College Students’ Revenge Goals Across Friend, Romantic Partner, and Roommate Contexts: The Role of Interpretations and Emotions

Abstract

Residential college environments provide young people with distinctive relationship opportunities and challenges. A major purpose of the present study was to learn whether college students respond differently to conflict-of-interest vignettes in three different relationship contexts. Students were more likely to make negative interpretations about their romantic partner’s behavior than they did about their friend’s or roommate’s behavior. They were also more likely to feel angry and hurt and to endorse hostile goals and strategies with romantic partners. A second major purpose was to learn about the types of interpretations and emotions associated with revenge goals in conflict-of-interest situations. Results indicated that interpreting the other person’s actions as disrespectful and as rejecting was related to revenge goals and also predicted to revenge goals beyond the contributions of anger and hurt feelings.

Introduction

The period of life when individuals transition from adolescence to adulthood, termed “emerging adulthood” by Arnett (2000), is a phase of development characterized by increased exploration of identity and social relationships (Arnett, 2000; Collins & van Dulmen, 2006; Erikson, 1968). When youth move away from home and family, they increasingly seek support, security, and need fulfillment from close peers to help them through major life changes (Berscheid, Snyder, & Omoto, 1989; Carbery & Buhrmester, 1998). Residential colleges offer an opportunity for youth to form new close relationships with several different types of relationship partners: friends, roommates, and romantic partners. These different types of relationships share the potential to become sources of intimacy, emotional support, and instrumental assistance (e.g., Berg, 1984).

Although all three types of relationships can be close and significant, romantic relationships during the college years have a special developmental significance (Furman, Brown, & Feiring, 1999). Romantic relationships carry higher expectations of various relationship provisions including intimacy, support, help, and, unique to this relationship context, exclusive commitment, passion, and sex (e.g., Connolly, Craig, Goldberg, & Pepler, 1999; Flannagan, Marsh, & Fuhrman, 2005; Hazan & Shaver, 1987). Furthermore, although students often have several good friends and may have more than one roommate, genuine romantic relationships (in contrast to more brief “hook up” relationships) typically have that “my one and only” quality that adds to their significance. This, in turn, may increase individual tendencies to negatively overreact to conflicts, both large and small (Flannagan et al., 2005). The heightened vulnerability and overreactions to conflict in romantic relationships during this life period is reflected in evidence that 25–30% of college students indicated that they had experienced psychological and/or physical abuse in their romantic relationships over a 12-month period (e.g., Sugarman & Hotaling, 1989).

Understanding the factors that increase negative responses within conflict, defined as an interpersonal, behavioral event involving opposition (Laursen & Collins, 1994), is important because how individuals respond in conflict affects their success in close relationships (e.g., Gottman, 1994; Laursen & Pursell, 2009) and may negatively affect the support that young people come to depend on from peers. Conflicts can raise questions about partners’ availability to provide support and security, test partners’ skills at maintaining a cooperative relationship, and lead individuals to evaluate and revise their feelings and beliefs about their partner and the relationship (Simpson, Rholes, & Phillips, 1996). Additionally, in conflict, individuals assess whether their social needs, including their needs for autonomy, relatedness, and competence, are being met (see Deci & Ryan, 2000; Shaver & Buhrmester, 1983). As such, even minor conflicts may be threatening, leading some individuals to overreact and behave badly.

This line of reasoning suggests that college students may be more challenged by conflict with romantic partners than with friends or roommates. More specifically, we hypothesize that in the romantic partner context, compared with in the friend or roommate context, college students are more likely to make negative interpretations about the meaning of the other person’s actions, experience greater hurt feelings and anger, endorse more problematic goals (most notably revenge goals), and endorse more maladaptive strategies, such as verbal aggression. The potentially powerful role of revenge goals is considered next.

Revenge Goals

Children and adolescents who choose to pursue revenge goals in conflict-of-interest situations, as well as in ambiguous provocation situations, are more likely to have problematic relationships with peers in general and with friends in particular (e.g., Erdley & Asher, 1996; Lochman, Wayland, & White, 1993; Rose & Asher, 1999; Slaby & Guerra, 1988). A demonstration of the potentially destructive power of revenge goals within a close relationship comes from Rose and Asher (1999) who examined whether children’s goals in conflict-of-interest situations were related to their success in friendship. Rose and Asher presented children with hypothetical situations depicting relatively benign conflict-of-interest situations with a friend, such as a disagreement over which game to play, or a child requesting help from a friend but the friend has other obligations or commitments. After each hypothetical scenario, children rated various goals and strategies in terms of how likely they would be to choose each goal and enact each strategy. Results indicated that the endorsement of revenge goals (i.e., “I would be trying to get back at my friend”) related to independent indices of children’s friendship adjustment even when statistically controlling for children’s overall level of acceptance by peers. Specifically, revenge goals were negatively associated with the number of friends children had (assessed sociometrically) and the quality of children’s best friendships as independently reported by the best friend.

The findings from this study suggest that the conflict-of-interest situation is an important context in that how individuals respond is predictive of their success at friendship. The findings also reinforce the claim that a comprehensive understanding of relationship competence involves attention not just to people’s overt behavioral responses to a conflict, but also to the underlying goals people are trying to accomplish in social situations (see Asher, MacEvoy, & McDonald, 2008; Crick & Dodge, 1994; Renshaw & Asher, 1983). In addition, the findings suggest that of the various goals that people might pursue in conflicts, the goal of revenge is particularly powerful in influencing the course of students’ relationships.

An important question not addressed to date is why some youth endorse revenge goals with a friend even when conflicts are relatively minor and benign. The present study examines the hypothesis that particular interpretations in conflict-of-interest situations are likely to lead to revenge goals even when the other person is a close relationship partner and when the stakes seem, on the surface, to be relatively small.

Interpretations

There is a long and productive history of research on interpersonal attributions and their associations with social behaviors (e.g., Dodge, 1980; Kelley & Thibaut, 1978). However, relatively little attention has been given to the situation-specific attributions or interpretations people make about what the other person’s actions imply regarding how the other person thinks and feels about the self (for an exception, see Bradbury & Fincham, 1990). Watzlawick, Beavin, and Jackson (1967) contended that the surface level, or topography, of behavior carries deeper level meanings that are communicated about how the partner defines the relationship, how the partner sees himself or herself in relation to the other, how the partner feels about the other, and how the partner wants to be viewed. The implication of Watzlawick et al.’s proposal is that people make interpretations of the other person’s behavior in terms of what the behavior communicates at one or more of these deeper levels.

To assess some of these deeper level messages, we assessed eight distinct interpretations that could be made in conflict-of-interest situations. These included interpretations of (1) being disrespected, (2) being rejected, (3) not being cared for, and (4) being betrayed. We assessed these four interpretative dimensions because they pertain to human beings’ social needs for respect, relatedness, trust, and security (Asher & McDonald, 2009; Bowlby, 1969; Deci & Ryan, 2000; Shaver & Buhrmester, 1983). We hypothesized that when individuals perceive that their partners do not respect them, do not like them, do not care for them, or have betrayed them, revenge is more likely (e.g., Buckley, Winkel, & Leary, 2004; McCullough, 2008; Miller, 2001).

In addition to interpretations regarding how the relational partner perceives the self, we were also interested in whether people would interpret the other’s behavior as (5) morally wrong and (6) a sign that the relationship lacked viability. Interpretations that the relational partner has done something morally wrong may increase revenge goals because perceived violations of moral standards may elicit desires to teach the violator a lesson about their perceived immoral behavior (Heider, 1958). Further, if individuals make the interpretation that the relationship is unlikely to last, they may be more likely to pursue revenge goals, perhaps because they are less invested in the continuation of the relationship (Finkel, Rusbult, Kumashiro, & Hannon, 2002). More optimistic interpretations were also assessed: (7) that the problem was easily solvable and (8) that the conflict was just what it seemed, namely that the other person just wanted something different or had different interests from the self and that no “relationship statement” (Watzlawick et al., 1967) was being made. We hypothesized that when individuals believe that the conflict can be resolved and that the conflict is benign they would be less likely to seek revenge (cf., Dodge, 1980).

Of these eight interpretations, we were particularly interested in the interpretations of disrespect and rejection based on theory and research linking these experiences to retaliation and aggression (e.g., Buckley et al., 2004; Miller, 2001). Considerable developmental research with children has focused on rejection, defined as being disliked or excluded by others (see Asher & McDonald, 2009, for a recent review). Children who are disliked by peers become more aggressive over time, especially if they are initially aggressive (e.g., Coie, Lochman, Terry, & Hyman, 1992; DeRosier, Kupersmidt, & Patterson, 1994). Additionally, in efforts to further understand how older adolescents react when they are rejected, researchers have designed paradigms in which college student participants are excluded by confederates. For example, Buckley et al. (2004) had participants fill out a personal information form and told them that another person would judge their information while they evaluated the same information from that person. Subsequently, they learned how the other person had supposedly evaluated them. People who were told that the other person chose not to work with them were more angry and more aggressive in response (i.e., by assigning aversive tasks to the offender). Together with the longitudinal work with children, this laboratory-based research with college students provides a foundation for hypothesizing that interpreting another person’s actions as rejecting of self leads to revenge goals, a hypothesis that has not been directly tested previously and is a primary goal of the current study.

Interpreting another person’s actions as disrespectful may also increase the probability of revenge being sought. McCullough (2008) suggests that desires for revenge are a universal human reaction to perceptions that one has been dishonored or disrespected. The noun “disrespect” is defined in the Merriam-Webster’s Collegiate Dictionary (2003) as “the lack of the quality … of being esteemed”. Generally, people believe that they have the right to be respected and to be treated with respect by others. To be disrespected means that one was not given consideration by others or treated at the level one’s status warrants, that an act was unfair, or that others have broken an implicit contract of what is expected and acceptable in a relationship (Miller, 2001). Ethnographers have observed that in certain communities aggression often follows from perceptions of disrespect. In these communities, salient norms about respect exist, and when these norms are violated revenge is expected and sanctioned. The set of shared beliefs about respect that legitimize retaliation for offenses is typically referred to as the “code of honor” (e.g., Anderson, 1999; Horowitz, 1983; Nisbett & Cohen, 1996). To test hypothesized relations between disrespect interpretations and revenge, the current study investigates whether interpretations of disrespect are associated with revenge goals within conflicts of interest.

Comparing Rejection and Disrespect Interpretations

We were also interested in comparing the relative influences on revenge goals of interpreting another person’s actions as rejecting vs. disrespectful. These two interpretations may often co-occur. When people feel rejected, they may also feel disrespected and vice versa. However, adults clearly distinguish the dimension of acceptance–rejection, that is, the dimension of liking–disliking, from the dimension of respect–disrespect when they speak about their preferred treatment from others (e.g., “You don’t have to like me, but you must respect me”).

Further, it is of theoretical interest to separate the contributions of rejection and disrespect interpretations to revenge goals. Baumeister and Leary (1995) posited that human beings have a strong need to be accepted and liked by others, and that individuals evolved to be sensitive to rejection and motivated to act in ways that increase acceptance by others. If perceptions of rejection are evolutionarily adaptive by motivating individuals to increase their acceptance, then individuals would be expected to act in more prosocial rather than aggressive ways when they feel rejected. However, as noted above, longitudinal and experimental studies have found rejection and aggression to be closely linked. How can this be reconciled?

Perhaps the answer is that in everyday life, and in some laboratory paradigms designed to study rejection, interpretations of rejection may often be accompanied by interpretations of disrespect. Furthermore, if the experience of being disrespected includes a perception that the interaction partner inappropriately believes himself or herself to have greater power or status than the perceiver and can treat the perceiver with disdain or disregard, revenge-seeking may be a means to reassert power or insure future fair treatment. Thus, when interpretations of disrespect accompany the interpretation of rejection, it may be that antisocial responses rather than prosocial responses are more likely. As a first step in examining this possibility, the present study examined the unique contributions of rejection and disrespect interpretations to revenge goals.

The Potential Influence of Anger and Hurt Feelings

Another major goal of the study was to investigate how anger and hurt feelings may contribute to desires for revenge. Lemerise and Arsenio (2000) called attention to the need to examine how emotions as well as cognitions affect the goals and strategies that individuals pursue in social interaction. Although anger has long been understood to be an emotion associated with revenge motivations and retaliation (e.g., Berkowitz, 1993), researchers recently have begun to examine the emotional experience of “hurt feelings” as well (e.g., Feeney, 2004; Leary, Springer, Negel, Ansell, & Evans, 1998). Hurt feelings can stem from rejection or other forms of harm that threaten positive self-views (Feeney, 2004; Leary et al., 1998). Hurt feelings, while a distinct experience, appears to be a form of negative affect that blends sadness and anger, with the latter especially occurring when the offense is seen as unjustified (Leary, Koch, & Hechenbleikner, 2001). In an investigation of events that led to hurt feelings, Leary et al. (1998) found that in 35 percent of episodes college students’ hurt feelings were attributed to either active or passive disassociation by another person, and in 33 percent of episodes hurt feelings were attributed to criticism (also see Feeney, 2004). Further, along with hurt feelings, college students reported that they felt and/or expressed anger (80 percent of episodes) toward their perpetrator and/or retaliated with a nasty or critical remark (62 percent of episodes), thereby linking hurt feelings with anger experiences and vengeful behaviors.

Drawing on this literature, we hypothesized that both anger and hurt feelings would be positively linked with revenge goals in conflicts of interest. Additionally, to examine if anger and hurt feelings were distinctive in their associations with revenge, we examined their relative associations with revenge goals. In addition, as interpretations and emotions both are likely to influence revenge goals (Crick & Dodge, 1994), we compared their relative associations with revenge goals.

Gender Differences

A final purpose of the study was to examine possible gender differences in the interpretations and emotional reactions college students make in conflicts of interest. Prior research has found that boys, more than girls, endorse revenge goals and aggressive strategies in conflict-of-interest situations with a friend (Rose & Asher, 1999; see Rose & Rudolph, 2006, for a broader review of gender differences in peer relationships). However, little research has focused on whether there are gender differences in the content of older adolescents’ interpretations or their emotional reactions in conflicts of interest. In the present study, we expected gender differences, not only in goals and strategies, but also in college students’ interpretations and emotional reactions, with males reporting more negative interpretations and affect than females.

Method

Participants

Participants were 157 college students (92 female) from a residential university in southeastern USA where students live in university housing for the majority of their college years. Participants came to the laboratory and completed all measures in small groups of no larger than five students. Students ranged in age from 17 to 22 years old, with the majority being 18 (42.3 percent) or 19 years old (33.0 percent); 86 percent of participants were in their first two years of college. Over 40 percent of participants were from racial/ethnic minority groups.

Conflict Vignettes

A hypothetical situations methodology was used to assess students’ responses to a variety of conflict-of-interest situations. With this methodology, features of situations can be controlled and a wide range of situations can be presented. Children’s responses to carefully constructed vignettes have been found to be meaningfully related to behavioral style, peer acceptance, and friendship quality in everyday life (e.g., Rose & Asher, 1999; Rubin, Daniels-Beirness, & Bream, 1984). In the present study, 15 vignettes were used to depict conflict-of-interest situations in three different relationship contexts: with a romantic partner, with a friend, and with a roommate. These 15 vignettes, adapted from Kaluk, Asher, and Parkhurst (2001), consisted of five different types of vignettes per relationship context. These vignette types represented five different types of conflict issues that friends, roommates, or romantic partners in college might encounter: (1) competing to use a resource, (2) needing help, (3) doing one’s fair share, (4) exclusivity, and (5) breaking a commitment to spend time together. So, for example, one of the vignettes involving competing to use a resource read as follows: “You and your roommate share a TV. Tonight you need to use the TV to watch a video that you missed seeing in class before your test tomorrow. However, your roommate also needs to use the TV to watch an election debate for his/her political science class. You really need the TV, but your roommate insists that he/she really needs it too”.

Participants were asked to read each of the 15 vignettes, imagine themselves in each situation, and for each situation to rate the likelihood that they would (1) make a variety of interpretations, (2) experience various emotions, (3) pursue a variety of goals, and (4) endorse various strategies (see Table 1 for the wording of each interpretation, emotion, goal, and strategy). All of the stories used in the study were designed to be applicable across all three relationship contexts, although different stories were used for each relationship context (with the constraint that the same five conflict issues were represented in each relationship context).¹ Furthermore, each type of interpretation, emotion, goal, and strategy was worded similarly for each story. For example, following all stories, the rejection interpretation was worded as “My [insert either ‘boyfriend/girlfriend’, ‘friend’, or ‘roommate’ here] is rejecting me”. Students made their ratings for each interpretation, emotion, goal, and strategy by making an “X” on a continuous line that was five and a half inches long with “strongly disagree” on the left and “strongly agree” on the right. Undergraduate assistants measured where participants marked the line in sixteenths of an inch increments. For the entire set of vignettes, please contact the authors.

Table 1.

Correlations of Interpretations, Emotions, Strategies, and Goals^a With the Number of People who Provide Social Support (SSQ-P), Satisfaction with Social Support (SSQ-S), and Social Provisions

	Examples	SSQ-P	SSQ-S	Social Provisions Scale
Interpretations
Disrespect	“My partner doesn’t respect me”	−.32***	−.24***	−.32***
Rejection	“My partner is rejecting me”	−.29***	−.23***	−.28***
Lack of caring	“My partner doesn’t care about my wants or needs”	−.32***	−.26***	−.26***
Betrayal	“My partner’s behavior is a kind of betrayal”	−.29***	−.25***	−.26***
Lack of viability	“My relationship with this person is falling apart”	−.29***	−.20**	−.28***
Moral judgment	“My partner’s behavior is wrong”	−.25***	−.20**	−.16
Solvability	“My partner and I can easily solve this situation”	.26***	.30***	.32***
Face value	“My partner just wants to use the TV”	.29***	.27***	.42***
Emotions
Anger	“I would feel angry”	−.24***	−.24***	−.17*
Hurt feelings	“My feelings would be hurt”	−.17*	−.17*	−.07
Goals
Revenge	“I would be trying to get back at my partner”	−.30***	−.13	−.26***
Relationship maintenance	“I would be trying to stay together”	−.01	.00	−.01
Fairness	“I would be trying to make sure that things are done fairly”	.16	.15	.24***
Tension reduction	“I would be trying not to get upset”	−.08	−.16	−.19*
Control	“I would be trying not to be pushed around”	−.12	−.10	−.10
Self-interest	“I would be trying to do what I want”	.01	−.04	.12
Strategies
Verbal aggression	“I would say something insulting to my partner”	−.23***	−.18*	−.26***
Self-interest assertions	“I would tell my partner to help me study now”	−.25***	−.13	−.14
Emotional manipulation	“I would act cold and distant to my partner”	−.32***	−.18*	−.27***
Threat of termination	“I would threaten to end the relationship”	−.28***	−.16*	−.32***
Termination	“I would end the relationship”	−.15	.01	−.10

Note: SSQ = social support questionnaire.

^aIllustrative examples are provided for each interpretation, emotion, goal, and strategy.

^*p < .05,

^**p < .01,

^***p < .001.

Social Support and Social Provisions

To assess the concurrent validity of the newly created conflict-of-interest vignettes, measures of social support and social provisions were also administered. Given previous evidence that revenge motivations are negatively related to relationship quality with friends (Rose & Asher, 1999), we hypothesized that revenge goals, as well as negative interpretations, negative emotions, and aggressive strategies, would be negatively associated with perceived social support and social provisions. The Social Support Questionnaire (SSQ) developed by Sarason, Levine, Basham, and Sarason (1983) was administered to identify the number of people who were perceived to provide support (SSQ-People; current sample α = .91) and to assess satisfaction with perceived support (SSQ-Satisfaction; current sample α = .87). The Social Provisions Scale (Cutrona & Russell, 1987; current sample α = .92) was administered to learn about the extent to which students believed that certain social provisions (i.e., attachment, social integration, nurturance, reliable alliance, reassurance, and guidance) were obtained from their relationships.

Results

Internal Reliabilities

Internal reliabilities were calculated for each specific subscale (e.g., the interpretation of disrespect, the emotion of anger, the goal of revenge, the strategy of relationship termination), first based on all 15 vignettes (i.e., pooled across the three relationship contexts), and then separately for each relationship context (five vignettes per context). For the data pooled across the three contexts, internal reliabilities ranged from .76 to .97, with all but two at .88 or above. When reliabilities were examined separately by relationship context, the internal reliabilities for accommodation and compromise strategies ranged from .33 to .51. Due to the lower internal reliabilities of these scales at the context level, these two strategies were removed from further analyses, leaving five rather than seven strategies. The reliabilities for the remaining interpretations, emotions, goals, and strategies at the relationship context level ranged from .73 to .93.

Correlations With Social Support and Social Provisions

Next, to assess the concurrent validity of the newly created conflict-of-interest vignettes, we examined how students’ responses to conflicts of interest related to the degree of social support and social provisions they experienced from relationships in their everyday lives (see Table 1). The conflict vignettes-based interpretations and emotions scales were significantly related to students’ perceived social support and provisions. For example, the more college students endorsed interpretations such as disrespect and rejection, and the more they reported emotional reactions of anger and hurt feelings, the less social support and social provisions they reported in their relationships. Three goals were also associated with social support and social provisions. The more individuals endorsed revenge goals, the less social supports and social provisions they perceived in their lives. Tension reduction goals (i.e., trying not to get upset) were also negatively related to reports of social provisions, and fairness goals were positively related to social provisions. Finally, various strategies, including verbal aggression, emotional manipulation, and threat of termination, were negatively associated with perceived social support and social provisions. These findings speak to the validity of the conflict vignettes measure in that responses to the vignettes were significantly associated with indexes of students’ general relationship adjustment.

Relationship Context and Gender

Next, we examined whether there were significant relationship context and gender differences in students’ interpretations, emotional reactions, goals, and strategies. For each interpretation, emotion, goal, and strategy, a 2 × 3 mixed analysis of variance was computed with gender serving as the between-subjects factor and relationship context serving as the within-subjects factor. Each table reporting analysis of variance results includes effect size estimates (partial eta-squared). No significant interactions were found between gender and relationship context.

Table 2 presents the results for gender. Males endorsed interpretations of disrespect, rejection, betrayal, and lack of relationship viability more than females, whereas females endorsed face value interpretations more than males. Males endorsed revenge goals at higher levels than females, whereas females rated fairness goals more highly than did males. Males endorsed verbal aggression and threatening to terminate the relationship more than females. There were no significant differences between males and females on emotion ratings.

Table 2.

Interpretations, Emotions, Goals, and Strategies as a Function of Gender

	Female		Male		F(1, 155)	ηp2
	M	SD	M	SD
Interpretations
Disrespect	.81	.64	1.03	.60	3.95*	.03
Rejection	.71	.60	.91	.64	4.00*	.03
Lack of caring	1.14	.73	1.33	.72	2.53	.02
Betrayal	.87	.62	1.09	.65	4.86*	.03
Lack relationship viability	.62	.47	.80	.51	5.46*	.03
Moral judgment	1.47	.75	1.63	.81	1.63	.01
Solvability	4.30	.76	4.08	.68	3.23	.02
Face value	4.36	.64	3.97	.66	13.55**	.08
Emotions
Anger	2.06	.89	2.17	1.04	.51	.00
Hurt feelings	1.73	.91	1.69	.92	.07	.00
Goals
Revenge	.63	.78	.92	.58	6.36*	.04
Relationship maintenance	4.41	1.02	4.23	.94	1.30	.01
Fairness	4.51	.94	4.07	.85	8.87**	.05
Tension reduction	2.72	1.32	2.56	1.22	.54	.00
Control	2.17	1.45	2.17	1.23	.01	.00
Self-interest	3.03	1.13	2.89	1.04	.69	.00
Strategies
Verbal aggression	.62	.56	.91	.60	9.09**	.06
Self-interest assertion	1.50	.92	1.75	.95	2.67	.02
Emotional manipulation	.83	.74	1.03	.73	2.63	.02
Threat of termination	.34	.32	.45	.31	4.90*	.03
Termination	.28	.22	.34	.26	2.53	.02

Note: Ratings ranged from 0 to 5.50 on a scale of inches.

^*p < .05,

^**p < .01.

Table 3 presents the results for relationship context. Bonferroni corrections were utilized when probing relationship context main effects. Main effects were found for each of the interpretations. Interpretations of disrespect, rejection, lack of caring, and lack of relationship viability were higher in romantic relationships than in the friendship and roommate contexts. Individuals also endorsed betrayal and moral judgment interpretations more highly in the romantic context, less in the friendship context, and least in the roommate context. Consistent with these findings, face value interpretations were lowest in the romantic context, and interpretations of the conflict being solvable were endorsed most highly in the roommate context, less in the friendship context, and least in romantic relationships.

Table 3.

Interpretations, Emotions, Goals, and Strategies as a Function of Relationship Context

	Romantic partner		Friend		Roommate		F(2, 310)	ηp2
	M	SD	M	SD	M	SD
Interpretations
Disrespect	.97a	.66	.88b	.65	.84b	.67	8.62**	.05
Rejection	.90a	.71	.76b	.63	.71b	.64	20.36**	.12
Lack of caring	1.34a	.80	1.19b	.78	1.12b	.76	17.26**	.10
Betrayal	1.07a	.74	.95b	.68	.86c	.66	15.86**	.09
Lack relationship viability	.80a	.58	.65b	.49	.64b	.52	19.63**	.11
Moral judgment	1.70a	.92	1.54b	.84	1.36c	.77	27.38**	.15
Solvability	4.12a	.79	4.20b	.81	4.31c	.75	12.26**	.07
Face value	4.05a	.73	4.27b	.73	4.28b	.71	23.55**	.13
Emotions
Anger	2.32a	1.04	2.10b	1.02	1.91c	1.00	29.20**	.16
Hurt feelings	2.06a	1.05	1.63b	.95	1.46c	.93	70.41**	.31
Goals
Revenge	.80a	.78	.69b	.71	.74ab	.75	7.72**	.05
Relationship maintenance	4.26a	1.09	4.35b	1.05	4.40b	.95	4.73*	.03
Fairness	4.32	.95	4.33	.97	4.33	.98	.12	.00
Tension reduction	2.85a	1.28	2.57b	1.33	2.53b	1.36	23.87**	.13
Control	2.28a	1.38	2.10b	1.41	2.12b	1.40	7.58**	.05
Self-interest	3.19a	1.12	2.97b	1.18	2.75c	1.19	28.19**	.15
Strategies
Verbal aggression	.83a	.72	.71b	.60	.68b	.59	10.07**	.06
Self-Interest assertion	1.78a	1.06	1.55b	1.02	1.50b	.98	11.19**	.07
Emotional manipulation	1.08a	.91	.83b	.75	.82b	.73	21.66**	.12
Threat of termination	.39ab	.32	.41a	.37	.36b	.33	4.35*	.03
Termination	.31	.25	.31	.24	.29	.24	1.95	.01

Note: Within rows, means with different superscripts are significantly different from one another using a Bonferroni correction. Ratings ranged from 0 to 5.50 on a scale of inches.

^*p < .05,

^**p < .01.

With regard to the emotions that participants reported that they would experience, college students were angrier and more hurt in the romantic context than they were in the other two contexts. They were also more angry and more hurt in the friend context than in the roommate context.

There were also relationship context differences in goals. Revenge, tension reduction, control, and self-interest goals were endorsed more in the romantic context than with friends or roommates. Additionally, relationship maintenance goals were less strongly endorsed in the romantic context than in the friendship and roommate contexts.

Finally, there were also relationship context differences in the strategies that participants endorsed. Participants endorsed verbal aggression, self-interest assertion, and emotional manipulation more highly in the romantic relationship context than with friends and roommates. Further, threatening to terminate the relationship was more highly endorsed in the friendship context than in the roommate context.

Associations between Interpretations, Emotions, Goals, and Strategies

In order to examine how interpretations, emotions, goals, and strategies were related to each other, correlations were computed pooled across contexts and separately for each relationship context. Fisher’s r-to-Z Pearson–Filon statistic (Raghunathan, Rosenthal, & Rubin, 1996) was used to test whether associations differed across relationship context. In no case was a correlation found to significantly differ from one context to another. Accordingly, for space reasons, only the correlations based on data pooled across contexts are presented (see Table 4).

Table 4.

Correlations Among Interpretations, Emotions, Goals, and Strategies

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20
	Interpretations								Emotions		Goals						Strategies
Interpretations
1. Disrespect
2. Rejection	.83
3. Lack of caring	.83	.78
4. Betrayal	.80	.87	.76
5. Lack of relationship viability	.88	.79	.76	.73
6. Moral judgment	.69	.68	.81	.77	.65
7. Solvability	−.57	−.56	−.63	−.59	−.58	−.63
8. Face value	−.53	−.44	−.40	−.40	−.51	−.31	.51
Emotions
9. Anger	.51	.54	.66	.64	.49	.78	−.62	−.23
10. Hurt feelings	.60	.62	.70	.68	.53	.74	−.55	−.31	.79
Goals
11. Revenge	.56	.51	.44	.52	.55	.40	−.34	−.43	.31	.33
12. Relationship maintenance	−.15	−.13	−.11	−.10	−.19	−.15	.26	.24	−.14	−.10	−.09
13. Fairness	−.20	−.22	−.19	−.19	−.20	−.16	.34	.39	−.10	−.13	−.17	.50
14. Tension reduction	.32	.28	.31	.31	.31	.32	−.18	−.21	.31	.31	.27	.21	.22
15. Control	.43	.41	.39	.39	.41	.49	−.26	−.22	.44	.45	.43	.15	.13	.66
16. Self-interest	.10	.17	.23	.23	.12	.33	−.17	.07	.34	.24	.09	−.08	.13	.18	.32
Strategies
17. Verbal aggression	.70	.70	.62	.66	.74	.60	−.58	−.44	.56	.49	.65	−.20	−.22	.27	.38	.18
18. Self-interest assertion	.48	.52	.54	.55	.47	.60	−.49	−.23	.51	.35	.37	−.10	−.02	.25	.45	.52	.57
19. Emotional manipulation	.67	.66	.64	.68	.66	.65	−.52	−.43	.60	.65	.62	−.15	−.33	.21	.36	.10	.72	.40
20. Threat of termination	.61	.58	.48	.52	.68	.38	−.41	−.48	.30	.28	.59	−.16	−.16	.25	.36	.04	.71	.43	.57
21. Termination	.42	.46	.36	.45	.53	.35	−.36	−.32	.24	.25	.32	−.16	−.10	.19	.27	−.04	.52	.26	.41	.74

Note: Correlations greater in magnitude than .16 are significant at p < .05. Correlations greater in magnitude than .21 are significant at p < .01.

The correlations between students’ goals and their strategies replicated those found in previous research with children (e.g., Rose & Asher, 1999). For example, revenge goals were positively related to verbal aggression, whereas relationship maintenance and fairness goals were negatively related to verbal aggression.

Unique to this study is the examination of interpretations and how interpretative content relates to people’s goals in conflict-of-interest situations. As can be seen in Table 4, interpretations of disrespect, rejection, lack of caring, betrayal, lack of relationship viability, and moral judgment were positively associated with revenge, tension reduction, and control goals. By contrast, solvability and face value interpretations were positively related to relationship maintenance and fairness goals, and were negatively associated with revenge, tension reduction, and control goals.

Additionally, the examination of emotions revealed that both anger and hurt feelings were positively correlated with all of the negative interpretations, including disrespect and rejection interpretations, as well as with the goals of revenge, tension reduction, control, and self-interest, and with the strategies of verbal aggression, self-interest assertion, emotional manipulation, threat of relationship termination, and relationship termination.

The Contributions of Interpretations and Emotions to Revenge Goals

Next, two sets of hierarchical regression analyses were conducted to compare how interpretations and emotions predict to revenge goals. Within each set, four regressions were conducted, with one analysis per relationship context and one analysis for variables pooled across all contexts. Three specific research questions were addressed: (1) What are the relative contributions of disrespect and rejection interpretations to the pursuit of revenge goals? (2) What are the relative contributions of anger and hurt feelings to the pursuit of revenge goals? and (3) What are the relative contributions of interpretations and emotions to the pursuit of revenge goals? To address the first question, the first set of analyses (reported in Table 5) was conducted in which gender was entered in step 1, and thereby controlled for, interpretations were entered in step 2, and emotions were entered in step 3. To address the second question, a parallel second set of hierarchical regression analyses (reported in Table 6) was conducted in which gender was again controlled for in step 1, but emotions were entered in step 2, and interpretations were entered in step 3. To address the third research question, both sets of hierarchical analyses were examined, comparing the contributions of emotions and interpretations to revenge goals.

Table 5.

Hierarchical Regressions Predicting Revenge Goals From Gender, Rejection and Disrespect Interpretations, and Emotional Reactions

Step	Predictor	Roommate			Best friend			Romantic partner			Pooled across contexts
		β	R2	ΔR2	β	R2	ΔR2	β	R2	ΔR2	β	R2	ΔR2
1	Gender	.17*	.03*	.03*	.21**	.04*	.04**	.20*	.04*	.04*	.20*	.04*	.04*
2	Rejection	.10	.35***	.32***	.15	.30***	.26***	.27*	.28***	.24***	.15	.33***	.29***
	Disrespect	.49***	—	—	.39***	—	—	.26*	—	—	.41**	—	—
3	Anger	.11	.35***	.01	.01	.31***	.00	.02	.28***	.00	.03	.33***	.00
	Hurt feelings	−.09	—	—	−.09	—	—	.09	—	—	−.03	—	—

^*p < .05,

^**p < .01,

^***p < .001.

Table 6.

Hierarchical Regressions Predicting Revenge Goals From Gender, Emotional Reactions, and Rejection and Disrespect Interpretations

Step	Predictor	Roommate			Best friend			Romantic partner			Pooled across contexts
		β	R2	ΔR2	β	R2	ΔR2	β	R2	ΔR2	β	R2	ΔR2
1	Gender	−.17*	.03*	.03*	.21**	.04*	.04**	.20*	.04*	.04*	.20*	.04*	.04*
2	Anger	.21	.16***	.13***	.12	.12***	.08***	.06	.16***	.12***	.08	.16***	.12***
	Hurt feelings	.18	—	—	.17	—	—	.30**	—	—	.28*	—	—
3	Rejection	.09	.35***	.19***	.19	.31***	.19***	.23*	.28***	.12***	.15	.33***	.17***
	Disrespect	.49***	—	—	.41***	—	—	.24*	—	—	.42***	—	—

^*p < .05,

^**p < .01,

^***p < .001.

The Independent Associations of Disrespect and Rejection Interpretations With Revenge Goals

Combined across contexts, rejection and disrespect interpretations were highly related (r = .83, p < .001; see Table 4). Nonetheless, as seen in Table 5, in all contexts, disrespect interpretations remained significantly related to revenge goals, even with gender and rejection interpretations in the model. By contrast, rejection interpretations uniquely contributed to revenge goals only in the romantic context.

The Independent Associations of Anger and Hurt Feelings With Revenge Goals

As seen in Table 4, anger and hurt feelings were highly related emotions (r = .79, p < .001). Still, it was of interest to learn whether these emotional experiences were independently associated with revenge goals. Table 6 presents the hierarchical regression analysis aimed at learning whether anger and hurt feelings uniquely contributed to revenge goals (with gender statistically controlled). In both the roommate and friend contexts, the two emotions of anger and hurt feelings improved the prediction model, however, neither emotion alone independently related to revenge goals. By contrast, in the romantic partner context, hurt feelings positively contributed to revenge goals, even with anger and gender in the model. This was also the case when vignettes were pooled across contexts.

Comparing the Contributions of Interpretations and Emotions to Revenge Goals

The hierarchical regression analyses in Tables 5 and 6 were used to examine the relative contributions of interpretations and emotions to the pursuit of revenge goals. The findings in Table 6 show that anger and hurt feelings together contributed to the variance in revenge goal pursuit, beyond the contribution of gender, with ΔR² ranging from a low of .08 for the friend context to a high of .13 for the roommate context. When rejection and disrespect interpretations were added in the third step, their contribution was equal or greater (depending on relationship context) to that of emotions, with ΔR² ranging from .12 to .19. More importantly, as seen in Table 5, when the interpretations of rejection and disrespect were entered in the second step (following gender), these two interpretations contributed greatly to the variance in revenge goals, with ΔR² ranging from .24 to .32, and when anger and hurt feelings were added in the third step, they added little to no variance in the prediction of revenge goals beyond the predictions made by interpretations.

Discussion

College is an important developmental context in which youth have opportunities to develop different types of new, close relationships with peers. A major goal of the present study was to examine how college students react in conflict-of-interest situations with romantic partners, friends, and roommates. In line with our hypotheses, students indicated that they would make more negative interpretations, experience more negative emotions, and pursue more hostile goals and strategies with romantic partners than with friends or roommates. In the romantic relationship context, college students were more likely to make negative interpretations about their partner’s behavior and about the likely stability/longevity of their relationship. They were also more likely to be angry and hurt, and to endorse hostile goals and strategies with romantic partners compared to with friends and roommates. This pattern held for females as well as males (there were no significant gender by relationship context interactions), and overall, this pattern suggests that conflicts with romantic partners are more threatening for college students compared with conflicts with friends or roommates. This can be understood as the consequence of romantic relationships being a particularly important context for young adults as they seek to establish intimate and lasting connections with potential partners. Expectations for romantic relationships involve passion and commitment (Connolly et al., 1999), and older adolescents may develop more idealized beliefs and expectations for their romantic relationships than for their friendships or roommate relationships (e.g., Flannagan et al., 2005; Hazan & Shaver, 1987). Higher expectations for a relationship partner could lead someone to interpret a partner’s behavior in a conflict more negatively and to feel more hurt and angry as well. The exploration of how expectations differ across relationship contexts, and how expectations relate to interpretations, emotions, and revenge goals, should prove fruitful.

The present study also examined how interpretations in conflicts-of-interest are related to revenge goals. Confirming our hypotheses, various types of interpretations were associated with revenge goals. Specifically, revenge goals were found to be positively related to interpretations of rejection, disrespect, lack of caring, betrayal, lack of relationship viability, and judgments of wrongdoing. Revenge goals were also negatively related to the more optimistic interpretation of the conflict being resolvable and to the benign interpretation of the partner meaning no harm (i.e., a “face value” interpretation). These correlational findings support the thesis that an individual’s interpretations of how the partner views the individual and their relationship help explain revenge motivations in response to conflict.

Based on prior theory and research about aggression, we gave particular attention to rejection and disrespect interpretations, comparing their relative prediction to revenge goals. Although disrespect and rejection interpretations were highly related, results from the regression analyses indicated that they have non-overlapping components that may uniquely contribute to revenge goal endorsement. With friends and roommates, interpretations of disrespect remained significantly associated with revenge goal endorsement even with rejection interpretations in the model, but rejection interpretations did not make an independent contribution to revenge goals beyond the contribution made by disrespect interpretations. Only within the romantic context were both rejection and disrespect interpretations independently related to revenge goals when controlling for the other. One possible explanation for this is that interpretations of rejection may be more likely to lead to negative expectations about the future of the relationship with romantic partners than with friends or roommates. In turn, negative evaluations about the relationship may decrease relationship commitment making individuals more likely to react in a hostile manner, such as seeking revenge (Finkel et al., 2002). However, casting doubt on this explanation, results did not indicate that rejection and lack of viability interpretations were more strongly correlated with each other in romantic contexts compared with friend and roommate contexts. Clearly, more research is needed to understand why rejection is uniquely related to revenge goals in romantic relationships but not in other types of relationships.

It is interesting to consider why perceptions of being disrespected by a relationship partner would lead to desires for revenge. As noted in the introduction, feeling disrespected may include perceptions that the partner inappropriately assumes that he or she has higher status and more power within the relationship, and can act with disregard for the perceiver’s feelings. Retaliating or seeking revenge in this situation may be a way to correct the partner’s assumptions of a power imbalance within the relationship. Furthermore, the fundamental unfairness of the partner’s assumption may underlie retaliatory responses (see Richman & Leary, 2009, for further discussion about the role of perceived unfairness).

The findings that the unique components of disrespect may be more related to revenge than the unique components of rejection have potential implications for some paradigms used to study how rejection experiences are related to aggression. Experimental paradigms that involve rejecting participants through exclusion have often included explicit or implicit criticism as an explanation for the exclusion (e.g., Buckley et al., 2004; Twenge, Baumeister, Tice, & Stucke, 2001; for an example of an exception, see Williams, Cheung, & Choi, 2000). Due to the disdainful manner with which some of these criticisms are delivered (e.g., participants received feedback about an essay they had written that said, “One of the worst essays I’ve read!”), some experimental manipulations may not only elicit a sense that one has been rejected but also a sense that one has been disrespected. If so, it is unclear whether it is perceived rejection or perceived disrespect that is linked with subsequent aggression. Our findings suggest the testable hypothesis that if individuals were to experience “pure” rejection, they would be less likely to respond vengefully than in situations in which rejection is accompanied by the experience of disrespect. If a rejection is given in a way that communicates respect (rather than disdain), then it may be less likely that a person would respond vengefully.

Research is also needed with children to learn how interpretations of rejection and disrespect are related to vengeful behavior. Correlational evidence exists that peer rejection in childhood, measured sociometrically, is associated with later aggressive behavior (e.g., DeRosier et al., 1994). However, it is not possible to know from sociometric measurements how disliked children are actually treated by their peers in their everyday lives. Sometimes, disliked children are not invited to join activities or are avoided by peers, yet peers may still treat them respectfully or even deferentially. Other times, disliked children may be included in activities but treated disrespectfully by peers (harshly teased or criticized). At still other times, disliked children may be both excluded and disrespected by peers. Observational research of in vivo interactions and conversations using audiovisual wireless transmission methodology (e.g., Asher, Rose, & Gabriel, 2001; Pepler & Craig, 1995), combined with self-report data on interpretations, is needed on the experiences disliked children have to better understand how specific types of social interactions and interpretations contribute to increases in aggressive behavior over time.

Moreover, individual differences in interpretational tendencies may affect whether disrespect interpretations and rejection interpretations are made. For example, rejection sensitivity is a social-cognitive bias involving the tendency to expect rejection from others (Downey & Feldman, 1996) and has been found to be associated with aggressive behavior (e.g., Downey, Lebolt, Rincon, & Freitas, 1998). Some individuals may be characterized by a parallel type of sensitivity, disrespect sensitivity (McDonald, 2008), that increases their tendency to interpret social information as disrespectful. McDonald (2008) found that this type of sensitivity was positively associated with revenge goals and aggressive strategies in conflict situations.

Another purpose of the present study was to examine how anger and hurt feelings are related to revenge goals. Although reports of anger and hurt feelings were similarly correlated with revenge goals in conflict-of-interest situations, when the unique contributions of anger and hurt feelings were examined in a regression analysis, hurt feelings uniquely contributed to revenge goals in conflicts with romantic partners and anger did not. However, neither emotional experience independently contributed to revenge goals with friends or roommates. In previous research, Leary et al. (1998) found that feelings of anger are often embedded in the experience of hurt feelings. Our results could mean that, in some relationship contexts (i.e., with friends and roommates), the component of hurt feelings that is associated with revenge goals is that of anger, whereas in other contexts (i.e., with a romantic partner), other dynamics are at play. It is important to learn whether this unanticipated finding replicates and, if it does, to investigate the processes that might give rise to this pattern.

An issue of great interest concerned the relative contributions of interpretations and emotions to revenge goals. Analyses revealed that disrespect and rejection interpretations contributed more variance, and more unique variance, to revenge goals than did the emotional experiences of anger and hurt feelings. These results imply that interventionists who aim to decrease revenge motivations may find that focusing on the broad range of interpretations youth make in conflict will be more productive than targeting emotion regulation alone.

The present study found noteworthy gender differences not just in goals and strategies, but also in the types of interpretations people make about conflict-of-interest situations. Males’ greater endorsement of rejection, disrespect, and betrayal interpretations, and their greater pessimism about the viability of relationships appear to set the stage for their more negative goals and strategies. Also of interest was the finding that males and females did not differ in their reports of anger or hurt feelings even though they differed in the interpretations they made. Viewed together with the findings that interpretations are uniquely related to revenge goals beyond the effect of emotions, and that emotions are not related to revenge goals beyond the effect of interpretations, the findings of gender differences in interpretations but not emotions suggest that interpretational processes may play a stronger role than emotional reactions in giving rise to revenge goals in conflict-of-interest situations. This possibility merits future investigation.

There is also a need to examine the linkages among conflict interpretations, emotions, and goals with other age groups and in contexts that differ in norms about aggression. For example, younger children, compared with older adolescents, may place less importance on respect and may not be as sensitive to (“tuned in to”) cues of disrespect. In addition, it is possible that in school/community contexts where aggressive behavior is more normative, youth may be more vigilant for signs of disrespect and believe that getting even is necessary to maintain status with peers (Anderson, 1999).

Future research could also address three methodological issues. First, although internal reliabilities were generally excellent for the wide range of vignette-based constructs, the internal reliabilities for accommodation and compromise strategies at the context level were lower and had to be dropped from analyses. These two strategies were the most highly endorsed, perhaps suggesting that they were the strategies that participants had the most experience with in their everyday lives. If participants more commonly compromise and accommodate in their everyday conflicts, they may find that their success depends on many conflict-specific contextual factors. This idea follows from Runions and Keating’s (2007) suggestion that responses to vignettes may reflect individuals own “idiosyncratic history with similar scenarios (p. 846)”. Thus, experiences using these strategies may lead individuals to develop more nuanced, situation-specific expectations of when it is best to accommodate or compromise, and this could be reflected in the low internal reliabilities observed herein.

Second, even though the new conflict-of-interest vignettes showed concurrent validity with regard to social support and social provisions, these associations should be interpreted cautiously. All of the data came from the research participants themselves, and it is possible that a general, negative worldview could explain why college students who report low amounts of support also report more negative interpretations in conflict. In future research, participants’ relationship partners, rather than the participants themselves, could be asked to report on the extent to which they provide social support and social provisions to the participant (see Parker & Asher, 1993, for an example of independent assessment).

Third, in order to examine within-person differences across relationship contexts, we used vignettes that involved the same conflict issues but did not have exactly the same content. We did not use the same stories for each type of relationship partner because research participants may have inferred that the research question pertained to how their responses changed in different relationship contexts. Fortunately, we found that the stories we created were not judged to be significantly different in severity across contexts. Still, in future research a between-groups design could be used by randomly assigning participants to receive vignettes in one of the three relationship contexts. In this way, the same vignettes could be used across contexts, varying only the type of relationship partner in the vignette (i.e., roommate, friend, romantic partner) and eliminating any variability in scenario content across relational contexts. We expect that the present findings of context differences would replicate.

Finally, although rejection and disrespect interpretations were the primary focus here, a broad array of interesting interpretations was related to revenge goals. This highlights the importance of understanding the multifaceted content of individuals’ interpretations during conflict (e.g., Bradbury & Fincham, 1990; Dodge, 1980). All of the assessed interpretations deserve further examination as potential within-relationship “triggers” for negative interpersonal behavior. Focusing research more intensely on the interpretative processes people engage in during social interaction should help us better understand vengeful behavior in conflict-of-interest situations with close relationship partners.