so easily trivialized their importance. In this case, perpetrators may be forced to rely on other forms of dissonance reduction, such as diminishment of free-choice. As we mention in experiment 1, we specifically chose to assess mild (and more ambiguous) forms of infidelity because this made it easier to manipulate feedback. Certainly, developing manipulations that use more serious forms of infidelity would be useful to establishing whether and what type of dissonance reduction routes are utilized.

It is important to note that not all relationships end after they experience infidelity. There are relationships that survive infidelity and function quite well in the months and years that follow the transgression. The thoughts and actions of perpetrators are at least partially, and perhaps largely responsible for whether relationships do well or falter following acts of infidelity (Schonbach, 1980). Therefore, understanding why and how perpetrators of infidelity think and behave following their transgressions is important to the study of relationships that are touched by infidelity. Based on the results of the present research, perpetrators of infidelity seem quick to trivialize their behaviors if given the chance. This seems more likely to harm rather than help relationship functioning following infidelity. It may communicate a lack of appreciation for the severity of the offense. Of course, there are other ways that perpetrators of infidelity may temper dissonance resulting from infidelity (see above discussion). One intriguing possibility is that some perpetrators of infidelity, who experience dissonance, may behave more faithfully in the future as a way to reduce dissonance stemming from the infidelity. Aronson and colleagues (Aronson, Fried, & Stone, 1991; Dickerson, Thibodeau, Aronson, & Miller, 1992; Stone, Aronson, Crain, & Winslow, 1994; Thibodeau & Aronson, 1992) have shown impressively that one way to reduce cognitive dissonance is to behave in ways that are consistent with one’s self-concept. For example, individuals who believe that water conservation is important, but who are reminded that they sometimes fail to conserve water, often conserve water more frequently in the future to reduce the dissonance stemming from the salient behavior/belief discrepancy. In the context of infidelity, perpetrators of infidelity, who experience dissonance resulting from the discrepancy between their behavior and their self-concepts, may act faithfully in the future to bring their behavior back in line with their self-relevant beliefs. An interesting test of this idea would be to employ Aronson’s hypocrisy manipulation (e.g. getting participants to advocate a pro-fidelity stance, then remind them about previous incidents when they did not live up to their standards). Doing so may promote faithful behavior in the future.

Although research on hypocrisy and behavior change suggests that perpetrators of infidelity may become especially faithful following the experience of cognitive dissonance, additional lines of research suggest that just the opposite may occur. We decided to investigate trivialization in the present research in part because we thought that other dissonance reduction tactics, such as attitude change, would be less likely to be used by perpetrators of infidelity. Research suggests that attitude change in response to cognitive dissonance is unlikely when attitudes are central to the self (e.g. Cooper & Mackie, 1983). We suspect that attitudes toward infidelity fall into this central-to-the-self category. Nevertheless, research should be conducted to determine whether attitude change by perpetrators of infidelity represents a viable dissonance reduction strategy (perhaps, for example, when more severe forms of infidelity have been committed). Consider that if perpetrators change their attitudes toward infidelity in response to cognitive dissonance, then the change is likely to be in the direction of possessing more positive/less negative attitudes toward infidelity. Thus, one would predict that individuals who commit infidelity, experience cognitive dissonance, and reduce their dissonance via attitude change will become even more likely to commit infidelity in the future. Certainly, the feasibility and consequences of attitude change by perpetrators of infidelity merit further research.

Clearly, there are a number of potential ways that perpetrators of infidelity may deal with their cognitive dissonance. Some of them are more positive than others. A challenge to infidelity researchers is to identify why some perpetrators select some dissonance reduction mechanisms but not others. Do perpetrators simply select whichever mechanism is available first (e.g. what comes to mind first)? The present research suggests that they will generally take the opportunity to trivialize if it is offered, so maybe they are purely opportunists. However, are some mechanisms more convenient than others? Are some mechanisms not possible? Although participants trivialized in the present studies, what would happen if more serious types of infidelity (e.g. sexual infidelity) were assessed? Would they be able to trivialize in these situations? Would they select mechanisms more suitable for serious discrepancies—perhaps behaving very faithfully in the future or perhaps adopting more positive attitudes toward infidelity? Do perpetrators use multiple mechanisms simultaneously? All of these are questions that can possibly be answered empirically. Some will be more difficult to answer than others, but all will contribute to our understanding of the thoughts and actions of those who perpetrate infidelity.

Footnotes

Authors’ note

Portions of this paper were used as a doctoral dissertation completed by the first author (JDF) at the University of Georgia under the direction of Keith Campbell (chair), Mike Kernis, Lenny Martin, and Dave Shaffer.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Aron

E. N.

Smollan

(1992). Inclusion of other in the self scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63, 596–612.

Aronson

(1999). Dissonance, hypocrisy, and the self-concept. In Harmon-Jones

Mills

(Eds.), Cognitive dissonance: Progress on a pivotal theory in social psychology (pp. 103–126). Washington, DC: American Psychological Association.

Aronson

Carlsmith

J. M.

(1962). Performance expectancy as a determinant of actual performance. The Journal of Abnormal and Social Psychology, 65, 178–182.

Aronson

Fried

Stone

(1991). Overcoming denial and increasing the intention to use condoms through the induction of hypocrisy. American Journal of Public Health, 81, 1636–1638.

Baumeister

R. F.

Stillwell

Wotman

S. R.

(1990). Victim and perpetrator accounts of interpersonal conflict: Autobiographical narratives about anger. Journal of Personality and Social Psychology, 59, 994–1005.

Bradlee

P. M.

Emmons

R. A.

(1992). Locating narcissism within the interpersonal circumplex and the five-factor model. Personality and Individual Differences, 13, 821–830.

Brehm

J. W.

(1966). A theory of psychological reactance. Oxford, England: Academic Press.

Buss

D. M.

Shackelford

T. K.

(1997). Susceptibility to infidelity in the first year of marriage. Journal of Research in Personality, 31, 193–221.

Buunk

(1987). Conditions that promote breakups as a consequence of extradyadic involvements. Journal of Social & Clinical Psychology, 5, 271–284.

10.

Cameron

J. J.

Ross

Holmes

J. G.

(2002). Loving the one you hurt: Positive effects of recounting a transgression against an intimate partner. Journal of Experimental Social Psychology, 38, 307–314.

11.

Cooper

Mackie

(1983). Cognitive dissonance in an intergroup context. Journal of Personality and Social Psychology, 44, 536–544. doi: 10.1037/0022-3514.44.3.536

12.

Cooper

Zanna

M. P.

Taves

P. A.

(1978). Arousal as a necessary condition for attitude change following induced compliance. Journal of Personality and Social Psychology, 36, 1101–1106. doi: 10.1037/0022-3514.36.10.1101

13.

Devine

P. G.

Froning

Elliot

A. J.

(1995). [Attitude importance, dissonance, and resistance to attitude change]. Unpublished raw data.

14.

Devine

P. G.

Monteith

M. J.

Zuwerink

J. R.

Elliot

A. J.

(1991). Prejudice with and without compunction. Journal of Personality and Social Psychology, 60, 817–830.

15.

Devine

P. G.

Tauer

J. M.

Barron

K. E.

Elliot

A. J.

Vance

K. M.

(1999). Moving beyond attitude change in the study of dissonance-related processes. In Harmon-Jones

Mills

(Eds.), Cognitive dissonance: Progress on a pivotal theory in social psychology (pp. 297–323). Washington, DC: American Psychological Association.

16.

Dickerson

C. A.

Thibodeau

Aronson

Miller

(1992). Using cognitive dissonance to encourage water conservation. Journal of Applied Social Psychology, 22, 841–854.

17.

Drigotas

S. M.

Safstrom

C. A.

Gentilia

(1999). An investment model prediction of dating infidelity. Journal of Personality and Social Psychology, 77, 509–524.

18.

Elliot

A. J.

Devine

P. G.

(1994). On the motivational nature of cognitive dissonance: Dissonance as psychological discomfort. Journal of Personality and Social Psychology, 67, 382–394.

19.

Feldman

S. S.

Cauffman

(1999a). Sexual betrayal among late adolescents: Perspectives of the perpetrator and the aggrieved. Journal of Youth and Adolescence, 28, 235–258.

20.

Feldman

S. S.

Cauffman

(1999b). Your cheatin' heart: Attitudes, behaviors, and correlates of sexual betrayal in late adolescents. Journal of Research on Adolescence, 9, 227–252.

21.

Festinger

(1957). A theory of cognitive dissonance. Oxford, England: Row, Peterson.

22.

Foster

J. D.

(2005). [Correlation between trivialization and infidelity severity]. University of Georgia. Unpublished raw data.

23.

Foster

J. D.

Shrira

Campbell

W. K.

(2005). When behavior does not match our beliefs: The case of infidelity. Paper presented at the Poster presented at the annual meeting of the Society for Personality and Social Psychology, New Orleans, LA.

24.

Glass

S. P.

Wright

T. L.

(1977). The relationship of extramarital sex, length of marriage, and sex differences on marital satisfaction and romanticism: Athanasiou's data reanalyzed. Journal of Marriage & the Family, 39, 691–703.

25.

Greene

B. L.

Lee

R. R.

Lustig

(1974). Conscious and unconscious factors in marital infidelity. Medical Aspects of Human Sexuality, 8, 97–105.

26.

Hansen

G. L.

(1987). Extradyadic relations during courtship. Journal of Sex Research, 23, 382–390.

27.

Harmon-Jones

(2000). Cognitive dissonance and experienced negative affect: Evidence that dissonance increases experienced negative affect even in the absence of aversive consequences. Personality and Social Psychology Bulletin, 26, 1490–1501.

28.

Harmon-Jones

(2001). The role of affect in cognitive-dissonance processes. In Forgas

J. P.

(Ed.), Handbook of affect and social cognition (pp. 237–255). Mahwah, NJ: Lawrence Erlbaum Associates Publishers.

29.

Harmon-Jones

Mills

(1999). Cognitive dissonance: Progress on a pivotal theory in social psychology. Washington, DC: American Psychological Association.

30.

Kinsey

A. C.

Pomeroy

W. B.

Martin

C. E.

(1948). Sexual behavior in the human male. Oxford, England: Saunders.

31.

Kinsey

A. C.

Pomeroy

W. B.

Martin

C. E.

Gebhard

P. H.

(1953). Sexual behavior in the human female. Oxford, England: Saunders.

32.

Kowalski

R. M.

Walker

Wilkinson

Queen

Sharpe

(2003). Lying, cheating, complaining, and other aversive interpersonal behaviors: A narrative examination of the darker side of relationships. Journal of Social and Personal Relationships, 20, 471–490.

33.

Lawson

Samson

(1988). Age, gender and adultery. British Journal of Sociology, 39, 409–440.

34.

Mikula

Athenstaedt

Heschgl

Heimgartner

(1998). Does it only depend on the point of view? Perspective-related differences in justice evaluations of negative incidents in personal relationships. European Journal of Social Psychology, 28, 931–962.

35.

Mongeau

P. A.

Hale

J. L.

Alles

(1994). An experimental investigation of accounts and attributions following sexual infidelity. Communication Monographs, 61, 326–344.

36.

Nisbett

R. E.

Wilson

T. D.

(1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review 84, 231–259.

37.

Schonbach

(1980). A category system for account phases. European Journal of Social Psychology, 10, 195–200.

38.

Schwartz

S. H.

(1994). Are there universal aspects in the structure and contents of human values? Journal of Social Issues, 50, 19–45. doi: 10.1111/j.1540-4560.1994.tb01196.x

39.

Schwartz

S. H.

Bilsky

(1990). Toward a theory of the universal content and structure of values: Extensions and cross-cultural replications. Journal of Personality and Social Psychology, 58, 878–891. doi: 10.1037/0022-3514.58.5.878

40.

Schwartz

S. H.

Cieciuch

Vecchione

Davidov

Fischer

Beierlein

Konty

(2012). Refining the theory of basic individual values. Journal of Personality and Social Psychology, 10, 663–688. doi: 10.1037/a0029393

41.

Sedikides

Gaertner

Toguchi

(2003). Pancultural self-enhancement. Journal of Personality and Social Psychology, 84, 60–79.

42.

Sherman

S. J.

Gorkin

(1980). Attitude bolstering when behavior is inconsistent with central attitudes. Journal of Experimental Social Psychology, 16, 388–403.

43.

Shrira

Foster

J. D.

(2005). Perpetrator and victim narrative accounts of infidelity. Paper presented at the Poster presented at annual meeting of the Society for Personality and Social Psychology, New Orleans, LA.

44.

Simon

Greenberg

Brehm

(1995). Trivialization: the forgotten mode of dissonance reduction. Journal of Personality and Social Psychology, 68, 247–260.

45.

Stillwell

A. M.

Baumeister

R. F.

(1997). The construction of victim and perpetrator memories: Accuracy and distortion in role-based accounts. Personality and Social Psychology Bulletin, 23, 1157–1172.

46.

Stone

Aronson

Crain

A. L.

Winslow

M. P.

(1994). Inducing hypocrisy as a means of encouraging young adults to use condoms. Personality and Social Psychology Bulletin, 20, 116–128.

47.

Thibodeau

Aronson

(1992). Taking a closer look: Reasserting the role of the self-concept in dissonance theory. Personality and Social Psychology Bulletin, 18, 591–602.

48.

Thompson

A. P.

(1983). Extramarital sex: A review of the research literature. Journal of Sex Research, 19, 1–22.