Emotion plays a critical role in the deception process. Researchers have examined outcomes associated with the target of deception--such as increased suspicion, uncertainty, negative affect, and diminished trust and relational satisfaction (e.g., Aune, Metts, & Hubbard, 1998; Cole, 2001; Jang, Smith, & Levine, 2002; Levine, McCornack, & Avery, 1992; McCornack & Levine, 1990). However, little empirical attention has been paid to the emotional consequences for those who engage in deception. Indeed, McCornack and Levine (1990) argue that "Perhaps no other factor relevant to relational communication is as significant and neglected as emotion" (p. 120). This is surprising in light of Buller and Burgoon's (1998) observation that "emotional processes are incorporated in every explanation of deceptive communication" (p. 381). To be sure, both classic and contemporary theories, ranging from Ekman and Friesen's (1969) seminal work on behavioral "leakage" during deception to Burgoon and Buller's (2004) more recent work on Interpersonal Deception Theory, suggest that emotion is a key component within deceptive interactions.
Ekman (2001) argued that several emotions are intertwined with deceit, including fear, "duping delight," guilt, and shame. The latter two, he argued, although distinct, are related. For example, whereas fear is derived from the possible consequences of being caught, shame and guilt result from transgressions of one's internalized moral standards. Based on this similarity, it makes sense to examine these emotions together, as Ekman did and as we do in this study. In addition, Ekman (2001) supposed that several variables should influence the extent to which deception is perceived to be a transgression that leads to the experience of shame and guilt. For example, he speculated that selfishly motivated lies would lead to more guilt and shame than would altruistically motivated lies, and that lies told to personal acquaintances would lead to more guilt than would lies to strangers. In addition, research shows that individuals' sex (Levine, McCornack, & Avery, 1992) and cultural background (Seiter, Bruschke, & Bai, 2002) affect their perceptions of the acceptability of deception, suggesting that these variables may also mediate guilt and shame during deception. Given these relationships and the notable lack of research in this area, this study investigates how communicators' motivations, relationship with the target, sex, and culture (United States vs. China) affect expected feelings of guilt and shame following imagined acts of deception.
Review of Literature
Guilt and Shame
Although guilt and shame share commonalities, the dominant view is that these two emotions involve distinct phenomenological experiences (e.g., Tangney, 1992). Whereas earlier conceptualizations viewed guilt as an internal reaction to some transgression and shame as the result of public exposure of impropriety, in a revised perspective Lewis (1971) suggested that an actual audience is not necessary for experiencing shame. Instead, "In guilt, the self is the source of evaluation, and some specific behavior is the object of that evaluation. In shame, the self is split into a focal object and an observing 'other.' Thus, the self is both the source and the object of evaluation, as one imagines how one would look to the other" (Tangney, 1992, pp. 199-200).
Although theories have focused on situational factors that distinguish guilt and shame (e.g., Benedict, 1946), ironically there has been little systematic study of how the concepts relate to deception (Tangney, 1992). In two studies, DePaulo and colleagues (DePaulo & Kashy, 1998; DePaulo, Kashy, Kirkendol, Wyer, & Epstein, 1996) asked participants to rate their feelings about telling lies on a scale labeled "very comfortable" to "very uncomfortable." In both studies participants reported negative feelings, particularly when lying in close relationships. Similarly, in three other studies, participants reported feelings of guilt associated with imagined or actual deception (Gozna, Vrij, & Bull, 2001; Hample, 1980; Peterson, 1996). In general participants noted that guilt was likely, though in some cases emotions depended on other factors (e.g., personality characteristics of the deceiver). Finally, we found only one study (Tangney, 1992) that examined both guilt and shame as they related to deception. Participants were asked to report three situations in which they were likely to feel shame and three in which they were likely to feel guilt. Results indicated that while lying was likely to induce both emotions, it was more likely to induce guilt than shame, perhaps because guilt is more typically induced by transgressions that harm others.
Despite their contributions, these studies are limited in important ways. First, the scales in the DePaulo studies assessed participants' "feelings" rather than focusing on specific emotions. As such, it is hard to know which emotions participants might have been referring to (e.g., fear, guilt, and so forth). Second, Gozna et al. (2001), Hample (1980), and Peterson (1996) focused on guilt but not shame. Because guilt and shame may be associated with different nonverbal cues (see Ekman, 2001), it would be valuable to know the degree to which one or the other emotion is associated with the enactment of deception. Finally, although Tangney (1992) examined both guilt and shame, the study focused on a large number of transgressions, including deception, that might lead to guilt and shame. The present study, in contrast, focuses exclusively on deception, thereby allowing us to examine complexities such as the various motivations one might have for lying. Given the limitations of past work, it is essential to determine:
RQI: Do individuals perceive that they would experience more guilt than shame (or shame than guilt) when engaging in deception?
We also address the complex nature of deception by examining several variables relevant to deception, including motives for deception, relationship type, differences between respondents for the United States and China, and sex differences.
Motives for Deception
A growing number of researchers (e.g., Hample, 1980; Lindskold & Waiters, 1983; Peterson, 1996; Seiter et al., 2002) have identified different motivations (e.g., to protect self, to avoid conflict, to manage impressions) for deceiving others. Although consensus has not emerged around a single typology of motives, there seem to be at least two basic components common to most: whether the deceptive message is directed at the communicator or the other and whether the deceptive message is told for altruistic or malicious reasons. Deceptive messages told for altruistic reasons are perceived as more acceptable than those told for selfish or malicious reasons (Seiter et al., 2002). Because some deceptions may be perceived as more serious moral transgressions than others, people might experience more negative emotions when telling some types of deceptions than others.
Unfortunately, previous research has neglected to examine this issue. We found only one study that investigated guilt as a result of telling different types of deceptive messages (Peterson, 1996), and none that investigated shame. Moreover, the one study focused less on motivations--although it did distinguish between "white lies" and "blatant lies"--and more on styles of lying (e.g., "failed lies," "omissions," and "half truths"). Participants reported that they would be less likely to experience guilt when telling a white lie than when telling a blatant lie. We thus propose:
H1: Deceptive messages told for malicious or selfish reasons will result in more expected shame and guilt than will mutually benefiting deception and deception that benefits others.
Deception theorists recognize that the nature of relationships impacts the deception process (see Burgoon & Buller, 2004). However, little research has examined how different types of relationships affect emotions during deception, and what research does exist presents an inconsistent pattern of findings. While some researchers suggest that people expect intimate others to be honest (Levine & McCornack, 1992) and feel more "uncomfortable" telling lies to friends than to strangers (DePaulo & Kashy, 1998), a majority of studies find more deception in high quality, personal relationships than in low quality, impersonal ones (Hample, 1980; Lippard, 1988; Millar & Tessar, 1988). Some have found the opposite (DePaulo & Kashy, 1998; Kalbfleisch, 2001), and others report no differences in the amount of deception between the two types of relationships (Williams, 2001, Studies 1 and 2). Moreover, while some studies indicate that deception is perceived less acceptable in personal relationships than in impersonal ones (Backbeir, Hoogstraten, & Terwogt-Kouwenhoven, 1997; Maier & Lavrakas, 1976), other research supports the contrary conclusion (Seiter et al., 2002). Finally, while some research suggests that trivial, harmless deceptions told in close relationships are sometimes perceived as acceptable (Boon & McLeod, 2001), other research indicates that all types of deception tend to be frowned upon (Lawson, 2000). Given these inconsistencies, we ask the following:
RQ2: Does the nature of the relationship between deceivers and those being deceived (child to parent, student to teacher, strangers, friends, employee to boss, spouses) influence the extent to which people perceive that they would experience guilt or shame when deceiving others?
Chinese versus United States Culture
Two theoretical traditions support the proposition that emotional reactions to deception may vary by culture. The first focuses on dimensions such as individualism/collectivism and high vs. low context communication as causal agents in determining cultural differences. Because members of individualistic cultures tend to seek private goals and put their personal interests above group interests, and members from collectivistic cultures tend to put the interests of the group over those of the individual (Hofstede, 1982), one might suspect that members of individualistic cultures might feel less guilt and shame when telling self-benefiting lies while members from collectivistic cultures would feel less guilt and shame when telling other-benefiting lies. Empirical evidence supports this notion. Aune and Waters (1994) found that people from individualistic cultures, compared to those from collectivistic cultures, are more likely to tell lies that protect privacy or benefit another person. Lee and colleagues (Lee, Cameron, Xu, Fu, & Board, 1997; Lee, Xu, Fu, Cameron, & Chen, 2001) found that Taiwanese and Chinese children rated lying about good deeds more positively than did Canadian children. On the other hand, although the effect size was small, a study by Seiter et al. (2002) found that participants from China, a collectivistic culture, perceived both self- and other-benefiting lies to be more acceptable than did participants from the United States, an individualistic culture.
A cultural dimension of high- versus low-context communication style points to a different prediction about the valuation of, or even what counts as, deception (see Yeung, Levine, & Nishiyama, 1999). Chinese culture emphasizes values that may take precedence over direct, explicit communication, leading members of its culture to feel fewer negative emotions when being deceptive. Gao and Ting-Toomey (1998) note that to Chinese, "providing the appropriate information at the appropriate time and context with the appropriate persons is more important than honest and truthful communication" (p. 64). If meaning is believed to exist in the context and not the individual, the entire notion of deception may seem less important across all interactions.
The second theoretical tradition is based on the idea that Western cultures tend to be more guilt oriented whereas Eastern cultures tend to be more shame oriented. Some scholars, however, have argued that this is a false dichotomy. As Creighton (1990) explained:
Ever since Ruth Benedict applied the terms "guilt culture" and "shame culture" to explain what she perceived as a fundamental contrast in the psychological makeup of Japanese and Americans, the applicability of the guilt-versus-shame dichotomy has been a controversial and much debated topic ... For over 40 years Benedict's interpretations have been repeatedly denounced, denied, refuted, and reclassified, but the issue is certainly not dead. (p. 279)
To be sure, although some writers defend Benedict's (1946) position (see Creighton, 1990), others report no difference between the experience of guilt and shame in members from Eastern and Western cultures (e.g., Hong & Chiu, 1991). Still others report that while guilt and shame may be universal emotions, the particular triggers for these emotions, as well as the form in which such emotions are expressed, can differ widely across cultures (Greenwald & Harder, 1998).
Although the literature suggests that people from different cultures vary in their evaluations of deception and expressions of emotion, to our knowledge no research has examined the deception-emotion link across cultures characterized by different traditions and different levels of collectivism/individualism and high/low context communication. Given that China and the United States represent such cultures, it is appropriate to ask the following question:
RQ3: Do people from China and the United States vary in how much they expect to experience guilt or shame when deceiving others?
Moreover, considering that collectivistic and individualistic cultures may differ with regard to their perceptions of motivations for lying (see above), we ask:
RQ4: Do culture (China versus United States) and motivation interact to influence the extent to which people perceive that they would experience guilt or shame when deceiving others?
Levine, McCornack, and Avery (1992) reasoned that because women are more likely than men to view particular verbal exchanges as defining the relationship, any event within the relationship related to verbal exchange, including deception, should be perceived as more important by women than by men. Consequently, women may be more likely than men to perceive deception as a serious relational transgression. Levine et al. (1992) found that women, compared to men, rated lying as less acceptable and more relationally significant. Furthermore, women reported significantly more negative reactions to lying than did men. Consistent with these findings, a study by Tangney and Dearing (cited in Tangney & Dearing, 2002, p. 154) found that females reported greater shame and guilt than their male counterparts. Based on these findings, we propose the following hypothesis:
H2: Compared to males, females will expect to experience significantly more shame and guilt as a result of deception.
The participants were 106 (54 male, 52 female) students from communication classes at a large university in the United States and 103 (51 male, 51 female, one unreported) Chinese students from sociology classes at a large university in the People's Republic of China. United States students received extra credit and Chinese students were paid money for participating.
Materials and Design
This study used a 2 (culture) by 2 (sex) design, with motivation (9 levels) and relationship type (6 levels) nested within each of these four cells. Culture and sex were between-subjects variables, whereas motivation and relationship type were within-subject variables. Participants were presented with hypothetical scenarios depicting nine possible motives for each of six different relational types (i.e., 54 different scenarios). For each scenario, respondents were asked to imagine themselves in the same or a similar situation and then, among other things, to use single item 9-point Likert-type scales to rate the degree to which they would expect to feel guilt and shame after engaging in such deception, with higher scores indicating greater levels of guilt or shame. To explore whether expected guilt and shame are related to expected behavior, participants were asked to indicate on 9-point Likert-type scales the likelihood that they would engage in the deception suggested in each scenario.
Examples of the deception scenarios utilized here have been reported elsewhere (Seiter et al., 2002). Each scenario involved different motivations for deceiving. Two major themes, whether the deception was altruistic or malicious and whether it was intended to benefit the self or the other, were incorporated into one continuum, ranging from malicious deceptions told to harm the other to altruistic deceptions told to benefit the other. Nine different motivation categories were utilized, including deceptions told to affiliate with another, benefit another, protect privacy, avoid conflict, protect self, manage impressions, benefit the self without harming another, benefit the self while harming another, and deceiving for malicious purposes. The scenarios were written so that the nine motivations for lying were presented explicitly (e.g., "in order to impress her," "because she wants him to feel good") and therefore appeared to have face validity. Each motivation was repeated for the six different targets (parent, teacher, stranger, friend, boss, and spouse), which represent relationship type. We stress that this typology was used for heuristic purposes and is not presented as a complete or conclusive scheme.
Following construction of the scenarios, two Chinese graduate students checked the scenarios to determine whether they were sensible to people in Chinese culture. Six were not and were rewritten until found to be realistic. Afterward, as verification that the scenarios represented the primary motive that they were intended to represent, someone known to the researchers read each of the scenarios and categorized them into nine piles, each pile representing one motivation. Seven scenarios were not accurately categorized so were rewritten to better represent the intended motivation. Afterward, two undergraduate students, one from China and one from the United States, completed the same task (i.e., categorizing scenarios into piles). Analysis revealed that the students' agreement regarding categories was high (Cohen's kappa = .96).
Finally, one Chinese graduate student translated the scenarios into Chinese, and another translated them back into English to make sure that the meaning of the scenarios had not changed during translation. Back translation continued until two versions of the scenarios, one in English and the other in Chinese, were complete.
Participants were told this project was investigating people's thoughts and feelings about deception and whether certain types of deception are associated with feelings such as guilt or shame. They were presented with a brief statement describing the difference between these two emotions (available from the authors upon request) and then completed the survey.
Because of the large number of items on the survey (over 400, including the 54 scenarios reported here and data collected for other purposes) respondent fatigue was a concern. The order of the items was not varied across respondents and thus direct tests of fatigue effects were not possible, but the data were divided into quartiles for guilt and shame to explore changes in responses over time. The means and standard deviations did not follow linear patterns, suggesting that there were not systematic changes over time. In addition, ratings were not consistently high or low, which suggest the absence of a halo effect. Respondent fatigue did not appear to contaminate the results.
The survey asked participants to self-report their expected emotional reactions to imagined deceptions. To do so, respondents read the 54 scenarios and reported how much shame and guilt they thought they would feel using two 9-point scales, one for shame and one for guilt, with 1 representing not at all guilty/shameful and 9 representing extremely guilty/shameful. Although using hypothetical scenarios has its drawbacks, including the possibility that people's perceptions of how they would feel might differ from how they would actually feel, we believe this method is appropriate. Specifically, using hypothetical scenarios enabled us to examine different motivation and relationship combinations in a way that using perceptions of actual deception would not. Such a method also measures people's expectations about how they would react, and these expectations are likely influenced by culture as well as one's past experiences in particular types of relationships.
To examine RQ1 expected guilt and shame ratings were summed across all 54 manipulations of relationship and motive. The overall mean for guilt was 6.04 and was 5.89 for shame. A paired t test revealed a significant correlation, r = .89, n = 208, p < .001, and difference between the two means, t(207) = 5.25, p < .001, suggesting that people expect to experience moderate to moderately high levels of guilt and shame when engaging in deception. Analysis of the remaining research questions and hypotheses required separate repeated-measure multivariate analysis of variance for each dependent variable. For expected guilt, significance was obtained for Mauchly's test of sphericity for both repeated measures (relationship and motive) and their interaction, and thus Greenhouse-Geisser adjustments to the degrees of freedom were applied to all subsequent tests. Significance was obtained for culture (F[1, 204] = 12.9, p<.001, partial [eta.sup.2] = .07), motive (F[5.09] = 791.67, p > .001, partial [eta.sup.2] = .80), and relationship (F[4.36, 5714.2] = 57.36, p < .001, partial [eta.sup.2] = .22), but not for sex (observed power = .42). For culture, the United States sample reported expecting more guilt (M = 6.27, SD = .83) than did the Chinese sample (M = 5.77, SD = .97).
The within-subjects variables had multiple levels that required follow-up tests. Repeated deviation contrasts compare each level with the grand mean for all levels up to the degrees of freedom. The advantage to this procedure is that it does not assume ordinality, and its drawback is that the final level is not tested. Given the lack of ordinality, the repeated deviation contrast was deemed most appropriate in this analytical context. Post hoc contrasts revealed significant differences between all relationship types and the grand mean except for friendships; strangers, as the final group, were not tested.
For the relationship variable, estimated marginal means revealed that the most guilt was expected when deceiving teachers (M = 6.45, SE = .08), followed by parents (M = 6.36, SE = .07), strangers (M = 6.09, SE = .07), friends (M = 6.00, SE = .07), bosses (M = 5.75, SE = .08), and spouses (M = 5.48, SE = .09) (estimated marginal means are used here to interpret the follow-up comparisons; original descriptive tables are available from the authors upon request).
For motives, estimated marginal means revealed that the least guilt was expected by deceptions for affiliation (M = 3.77, SE = .10), followed by benefiting the other (M = 4.03, SE = .09), privacy (M = 4.66, SE = .08), avoiding conflict (M = 5.07, SE = .09), protecting the self (M = 6.58, SE = .09), impression management (M = 6.82, SE = .08), benefiting the self without harming the other (M = 7.20, SE = .08), benefiting the self and harming the other (M = 7.93, SE = .06), and deceiving out of maliciousness (M = 8.13, SE = .06). Follow-up tests confirmed significant differences between every condition tested.
For expected shame, significance was again obtained for Mauchly's test of sphericity and the repeated deviation procedure was again the follow-up test of choice. Significance was obtained for relationship, F(4.58, 5643.09) = 65.70, p K .001, partial [eta.sup.2] = .25, motive F(5.07, 5643.09) = 915.31, p < .001, partial [eta.sup.2] = .82, and culture, F(1, 203) = 6.06, p < .05, partial [eta.sup.2] = .03, but not for sex (observed power = .12). The differences for culture were slight, with the United States sample reporting slightly more expected shame (M = 6.05, SD = .93) than the Chinese sample (M = 5.75, SD = .91). For relationship, estimated marginal means revealed that the most shame was expected when deceiving parents (M = 6.47, SE = .07), followed by teachers (M = 6.10, SE = .06), strangers (M = 6.02, SE = .07), friends (M = 5.86, SE = .08), bosses (M = 5.51, SE = .08), and spouses (M = 5.42, SE = .09). Follow-up tests revealed significant differences between all levels and the grand mean except for friendship, which was nonsignificant, and stranger, which was not tested. For motive, the least shame was expected when deceiving for affiliation (M = 3.62, SE = .09), followed by benefit other (M = 3.79, SE = .09), privacy (M = 4.60, SE = .07), avoiding conflict (M = 4.91, SE = .09), protecting the self (M = 6.44, SE = .09), impression management (M = 6.71, SE = .08), benefiting the self without harming the other (M = 7.13, SE = .08), benefiting the self while harming the other (M = 7.82, SE = .06), and malice (M = 8.06, SE = .06). Follow-up tests revealed all comparisons significantly different from the grand mean.
Of the interaction effects, only two contributed nontrivially to the overall variance accounted for. The first was the relationship by culture interaction (F[4.58, 5643.09] = 31.05, p < .001, partial [eta.sup.2] = .14). As with the guilt interaction, the United States sample expected greater shame for all conditions except the teacher relationship, where the Chinese sample expected greater amounts of shame. Second, the relationship by motive interaction was significant (F[27.80, 5643.09] = 65.97, p < .001, partial [eta.sup.2] = .25). Estimated marginal means, standard errors, and follow-up test results for all 54 conditions are reported in Table 1.
Finally, likelihood of use was correlated with guilt and shame. Both correlations were significant and negative (guilt r = -.58, shame r = -.50). The Chinese sample (M = 3.93) indicated a significantly greater likelihood of use summed across all 54 scenarios as compared to the United States sample (United States M = 3.30; t(207) = -5.27).
Our results can be summarized this way: When people make judgments about how much shame and guilt they are likely to experience when deceiving others, those judgments are influenced by their perceived motivation and, to a lesser extent, the relationship they share with the person they imagine themselves to be deceiving. These effects remain constant across cultures, although the overall emotional reaction across cultures differ with people from the United States reporting that they expect more guilt and shame when deceiving others than do people from China. The notable difference is for teacher relationships, for which the United States participants reported expecting less guilt and shame when behaving deceptively than did Chinese participants. Not surprisingly, expected guilt and shame were negatively correlated with the intention to engage in a particular deception, and a model demonstrating the relationship between culture, guilt, and shame will no doubt include a number of dependent variables mediating one another (although it is of course the case that the ordering of these variables is a theoretical choice and correlations by themselves cannot indicate which variables are of terminal importance and which ones serve a mediating role). The remainder of this section will summarize these conclusions and discuss their implications and theoretical significance.
First, previous work suggests that people render judgments about the morality of deceptive acts, experience negative emotions when communicating deceptively, and, as a consequence, behave differently when deceiving than when telling the truth. Although the present study did not explore this entire process, it sheds light on one previously neglected link. Specifically, this is one of few empirical studies connecting deception to expected guilt and shame. Importantly, this study supports previous theoretical principles suggesting that negative emotions are a vital explanatory mechanism in the enactment of deception.
Second, although previous research indicates that acts of deception can result in negative outcomes for people who are deceived, this study suggests that deceivers may be affected negatively as well. Specifically, guilt and shame are uncomfortable and possibly painful phenomenological experiences (Tangney & Dearing, 2002). As such, this study indicates that the potentially negative consequences of deception should not be ignored.
Third, it has been argued that shame and guilt are emotions that guide people's behavior, decreasing the frequency of acts that might lead to such negative emotions (Greenwald & Harder, 1988). Indeed, we found a negative correlation between guilt and shame and the likelihood of engaging in deception. Such thinking suggests that if deception leads to guilt and shame it might occur less frequently than other forms of communication. At the same time, previous research indicates that deception is indeed common (e.g., DePaulo & Kashy, 1998). How might this be explained in light of our findings? Perhaps people contemplating deception engage in a cost/reward analysis. If the reward for deceiving (escaping trouble) exceeds the cost (expected guilt or shame), the person may choose to deceive. If not, the person may choose to behave honestly. Future research should explore this possibility as well as its implications. For instance, previous research (Zuckerman, DePaulo, & Rosenthal, 1981) found that deception is marked by long response latencies that may be due to cognitive load (i.e., it is cognitively easier to tell the truth than to make something up). An alternative or accompanying explanation is that potential deceivers are weighing the possible costs of deception (e.g., how guilty will I feel afterward?) before engaging in the behavior.
Analysis of our first research question indicated that expected guilt and shame ratings were highly correlated and that guilt ratings were slightly higher (6.04 for guilt and 5.89 for shame). This is consistent with previous research indicating that guilt and shame, though related (our data, in fact, show considerable overlap), are distinct emotions, and that deception is more likely to lead to guilt than to shame (see Tangney & Dearing, 2002). Because these different emotions may be associated with different nonverbal expressions (Ekman, 2001) one implication might be that deception detection may be more successful when monitoring one emotion rather than another.
As predicted in Hypothesis 1, people expected to feel more guilt and shame following deceptions told for malicious and selfish reasons than following deceptions told to benefit the self or the other. This finding is consistent with previous research (see above) and has important implications. For instance, the fact that some messages are associated with less expected guilt and shame might explain why research has found that deception is common. A next logical step is to ask whether deceptions associated with less negative affect are the most common. For instance, whereas Metts and Hippensteele (1987) found that socially approved lies (e.g., other-benefiting lies) are used more often than self-benefiting lies, Hample (1980) found the opposite.
In addition, there is a need for research exploring the association between motivation for deceiving and the enactment of deception. According to Lindskold and Waiters (1983) motives may affect one's ability to deceive successfully. If different motivations are associated with different levels of expected guilt and shame, as this study has found, different motivations may also be associated with different nonverbal leakage during deception. Thus, the current failure to identify a reliable set of cues that differentiates deceivers from truthtellers (see Feeley & Young, 1998; McCornack, 1997) could be due to previous researchers' failure to distinguish between research participants' motivations for deceiving. Future research should explore this issue.
Tests of our second research question suggest that a deceiver's relationship with the target of deception does influence the degree of expected guilt and shame. This finding was tempered by an interaction with culture, so that United States participants felt more guilt and shame for all relationships except for teachers, suggesting that deceiving teachers does not evoke as much guilt and shame in United States respondents as it does in Chinese respondents. In addition, the relationship variable interacted with the motivation variable as reported in Table 1.
Given that studies examining issues such as these are uncommon, our results indicate that the study of deception and emotion as situated within relationships is a vital area for future examination. Specifically, future research should examine which relational dimensions contribute to the experience of emotions during deception. To date, theorists have focused on intimacy, claiming that guilt tends to be elicited in the context of relatively intimate relationships (Vangelisti & Sprague, 1998), but our results indicate that relationships characterized by less intimacy (such as teachers) relative to others (such as spouses and friends) seem to be associated with more guilt and shame following deception. It is interesting to note that the most guilt and shame were expected when deceiving some high-status others (parents and teachers) and relatively less was felt when deceiving intimates (friends and spouses). Perhaps status or power distance may be an important dimension affecting emotions during deception. Whatever the case, although our results are probably best seen as exploratory, the theoretically powerfully point they support is that relationships and motivations interact with one another, and theories that consider each factor in isolation may be incomplete.
Analysis of our final two research questions revealed significant differences for United States versus Chinese participants for both expected shame and guilt but no significant interactions between culture and motive. The culture and relationship interaction, however, was significant. Generally, the United States sample reported more guilt and shame for all relationships except for teachers. This is consistent with Greenwald and Harder's (1998) assertion that, although guilt and shame may be universal emotions, the "particular triggers for these emotions, that is, what makes an individual feel guilty or ashamed, as well as the form in which these affects are expressed, may well differ widely from culture to culture" (p. 226). Thus, deception may not trigger these emotions in China to the same extent as in the United States.
The finding that individuals from Chinese culture may expect less intense negative emotion following deception is consistent with literature reporting that individuals from China find deception to be more acceptable than individuals from the United States and it also may help explain the results of previous studies. For example, Feldman (1979) reported that Korean subjects (who, like the Chinese subjects in our study, represent a high-context culture) may be able to control nonverbal behavior during deception to a greater extent than United States subjects, and Cody, Lee, and Chao (1989) found that Chinese participants rarely communicated negative feelings when lying. This discussion all points to the importance of the high- versus low-context dimension rather than individualism-collectivism explaining assessments of deception. Importantly, however, these dimensions were not tested directly in the present study. Future work should do so.
Our final hypothesis predicting sex differences was not confirmed. No significant main effects were present for either guilt or shame, and when sex did emerge as significant term in an interaction the magnitude was trivial. These findings, though tempered by the relatively low power of the tests (especially for shame), are inconsistent with previous research. Future studies should explore reasons for this inconsistency.
Before concluding we should note the limitations of this study. We already mentioned the issues of possible respondent fatigue and the problems associated with using hypothetical scenarios. The fact that Chinese students were from sociology classes and United States students were from communication classes should also be noted as a possible confounding variable, as should the fact that the two samples had different motivations for participating in the study (extra credit vs. money). Moreover, it may have been inappropriate to ask participants to speculate about interactions with a spouse since it is unknown how many were married. Also, the scenarios were always presented in the same order, thereby leading to the possibility of order effects. Finally, it is important to note that single messages were used to represent each type of motivation. Jackson (1992) persuasively argued that single-message designs introduce case-category confounds and the possibility that the insufficiency of data may be concealed. Although we followed Jackson's advice of using "exemplars" to represent each category, we recognize that our findings could be specific to the scenarios we created and that replication is necessary to show that our findings are generalizable.
One strength of this study, however, is the inclusion of several different variables that likely affect the extent to which people expect to feel guilt or shame following deception. These expectations may influence when and how people deceive others. Few deception studies have spanned the various levels of abstraction to study cultural and relational variables simultaneously with communicator characteristic (i.e., sex), goal (i.e., motivational), and emotional outcome (i.e., guilt and shame) variables. This study is an initial attempt at understanding how these variables work together. Deception, depending on motivation, may provoke shame and guilt two emotions that play important roles not only in relationships but in how individuals perceive themselves and regulate their own behavior. As such, it is time that systematic empirical work begins to focus on this rich area of study.
This manuscript was accepted by the previous editor, Professor Jim L. Query.
Aune, R. K., Metts, S., & Hubbard, A. S. E. (1998). Managing the outcomes of discovered deception. Journal of Social Psychology, 138, 677-689.
Aune, R. K. & Waters, L. L. (1994). Cultural differences in deception: Motivations to deceive in Samoans and North Americans. International Journal of Intercultural Relations, 18, 159-172.
Backbier, E., Hoogstraten, J., & Terwogt-Kouwenhoven, K. M. (1997). Situational determinants of the acceptability of telling lies. Journal of Applied Social Psychology, 27, 1048-1062.
Benedict, R. (1946). The Chrysanthemum and the sword: Patterns of Japanese culture. Boston: Houghton Mifflin.
Boon, S. D. & McLeod, B. A. (2001). Deception in romantic relationships: Subjective estimates of success at deceiving and attitudes toward deception. Journal of Social and Personal Relationships, 18, 463-476.
Buller, D. B. & Burgoon, J. K. (1998). Emotional expression in the deception process. In P. A. Andersen & L. K. Guerrero (Eds.), Handbook of communication and emotion: Research, theory, applications, and contexts (pp. 382-402). San Diego, CA: Academic Press.
Burgoon, J. K. & Buller, D. B. (2004). Interpersonal deception theory. In I. S. Seiter & R. H. Gass (Eds.), Perspectives on persuasion, social influence, and compliance gaining (pp. 239-264). Boston, MA: Allyn & Bacon.
Cody, M. J., Lee, W., & Chao, E. Y. (1989). Telling lies: Correlates of deception among Chinese. In J. P. Forgas & M. J. Innes (Eds.), Social psychology (pp. 309-368). North Holland: Elsevier Science Publishers.
Cole, T. (2001). Lying to the one you love: The use of deception in romantic relationships. Journal of Social and Personal Relationships, 18, 107-129.
Creighton, M. R. (1990). Revisiting shame and guilt cultures: A forty-year pilgrimage. Ethos, 18, 279-307.
DePaulo, B. M. & Kashy, D. A. (1998). Everyday lies in close and casual relationships. Journal of Personality and Social Psychology, 74, 63-79.
DePaulo, B. M., Kashy, D. A., Kirkendol, S. E., Wyer, M. M., Epstein, J. A. (1996). Lying in everyday life. Journal of Personality and Social Psychology, 70, 979-995.
Ekman, P. (2001). Telling lies: Clues to deceit in the marketplace, politics, and marriage (3rd ed.). New York: W.W. Norton & Company.
Ekman, P. & Friesen, W. V. (1969). Nonverbal leakage and clues to deception. Psychiatry, 32, 88-106.
Feeley, T. H. & Young, M. J. (1998). Humans as lie detectors: Some more second thoughts. Communication Quarterly, 42, 109-126.
Feldman, R. S. (1979). Nonverbal disclosure of deception in urban Koreans. Journal of Cross-Cultural Psychology, 10, 73-83.
Gao, G. & Ting-Toomey, S. (1998). Communicating effectively with the Chinese. Thousand Oaks, CA: Sage Publications.
Gozna, L. F., Vrij, A., & Bull, R. (2001). The impact of individual differences on perceptions of lying in everyday life and in a high stake situation. Personality and Individual Differences, 31, 1203-1216.
Greenwald, D. F. & Harder, D. W. (1998). Domains of shame: Evolutionary, cultural, and psychotherapeutic. In P. Gilbert & B. Andrews (Eds.), Shame: Interpersonal behavior, psychopathology, and culture (pp. 225-245). Oxford, UK: Oxford University Press.
Hample, D. (1980). Purposes and effects of lying. The Southern Speech Communication Journal, 46, 33-47.
Hofstede, G. (1982). Culture's consequences. Beverly Hills, CA: Sage Publications.
Hong, Y. & Chiu, C. (1991). A study of the comparative structure of guilt and shame in Chinese society. The Journal of Psychology, 126, 171-179.
Jackson, S. (1992). Message effects research: Principles of design and analysis. New York: Guilford Press.
Jang, S. A., Smith, S. W., & Levine, T. R. (2002). To stay or to leave? The role of attachment styles in communication patterns and potential termination of romantic relationships following discovery of deception. Communication Monographs, 69, 236-252.
Kalbfleisch, P. J. (2001). Deceptive message intent and relational quality. Journal of Language and Social Psychology, 20, 214-230.
Lawson, T. J. (2000). Are kind lies better than unkind truths? Effects of perspective and closeness of relationship. Representative Research in Social Psychology, 24, 11-19.
Lee, K., Cameron, C. A., Xu, F. Fu, G., & Board, J. (1997). Chinese and Canadian children's evaluations of lying and truth-telling. Child Development, 64, 924-934.
Lee, K., Xu, F., Fu, G., Cameron, C. A., & Chen, S. (2001). Taiwan and mainland Chinese and Canadian children's categorization and evaluation of lie- and truth-telling: A modesty effect. British Journal of Developmental Psychology, 19, 525-542.
Levine T. R. & McCornack, S. A. (1992). Linking love and lies: A formal test of the McCornack and Parks model of deception detection. Journal of Social and Personal Relationships, 9, 143-154.
Levine T. R. & McCornack, S. A, & Avery, B. (1992). Sex differences in emotional reactions to discovered deception. Communication Quarterly, 40, 289-296.
Lewis, H. B. (1971). Shame and guilt in neurosis. New York: International Universities Press.
Lindskold, S. & Han, G. (1986). Intent and the judgement of lies. Journal of Social Psychology, 126, 129-130.
Lindskold, S. & Walters, P. S. (1983). Categories for acceptability of lies. Journal of Social Psychology, 120, 129-136.
Lippard, P. V. (1988). "Ask me no questions, I'll tell you no lies": Situational exigencies for interpersonal deception. Western Journal of Speech Communication, 52, 91-103.
Maier, R. A. & Lavrakas, P. J. (1976). Lying behavior and the evaluation of lies. Perception and Motor Skills, 42, 575-581.
McCornack, S. A. (1997). The generation of deceptive messages: Laying the groundwork for a viable theory of interpersonal deception. In J. O. Green (Ed.), Message production: Advances in communication theory (pp. 91-126). Mahwah, NJ: Erlbaum.
McCornack, S. A. & Levine, T. R. (1990). When lovers become leery: The relationship between suspicion and accuracy in detecting deception. Communication Monographs, 57, 219-230.
Metts, S. (1989). An exploratory investigation of deception in close relationships. Journal of Social and Personal Relationships, 6, 159-179.
Metts, S. & Hippensteele, S. (1987, February). Characteristics of deception in close relationships. Paper presented at the annual meeting of the Western States Communication Association, San Diego, CA.
Millar, K. U. & Tesser, A. (1988). Deceptive behavior in social relationships: A consequence of violated expectations. The Journal of Psychology, 122, 263-273.
Peterson, C. (1996). Deception in intimate relationships. International Journal of Psychology, 31, 279-288.
Seiter, J. S., Bruschke, J., & Bai, C. (2002). The acceptability of deception as a function of perceivers' culture, deceivers' intention, and deceiver-deceived relationship. Western Journal of Communication, 66, 158-180.
Tangney, J. P. (1992). Situational determinants of shame and guilt in young adulthood. Personality and Social Psychology Bulletin, 18, 199-206.
Tangney, J. P. & Dearing, R. L. (2002). Shame and guilt. New York: Guilford Press.
Vangelisti, A. L. & Sprague, R. J. (1998). Guilt and hurt: Similarities, distinctions, and conversational strategies. In P. A. Andersen & L. K. Guerrero (Eds.), Handbook of communication and emotion: Research, theory, applications and contexts (pp. 123-154). San Diego, CA: Academic Press.
Williams, S. S. (2001). Sexual lying among college students in close and casual relationships. Journal of Applied Social Psychology, 31, 2322-2338.
Yeung, L. N. T., Levine, T. R., & Nishiyama, K. (1999). Information manipulation theory and perceptions of deception in Hong Kong. Communication Reports, 12, 1-11.
Zuckerman, M., DePaulo, B. M., & Rosenthal, R. (1981). Verbal and nonverbal communication of deception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 14, pp. 1-59). New York: Academic Press.
John S. Seiter (Ph.D., University of Southern California, 1993) is a professor at Utah State University, and Jon Bruschke (Ph.D., University of Utah, 1994) is a professor at California State University, Fullerton. The authors would like to thank Chunsheng Bai, Jacey Skinner, Harold Kinzer, Debora Seiter, Sally Yang, Chuhui Wang, and Dan Wang for their invaluable assistance with data collection and/or cross translation, and anonymous Reviewer B for especially helpful input. Correspondence to: John S. Seiter, Department of Languages, Philosophy, and Speech Communication, Utah State University, Logan, UT 84322-0720, U.S.A. Tel.: (435) 797-0138; E-mail: email@example.com
Table 1 Estimated Marginal Mean Shame Ratings and Standard Errors for Relationships and Motivations Relationship Motive Boss Parent Teacher Affiliation 4.44 (.15) 4.01 (.16) * 3.93 (.15) Benefit other 5.14 (.15) * 2.29 (.10) * 4.36 (.14) * Privacy 5.33 (.14) * 7.36 (.12) * 2.96 (.13) * Avoid 5.75 (.15) * 4.65 (.15) * 4.28 (.14) * conflict Protect Self 5.66 (.15) * 6.09 (.16) * 7.67 (.13) * Impression 7.82 (.11) * 7.52 (.13) * 6.51 (.14) * management Benefit self 7.52 (.13) 7.35 (.14) 7.67 (.13) * No harm other Benefit self 8.14 (.09) * 7.06 (.12) * 8.44 (.08) * Harm other Malice 8.41 (.08) 8.61 (.10) 8.30 (.09) Relationship Motive Friend Spouse Stranger Affiliation 3.81 (.14) * 2.78 (.12) * 2.73 (.13) Benefit other 5.00 (.16) * 3.13 (.13) * 2.80 (.13) Privacy 2.97 (.15) * 5.69 (.15) * 3.29 (.15) Avoid 5.06 (.15) 4.13 (.14) * 5.57 (.15) conflict Protect Self 6.42 (.15) 7.30 (.13) * 5.51 (.18) Impression 6.05 (.15) * 5.55 (.15) * 6.82 (.14) management Benefit self 6.99 (.14) 6.88 (.14) 6.35 (.14) No harm other Benefit self 8.46 (.08) * 6.69 (.13) * 7.86 (.11) Harm other Malice 7.99 (.11) 7.20 (.14) 7.85 (.12) Note. Standard errors appear in parentheses. No follow-up tests were conducted on Malice motivation or Stranger relationship because deviation contrasts are only conducted up to the degrees of freedom. An asterisk indicates significant difference from the grand mean.…