Abstract
When does power in intimate relationships shape important interpersonal behaviors, such as psychological aggression? Five studies tested whether possessing low relationship power was associated with aggressive responses, but (1) only within power-relevant relationship interactions when situational power was low, and (2) only by men because masculinity (but not femininity) involves the possession and demonstration of power. In Studies 1 and 2, men lower in relationship power exhibited greater aggressive communication during couples’ observed conflict discussions, but only when they experienced low situational power because they were unable to influence their partner. In Study 3, men lower in relationship power reported greater daily aggressive responses toward their partner, but only on days when they experienced low situational power because they were either (a) unable to influence their partner or (b) dependent on their partner for support. In Study 4, men who possessed lower relationship power exhibited greater aggressive responses during couples’ support-relevant discussions, but only when they had low situational power because they needed high levels of support. Study 5 provided evidence for the theoretical mechanism underlying men’s aggressive responses to low relationship power. Men who possessed lower relationship power felt less manly on days they faced low situational power because their partner was unwilling to change to resolve relationship problems, which in turn predicted greater aggressive responses to their partner. These results demonstrate that fully understanding when and why power is associated with interpersonal behavior requires differentiating between relationship and situational power.
Keywords: relationship power, situational power, aggression, dependence, relationship conflict, support
People possess social power when they are able to control or influence others’ desired outcomes, and they lack power when their needs and goals are dependent on the actions and preferences of others (Keltner, Gruenfeld & Anderson, 2003; Thibaut & Kelley, 1959). Power features particularly strongly in romantic relationships because people’s goals, desires and happiness inevitably depend on their partner’s cooperation and investment (Kelley & Thibaut, 1978). But, goals and desires often conflict in relationships, which produces the need to influence the partner’s thoughts and behaviors to reach one’s own desired outcomes (Rusbult & Van Lange, 2003). Partners are also a primary source of comfort and support, but this places people in the vulnerable position of depending on their partner to be responsive in times of need (Kelley et al., 2003). Thus, power dynamics are central to many relationship interactions, including situations in which (a) people need to influence their partner to achieve desired goals, such as during conflict, and (b) people are dependent on their partner for the fulfilment of core needs, such as when they need support (Kelley et al., 2003).
The difficulty and importance of these power-relevant situations—those that involve the need to influence or depend on the partner—is shown by the mass of research examining how couples can best resolve conflict and support one another. Yet, the ways in which power shapes people’s responses within these key situations has been largely overlooked. This is a significant oversight because research in non-romantic contexts has shown that threats to power often lead to aggression to assert and restore power (e.g., Bugental, 2010; Bugental & Lin, 2001; Case & Maner, 2014; Georgesen & Harris, 2006; Fast & Chen, 2009; Maner & Mead, 2010). In contrast, prior research examining the links between power and aggression in romantic relationships has produced mixed findings. We propose that these inconsistencies emerge for two reasons. First, the effects of relationship power (a) must be examined within actual power-relevant situations in which people need to influence or depend on their partner, and (b) will also vary according to how much power (influence or dependence) people are currently experiencing within those power-relevant situations. We predict that low relationship power will produce aggressive responses to restore power, but only when people experience low situational power because they are unable to influence their partner or are dependent on their partner for core relational needs, such as support. Second, the importance and thus effects of power differ across men and women. In particular, because masculinity (but not femininity) involves the possession and demonstration of power, we predict that the effects of low relationship and situational power will occur for men and not women.
Relationship Power and Aggression
Relational dependence and influence are central to the way power is conceptualized and measured in close relationships (Kelley & Thibaut, 1978; Kelley et al., 2003). Although partners are mutually dependent, asymmetries in levels of dependence commonly result in one partner possessing more power (Attridge, Berscheid & Simpson, 1995; Felmlee, 1994; Loving et al., 2004; Sprecher, Schmeeckle, & Felmlee, 2006). According to the principle of least interest (Kelley & Thibaut, 1978; Waller & Hill, 1951), the person who is less invested in the relationship, and thus less affected by their partner’s actions, possesses relatively greater levels of power to influence their partner and produce relationship outcomes in their favor (Oriña et al., 2011; Sprecher et al., 2006). In contrast, the person who is more invested, and whose goals and happiness are more dependent on the relationship, is less able to exert influence because the rewards and punishments he or she can enact do not hold as much sway over the partner’s desired outcomes (Kelley & Thibaut, 1978; Simpson et al., 2013).
Thus, low power in relationships involves being more dependent on the partner and being less able to influence the partner to achieve desired outcomes. For this reason, low relationship power has been assessed in two ways: (1) relative dependence and (2) perceived ability to influence the partner. Research using these two assessment methods have provided converging evidence that people who have low relationship power face greater difficulties in getting their relational needs and desires met. Intimates higher in power are less motivated to sacrifice and support their partner (Righetti et al., 2015; Van Willigen & Drentea, 2001) and place less value on the support they receive from their partner (Inesi, Gruenfeld & Galinsky, 2012). Accordingly, intimates lower in power have less control over important relationship decisions (Farrell, Simpson & Rothman, 2015; Grady et al., 2010; Vanderdrift et al., 2013), are less secure and confident with regard to their relationship (Caldwell & Peplau, 1984; Sprecher et al., 2006), and react more negatively when they face challenges in their relationship (e.g., Attridge et al., 1995; Berman & Frazier, 2005; Karremans & Smith, 2010; Kuehn, Chen & Gordon, 2015). People with lower relationship power also experience lower relationship satisfaction and subjective well-being (e.g., Gray-Little & Burks, 1983; Kifer, Heller, Perunovic, & Galinsky, 2013; Sprecher & Felmlee, 1997; Sprecher et al., 2006).
In contrast to these consistent effects, the associations between relationship power and aggression are less clear. Some studies support that the difficulties associated with low relationship power, and the need to redress power imbalances, spur aggression as a means for powerless individuals to assert and gain power (Bugental, 2010; Bugental & Lin, 2001; Worchel et al., 1978).1 For example, felt dependence is associated with more aggressive motives and behaviors (Bornstein, 1996; Dutton, 1995; Murphy, Meyer, & O’Leary, 1994), and lower perceived influence is associated with greater self-reported aggression (e.g., Babcock, Waltz, Jacobson & Gottman, 1993; Sagrestano, Heavey & Christensen, 1999). Yet, influential reviews of the literature have concluded that the links between relationship power and aggression “are mixed and often contradictory” (Holtzworth-Munroe, Bates, Smutzler, & Sandin, 1997, p. 83; also see Bornstein, 1996; Hotaling & Sugarman, 1986). Studies since these reviews have also revealed mixed findings. Relationship power has been negatively (e.g., Kaukinen, 2004; Rogers, Bidwell, & Wilson, 2005; Sagrestano et al., 1999) and positively (e.g., Bentley, Galliher, & Ferguson, 2007) associated with self-reported aggression, and these significant effects occur alongside as many (if not more) null associations between power and aggression, often within the same data set (e.g., Babcock et al., 1993; Bentley et al., 2007; Rogers et al., 2005; Ronfeldt, Kimerling, & Arias, 1998).
We propose there are two central reasons for these inconsistencies. First, prior research has overlooked that the effects of power on aggression should emerge when people are experiencing low power in important power-relevant situations, such as when people are unable to influence their partners or are highly dependent on their partners to satisfy key needs and goals. Second, prior research has offered an inconsistent treatment of gender differences, but the loss of power should be particularly challenging for men rather than women. We elaborate on these two factors in the following sections.
Power in Context: The Role of Situational Power
Prior research has typically considered how general perceptions of dependence or influence across the relationship—what we refer to as relationship power—shape aggression across prior relationship interactions. However, possessing low relationship power should be most problematic and produce aggressive responses within power-relevant situations in which people must influence their partner to attain desired outcomes or are dependent on their partner for important needs (Huston, 1983; also see Bugental & Lin, 2001). For example, low power intimates tend to exhibit greater physiological stress (Loving et al., 2004) and more dominant communication (Dunbar & Burgoon, 2005) during conflict discussions in which partners are trying to influence or resist influence from each other.
Moreover, even in specific power-relevant situations people’s ability to influence the partner and levels of dependence will vary, and will do so regardless of general levels of relationship power (Kelley et al., 2003). For example, people’s experience of power in the moment can be experimentally primed (e.g., recalling a situation of power) revealing situational expressions of power independent of global levels (e.g., Chen et al., 2001; Galinsky et al., 2003, 2006; Inesi et al., 2012). Daily assessments of power also demonstrate that people’s perceptions of power in their relationship vary across days (Gordon & Chen, 2013), and feeling less powerful on a given day predicts more negative emotional reactions to partners’ hostility (Kuehn et al., 2015). These situational shifts in power reflect the interdependent nature of relationships: intimates regularly encounter power-relevant situations that vary in their level of influence and dependence and thus experience situation-specific differences in power—what we refer to as situational power (Kelley et al., 2003).
Our contextual analysis of power suggests that aggressive responses within relationship interactions will be a function of both the power people generally hold in their relationship (relationship power) and their experience of power within specific power-relevant situations, such as when attempting to influence their partner or when dependent on their partner for the fulfillment of core needs (situational power). This distinction is crucial because low power intimates are unlikely to behave in aggressive ways to restore power in all relationship interactions (Shaver, Segev, & Mikulincer, 2011). Instead, low relationship power should only be problematic and lead to aggressive responses when individuals are in situations in which their lack of power has detrimental consequences, such as when they are unable to influence the partner or are dependent on the partner to attain key needs and goals (i.e., when situational power is low). In contrast, when situational power is not undermined, then low power partners will have less need to redress lack of power in aggressive ways.
Prior research has neglected the distinction between relationship and situational power, which we think contributes to the array of inconsistent findings regarding power and aggression. For example, most prior studies have relied on self-reported aggression assessed using the Conflict Tactics Scale (Straus, Hamby, Boney-McCoy, & Sugarman, 1996), which asks people to report on the frequency of aggression over the past year. Such self-reports gloss over the contexts in which power is consequential and fail to capture differences in the experience of power within power-relevant situations. It is also not enough to just examine the links between relationship power and aggression within power-relevant interactions (e.g., Sagrestano et al., 1999) without taking into account the degree to which people lack influence or are highly dependent within that interaction. If power-relevant interactions are not characterized by low influence or high levels of dependence, then low power partners will not experience the need or desire to respond aggressively in order to rebalance power.
In the current studies, we provide the first tests of the interaction between relationship power and situational power by (1) examining the associations between relationship power and aggressive responses within couples’ power-relevant interactions, and (2) assessing the degree to which individuals are facing low situational power because they are unable to influence their partner or are highly dependent on their partner within those interactions. We predicted that possessing low relationship power would be associated with greater aggressive responses, but only when low power partners were experiencing low situational power and not when situational power was high and thus there was no need to try to assert or establish power via aggression. As discussed next, we also expected this pattern to differ by gender.
Gender, Power and Aggression
If aggression serves as a means to redress power imbalances, aggressive responses should be exhibited most by people who find the loss of power particularly challenging. One principal factor that captures different sensitivities to the loss of power is gender because masculinity (but not femininity) involves possessing and demonstrating power (Bosson & Vandello, 2011; Kimmel, 2008; Vescio, Schlenker, & Lenes, 2010). Masculine traits include being independent, self-reliant, assertive, and dominant (Bem, 1974, 1981), and masculine norms involve having greater power within work and family roles, particularly in comparison to women (Thompson & Pleck, 1986; Vescio, et al., 2010). Masculinity also involves being physically, mentally and emotionally tough (Thompson & Pleck, 1986; Vescio, et al., 2010). Masculine norms emphasize that a man should “…not disclose pains”, “stand on his own two feet”, “never back down” (Levant et al., 1992), and should not be dependent or weak because “asking for help is a sign of failure” (Mahalik et al., 2003). Thus, the loss of power due to either a lack of influence or high levels of dependence threatens masculinity.
Masculinity also involves assertive and forceful responses when power (and thus masculinity) is threatened. This is because the central components of masculinity—being powerful, influential, tough and independent—can only be achieved if they are acknowledged by others (Kimmel, 2008; Vescio, et al., 2010). These social requirements of ‘being a man’ mean that masculinity is precarious; masculinity can be easily lost when power is threatened and so it must be actively demonstrated to others (see Bosson & Vandello, 2011; Kimmel, 2008; Vescio, et al., 2010). Not only does this render men sensitive to threats to their power, the need to prove masculinity via the demonstration of power promotes active, overt and aggressive displays to restore masculinity when power and masculinity is threatened. Indeed, aggression offers a clear demonstration of power, toughness and independence and thus is an effective way to reestablish masculinity (Bosson & Vandello, 2011; Vandello et al., 2008).
There is a great deal of evidence that threats to masculinity lead to aggressive responses by men to demonstrate and repair masculinity. In these studies, masculinity threat is experimentally induced by, for example, giving false feedback to men that they know less than the average man or are more like the average women, asking men to engage in feminine tasks, or men being outperformed by a woman (e.g., Bosson et al., 2009; Dahl, Vessio & Weaver, 2015; Glick et al., 2007; Maass, Cadinu, Guarnieri & Grasselli, 2003; Vandello et al., 2008; Weaver & Vescio, 2015). Such masculinity threats lead to public discomfort and anger by men and, in turn, a range of aggressive responses to restore masculinity, such as derogation of women or effeminate men, greater endorsement of ideological dominance, greater sexual harassment, and greater aggressive cognitions and behaviors in experimental tasks. In contrast, women do not react aggressively to threats to their femininity because being a women does not require demonstrations of power (Bosson & Vandello, 2011).
The centrality of power to masculinity and the evidence that masculinity threat instigates aggression to restore masculinity indicate that men will be more likely than women to respond aggressively to low power situations in their relationships. There is some evidence for this hypothesis. For example, men who have low social power (e.g. low income, low status jobs, unemployed) are more likely to physically aggress in their relationships (Magdol et al., 1997; Straus & Gelles, 1990). However, the links between relationship power and aggression are mixed. Some studies report similar associations between power and aggression for men and women (e.g., Bentley et al., 2007; Choi & Ting, 2008; Leonard & Senchak, 1996; Ronfeldt et al., 1998) whereas others have found low perceived influence or decision-making power to be linked with self-reported aggression for men, but not women (e.g., Babcock et al., 1993; Rogers et al., 2005; Sagrestano et al., 1999). However, none of these prior studies tested whether there were significant gender differences, and many of the existing studies also report several null effects for both men and women.
A key reason for these inconsistent patterns is that, as we have argued above, aggressive responses need to be examined within important power-relevant situations when low power is acute and consequential. If displays of aggression operate to restore masculinity, then aggression should emerge when masculinity is threatened, and masculinity should be most threatened within situations in which men experience low levels of power because they lack influence or are highly dependent on their partners. Accordingly, testing whether low power activates manhood-restoring aggression requires assessing the links between relationship power and aggression within interactions (1) in which men lack situational power, and (2) when aggressive responses could be enacted to reestablish power and restore masculinity. We performed exactly these tests in the current research.
Research Overview
The current research was designed to reconcile the inconsistent links between power and aggression in relationships by distinguishing between the power individuals generally possess in their relationship (relationship power) and the power they experience within important power-relevant situations in which individuals need to influence or depend on their partner to attain desired needs and goals (situational power). We expected that possessing low relationship power would predict greater aggressive responses toward romantic partners, but only when individuals were experiencing low situational power because they were unable to influence their partner in desired ways or were dependent on their partner for needed support. We also predicted this relationship power x situational power interaction to emerge for men, and not women, because low influence and high dependence threatens masculinity and responding aggressively is a way to demonstrate power and thereby restore masculinity.
We tested our predictions in five studies using multiple methods and operationalizations of relationship and situational power (see Table 1). First, we wanted to demonstrate that the predicted effects emerge using the two different ways relationship power has been assessed in prior research, including (1) assessing relative power by identifying the partner who is more invested and thus dependent on the relationship (Studies 1, 2 and 4; see Table 1), and (2) assessing perceptions of influence or decision-making power (Studies 3, 4 and 5; see Table 1). Second, we wanted to demonstrate that the predicted effects replicate across important power-relevant situations that involve (1) the need to influence the partner to achieve desired goals, such as during conflict, and (2) depending on the partner for the fulfilment of core needs, such as when people need support. Thus, our measures of situational power assessed the same defining elements of power as our measures of relationships power—dependence and influence—but captured the degree to which people (1) were able to influence their partner (Studies 1, 2, 3 and 5) or (2) were dependent on the partner (Studies 3 and 4) in specific power-relevant situations (see Table 1).
Table 1.
Operational Measurement of Power and Aggression across Studies
| Low Relationship Power | Low Situational Power | Aggression | |
|---|---|---|---|
| Study 1 |
High Dependence in Relationship Including the partner in one’s identity more than the partner includes the self in his/hers |
Low Influence in Situation Unsuccessful attempts to influence partner during conflict discussions |
Observer-rated aggressive communication during conflict discussions |
| Study 2 |
High Dependence in Relationship Experiencing more rewards from the relationship than the partner does |
Low Influence in Situation Partner disengagement and avoidance during conflict discussions |
Observer-rated aggressive communication during conflict discussions |
| Study 3 |
Low Influence in Relationship Perception of low influence and decision-making power |
Low Influence in Situationm Unable to influence partner High Dependence in Situation Needing support from partner |
Self-reported daily aggressive responses toward partner |
| Study 4 |
High Dependence in Relationship Including the partner in one’s identity more than the partner includes the self in his/hers Low Influence in Relationship Perception of low influence and decision-making power |
High Dependence in Situation Needing support from partner during discussions of important personal goals |
Observer-rated aggressive communication and self-reported aggressive feelings toward partner |
| Study 5 |
Low Influence in Relationship Perception of low influence and decision-making power |
Low Influence in Situation Partners’ unwillingness to change to resolve relationship problems |
Self-report daily descriptions of aggressive behavior toward partner |
To assess aggressive responses to low power, we assessed well-studied behaviors that (1) are relevant to the power-relevant situations investigated, (2) have been shown to be destructive and harmful to partners, and (3) represent psychological aggression involving acts or communications that are intended, or would be reasonably perceived as intended, to hurt the partner and cause psychological pain (Gelles & Straus, 1979; Straus, 1979; Vissing, Straus, Gelles, Harrop, 1991). In Studies 1, 2 and 4, independent observers rated the degree to which individuals exhibited aggressive communication toward their partner during couples’ video-recorded interactions, such as derogating, criticizing and insulting the partner. In Studies 3 and 5, we assessed more general aggressive responses that are relevant to the daily course of relationships, such as being critical, hurtful or yelling at and insulting the partner.
STUDY 1
In Study 1, we tested the link between relationship power and observer-rated aggressive communication when individuals were trying to influence their partner during couples’ video-recorded conflict discussions. We used the principle of least interest (Kelley & Thibaut, 1978; Waller & Hill, 1951) to assess relationship power. Regardless of absolute levels of investment, the person who is more dependent on their relationship—identifies more strongly and is more emotionally or cognitively involved than their partner—possesses less power to distribute rewards and punishments to their partner (Oriña et al., 2011; Sprecher et al., 2006). Thus, we indexed low relationship power as including the partner in one’s identity more than the partner includes the self in his/hers (see Table 1). To assess situational power, we measured the degree to which individuals were able to influence their partners’ attitudes and behavior in desired ways. We predicted men (but not women) low in relationship power would exhibit greater aggressive communication, but only when situational power was low because men were unable to influence their partner in desired ways.
Method
Participants
Study 1 involved new analyses of an existing sample and assessment of aggressive communication (Overall, Fletcher, Simpson & Sibley, 2009). Sixty-one heterosexual couples (total N = 122 individuals) replied to campus-wide advertisements and received $40NZD for participating. Participants were on average 23.48 years old (SD = 4.99) and involved in serious (61% married or cohabitating), long-term (M = 2.81 years, SD = 2.82) relationships.
Materials and Procedure
Relationship Power
Participants completed the Inclusion of the Other in the Self (IOS) Scale (Aron, Aron, & Smollan, 1992). The IOS involves selecting one of seven Venn diagrams depicting two circles that represent the self and the partner and vary in their degree of overlap (1 = no overlap, 7 = almost complete overlap; M = 5.59, SD = 1.18). Higher scores indicate greater incorporation of the partner into the self, and thus higher levels of cognitive dependence (also see Agnew, Van Lange, Rusbult, & Langston, 1998). Consistent with prior definitions of relationship power (Kelley & Thibaut, 1978; Simpson et al., 2013; Waller & Hill, 1951) and dyadic-based indices of relative dependence (Anderson, Keltner & John, 2003; Attridge et al., 1995; Loving et al., 2004; Oriña et al., 2011; Sprecher et al., 2006), we assessed power by measuring partners’ relative IOS. When individuals incorporate their partner into the self-concept more than their partner incorporates the individual into his/her self-concept, their outcomes are inherently more strongly influenced and bound by the partner’s actions and desires than vice versa. Thus, we operationalized low power as including the partner into one’s identity more than the partner includes oneself, and high power as the reverse. To do this, we compared dyad members’ IOS scores and scored each participant as low in relative power (−1) if they reported a higher IOS score than their partner, equal in power (0) if they reported an IOS score equal to their partner, or high in relative power (1) if they reported an IOS score lower than their partner. This approach captures the within-couple interaction between individuals’ and their partner’s level of IOS (Kenny & Cook, 1999), which is important because power within a relationship is relative to one’s partner and not whether individuals’ IOS scores are higher or lower compared to the rest of the sample (see Online Supplemental Material for more detail). Men were the low power partner in 38% of the couples, equal in 21%, and high in power in 41%.
Alternative Moderators
To rule out the possibility that our measure or the effects of relationship power were due to individual differences in relationship security, we also assessed self-esteem (Rosenberg, 1965) and attachment insecurity (Simpson, Rholes & Phillips, 1996). All scales were reliable (Cronbach’s alphas [α] > .80), and correlations across measures (see Table 2) reveal that the power index was not simply capturing insecurity.
Table 2.
Descriptive Statistics and Correlations between Relative Relationship Power and Questionnaire Measures (Studies 1, 2 and 4)
| Questionnaire Measures | Study 1 | Study 2 | Study 4 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||||
| 1. | 2. | 3. | 4. | 1. | 2. | 3. | 4. | 1. | 2. | 3. | 4. | ||
| 1. Relationship Power | - | .07 | −.19 | −.02 | - | .04 | −.17 | −.02 | - | −.03 | −.02 | −.08 | |
| 2. Self-Esteem | −.12 | - | −.36* | −.24 | −.02 | - | −.43* | −.41* | .16 | - | −.30* | −.32* | |
| 3. Attachment Anxiety | .22 | −.44* | - | .13 | .16 | −.52* | - | .47* | −.04 | −.39* | - | .09 | |
| 4. Attachment Avoidance | .27* | −.40* | .21 | - | .26* | −.16 | .46* | - | −.04 | −.24* | .16 | - | |
|
| |||||||||||||
| Men | Mean (SD) | 0.04 (0.89) | 5.43 (1.01) | 2.91 (0.92) | 2.85 (1.00) | 0.08 (0.91) | 3.35 (0.49) | 2.03 (0.96) | 2.05 (0.90) | −0.10 (0.83) | 5.10 (1.11) | 2.95 (0.99) | 2.78 (1.02) |
| Women | Mean (SD) | −0.04 (0.89) | 5.31 (1.01) | 3.05 (1.25) | 3.06 (0.91) | −0.08 (0.91) | 3.31 (0.53) | 2.00 (0.99) | 1.79 (0.85) | 0.10 (0.83) | 4.72 (1.05) | 3.19 (1.10) | 2.94 (1.02) |
Note. Correlations for men are above the diagonals, correlations for women are below the diagonals. The scale for relative relationship power ranged from −1 = low relative power, 0 = equal power, to 1 = high relative power. All other measures were assessed on 1–7 scales except self-esteem in Study 2, which was assessed on a 1–4 scale. Because relationship power was assessed relative to partners, correlations with relationship power control for its component parts of individuals’ and their partner’s IOS (Study 1 and 4) or SMD (Study 2) scores.
p < .05.
Situational Power
Couples were video-recorded having two 5-minute discussions about conflicts arising from: (1) male partners wanting to change something about female partners (male is agent of change and female is target of change), and (2) female partners wanting to change something about male partners (female is agent and male is target; order counterbalanced across the sample). Each discussion represents a power-relevant situation for both agents and targets of change because both positions involve trying to influence, and resist influence from, the partner (Overall, Sibley & Tan, 2011). However, assessing situational power requires examining the level of influence people experience within power-relevant situations, and in this study we only assessed the degree to which agents’ of change were successful in their influence attempts. Immediately after each discussion, agents of change rated: (a) how much their partner moved toward their own position (1 = have not moved at all, 7 = moved very much), (b) how successful they were in bringing about change or intention to change (1 = not at all successful, 7 = extremely successful), and (c) how successful the discussion was in bringing about change or intention to change (1 = not at all successful, 7 = extremely successful). Items were averaged to assess influence success (α was .90 for men and .92 for women; descriptive statistics shown in Table 3). We operationalized low situational power as agents of change perceiving their attempts to influence their partner as being unsuccessful and high situational power as agents perceiving their influence attempts to be relatively successful. The degree of influence success (situational power) was not associated with relationship power (B = −.10, t = −.37, p =.71; gender diff. t = .69, p = .49).
Table 3.
Descriptive Statistics of Measures of Situational Power and Aggression (Studies 1 to 4)
| Situational Power and Aggression Measures | Men | Women | Gender Differences |
|---|---|---|---|
|
| |||
| Mean (SD) | Mean (SD) | t | |
| Study 1 Self-Report and Observer Ratings | |||
| Influence Success | 4.41 (1.59) | 4.22 (1.39) | −0.69 |
| Derogation | 1.41 (0.72) | 1.75 (1.09) | 4.25** |
| Invalidation | 1.69 (0.98) | 2.34 (1.24) | 2.45** |
|
| |||
| Study 2 Observer Ratings | |||
| Avoid | 2.21 (1.23) | 1.78 (1.05) | −4.02** |
| Engaged | 4.51 (1.05) | 5.01 (0.95) | 5.35** |
| Rejection | 0.01 (0.02) | 0.01 (0.03) | 1.69 |
|
| |||
| Study 3 Daily Diary Reports | |||
| Inability to Influence Partner | 2.69 (0.91) | 2.70 (1.00) | 0.08 |
| Needing Support | 3.59 (1.30) | 4.28 (1.29) | 4.50** |
| Daily Aggressive Responses | 2.32 (0.71) | 2.46 (0.58) | 1.61 |
|
| |||
| Study 4 Self-Report and Observer Ratings | |||
| Needing Support | 0.83 (1.08) | 1.15 (1.15) | 1.96* |
| Aggressive Responses | 1.58 (0.84) | 1.75 (1.17) | 1.14 |
|
| |||
| Study 5 Daily Diary Reports | |||
| Partner Willing to Change | 6.03 (1.08) | 5.88 (1.20) | −3.47** |
| Feelings of Masculinity | 4.09 (1.92) | 1.33 (0.92) | −49.04** |
Note. All measures were assessed on 1–7 scales, except for rejection in Study 2 which represents a proportion of speaking turns during problem-solving discussions. Gender difference ts are tests of whether average levels of situational power and aggression differ across men and women.
p = .05.
p < .01.
Aggressive Communication
Two trained coders independently rated two forms of aggressive communication using an established coding scheme that incorporates the most commonly-assessed hostile and destructive conflict behaviors (see Overall et al., 2009) and have been shown to have detrimental effects on the partner’s wellbeing (Gottman, 1998; Karney & Bradbury, 1995; Overall & Simpson, 2013). Coders rated the degree which agents’ of change exhibited: (1) derogation (criticizing, derogating and threatening the partner, expressing anger and hostility) and (2) autocracy (being domineering, rejecting, invalidating the partner; 1= low, 7 = high). Men and women were coded independently in separate viewings, with order of coding counterbalanced across the sample. Inter-rater reliability was high (intraclass correlation coefficients [ICCs] > .97) and descriptive statistics are shown in Table 3. Derogation and autocracy were strongly correlated for men and women (r = .64 and .51, p < .01) so we standardized (z-scored) and averaged derogation and autocracy to assess aggressive communication during the discussion (also see Overall et al., 2009).
Results
To test our predictions, we regressed agents’ aggression on their relative power score (relationship power), mean-centered ratings of influence success (situational power), and the interaction between relative power and influence success (relationship power x situational power). The relative power score captures the within-couple interaction between the partners’ IOS scores and so to identify the unique effect of relationship power the components used to generate the relative power score need to be controlled (Kenny & Cook, 1999). Thus, we included the main and interactive effects of individuals’ own and their partners’ IOS score to ensure any effects of relative power were not simply due to individuals’ or their partners’ having high or low IOS scores. Following the procedures outlined by Kenny, Kashy and Cook (2006), we used the MIXED procedure in SPSS 21 to run a dyadic regression model estimating the effects pooled across men and women, and tested whether the effects significantly differed across men and women by modeling the main and interaction effects of gender (−1 women, 1 men). We then simultaneously estimated the effects for men and women using a two-intercept model that accounts for the dyadic dependence in the data (equivalent to decomposing the gender interactions; see Kenny et al., 2006).
The top section of Table 4 presents the main and interaction effects of relationship power and situational power for men (first column) and women (second column) as well as the gender interaction effects testing whether these effects differed across men and women (final column). As predicted, the interaction between relationship power and situational power was significant for men (p = .01) and not women (p = .60). The 3-way relationship power x situational power x gender interaction effect (see bold predicted effect in final column of Table 4) also revealed this difference was marginally significant (p < .08). Shown in Figure 1, men low in relationship power exhibited greater aggressive communication when they were unable to influence their partner and thus had low situational power (b = −1.48, t = −3.98, p < .01, r = −.48), but men low in relationship power did not display more aggression when they were able to influence their partner (b = −0.25, t = −0.84, p = .41, r = −.11).
Table 4.
The effects of Relationship Power and Situational Power on Observer-Rated Aggressive Communication during Couples’ Conflict Discussion (Studies 1 and 2)
| Power Variables | Main and Interaction Effects for Men | Main and Interaction Effects for Women | Interaction Effects testing Gender Differences | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| B | t | r | B | t | r | B | t | r | |
| Study 1 | |||||||||
| Relationship Power | −.87 | −3.58** | −.44 | .38 | 1.25 | .17 | −.62 | −3.10** | .38 |
| Situational Power | −.11 | −1.73 | −.23 | −.13 | −1.52 | −.20 | .02 | 0.28 | .03 |
| Relationship Power × Situational Power | .41 | 2.64* | .34 | −.14 | −0.53 | −.07 | .28 | 1.79† | .19 |
| Study 2 | |||||||||
| Relationship Power | −.01 | −2.32* | −.21 | −.01 | −1.20 | −.11 | −.00 | −0.16 | .01 |
| Situational Power | −.01 | −3.55** | −.30 | −.00 | −1.15 | −.10 | −.01 | −1.15 | .09 |
| Relationship Power × Situational Power | .01 | 3.08** | .27 | −.01 | −1.18 | −.11 | .01 | 2.58** | .18 |
Note. Effects that test our specific predictions are displayed in bold. The final column presents the interaction effects testing whether the main and interaction effects of relationship power and situational power significantly differ across men and women. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √(t2 / t2 + df).
p < .08.
p < .05.
p < .01.
Figure 1. The interaction between relationship power and situational power on men’s aggressive communication exhibited within conflict discussions in which men were trying to change their partner (Study 1).
Note: Low (versus high) relationship power was operationalized as incorporating the partner into one’s identity more (versus less) than the partner incorporated the individual into their identity. Low situational power was operationalized as being relatively unsuccessful (−1 SD) in influencing the partner during the discussion and high situational power as being relatively successful in influencing the partner (+1 SD).
Alternative moderators and explanations
Self-esteem, attachment anxiety and avoidance did not demonstrate the same effects as relationship power did and the significant effects shown in Figure 1 remained robust when controlling for any of these variables. Age, relationship status (married/cohabiting or not) and relationship length were also not significantly associated with relationship or situational power, did not demonstrate the same effects as power did, and controlling for these demographic variables did not alter the effects.
Discussion
Study 1 used an existing sample to provide an initial test of our predictions. Men who possessed low relationship power because they were more dependent than their partner exhibited greater aggressive communication during couples’ conflict discussions when they were unable to influence their partner in desired ways and thus had low situational power. Low relationship power was not associated with aggression when men were able to influence their partner, and women’s relationship power was not associated with aggression.
STUDY 2
In Study 2, we aimed to replicate the effects shown in Study 1 employing different measures of power. To assess relationship power, we again followed the principle of least interest (Waller & Hill, 1951) but in Study 2 identified low power partners as those who received more rewards from their relationship and were thus more invested in and dependent on their relationship. We again examined aggressive communication during couples’ discussions but in Study 2 these discussions focused on general relationship problems rather than conflicts arising from specific desires to change the partner. To assess situational power, we measured the degree to which partners avoided and disengaged during the discussion. Avoidance is a particularly effective way of reducing the power of the other partner because it communicates that the avoider is (a) unwilling to be influenced and (b) willing to reduce closeness and intimacy to maintain their position. Indeed, when the partner refuses to engage and avoids discussing the issue, this directly thwarts the individual’s attempts to influence their partner and resolve conflict (Christensen & Heavey, 1990; Heavey, Christensen, & Malamuth, 1995) and this type of partner resistance has been previously associated with greater aggressive responses by men (Babcock et al., 1993). We predicted that men, but not women, who possessed low relationship power would exhibit more aggressive communication when their partner’s high level of avoidance reduced their situational power.
Method
Participants
Participants were 132 newlywed couples (total N = 264 individuals) participating in a broader study of marriage (see McNulty & Russell, 2010). Husbands were an average age of 25.9 years (SD = 4.6), 70% were employed full time and 26% were full time students. Wives were an average age of 24.2 years (SD = 3.6), 56% were employed full time, and 28% were full time students. Ninety-one percent of husbands and 93% of wives identified as Caucasian.
Materials and Procedure
Relationship Power
This study did not include measures of IOS as used in Study 1 to assess relative power. However, another important element underpinning levels of relationship dependence (and thus power) is how much each partner finds the relationship rewarding and satisfying (Rusbult, 1983; Rusbult, Arriaga & Agnew, 2001). When individuals find the relationship more rewarding than their partners, their partners will be relatively less invested in the continued future of the relationship, put relatively less effort into maintaining the relationship, and be relatively more likely to withdraw from the relationship (Murray & Holmes, 2009; Rusbult et al., 2001). This study included a good measure of how rewarding the relationship was experienced; a semantic differential scale (SMD; Osgood, Suci, & Tannenbaum, 1957) that involved each spouse rating their perceptions of their relationship on 7-point scales involving fifteen pairs of opposing adjectives (e.g., “Bad-Good,” “Unpleasant-Pleasant”). Scores ranged from 15 to 105, with higher scores reflecting more positive relationship evaluations (for husbands, M = 97.40, SD = 8.08, α = .89; for wives, M = 97.57, SD = 8.40, α = .91). Following the same procedure in Study 1, we assessed relationship power by measuring partner’s relative SMD. Each participant was identified as low in relationship power (−1) if they found the relationship more rewarding then their partner, equal in power (0) if they reported the relationship as equally rewarding, or high in power (1) if they experienced the relationship as less rewarding than their partner. Similar to the distributions in Study 1, men were the low power partner in 38% of the couples, equal in 17%, and the high power partner in 45%.
Alternative Moderators
Participants also completed scales (αs > .80) assessing self-esteem (Rosenberg, 1965) and attachment insecurity (Brennan, Clark, & Shaver, 1998). The patterns of correlations were similar to Study 1 (see Table 2).
Situational Power
Couples were video-recorded having two 10-minute discussions about current relationship problems; one identified by the husband and one identified by the wife. Unlike Study 1, husbands and wives were instructed to choose a problem with the relationship rather than something specifically they wanted to change about their spouse. Nonetheless, discussions of relationship problems involves both partners trying to influence and resist influence, and thus is a power-relevant situation regardless of who selected the topic (Christensen & Heavey, 1990; Gottman, 1998). In addition, unlike Study 1, we did not assess perceptions of influence of one of the partners. Instead, in Study 2 we assessed both partners’ situational power by assessing the degree to which each individual’s influence was forestalled by their partner’s avoidance and disengagement. As described above, avoidance and disengagement has been previously established as a powerful way of undercutting others’ power and influence (Christensen & Heavey, 1990; Gottman, 1998; Heavey et al., 1995) and our assessment of these partner behaviors followed this prior literature. After viewing the entire discussion, four trained coders independently rated the degree to which each spouse (a) avoided and (b) was engaged during the discussion (1= Not at all, 7 = Extremely). To assess reliability, 24% of the conversations were double-coded (average ICC across coders = 0.66). Ratings of the degree which each spouse avoided and was engaged were negatively correlated (r = −.57) and so we reverse-coded ratings of avoid and then standardized (z-scored) and averaged both ratings so that lower scores represented lower situational power (greater partner avoidance and disengagement). As in Study 1, relative relationship power did not significant predict levels of situational power for men (B = −.11, t = −.90, p =.37) or women (B = −.15, t = −1.06, p =.29; gender difference t = .19, p = .85).
Aggressive Communication
Four trained coders coded each speaking turn according to whether it met the definition for rejection as described in the Verbal Tactics Coding Scheme (VTCS, Sillars, Coletti, Parry, & Rogers, 1982). This code was focused on because the VTCS assigns precedence to rejection over all other behaviors in recognition of the importance of these explicitly hostile behaviors. Similar to the elements assessed in Study 1, speaking turns that directly insulted or pointed out personal flaws in the partner (“You’re so immature”) and directly undermined the partner’s point of view (“I don’t care what you think”, “whatever”) were coded as rejection. The number of rejection codes assigned to each spouse was divided by the total number of speaking turns by that spouse to provide an estimate of the overall proportion of speaking turns that were classified as rejection. To assess reliability, 24% of the conversations were double-coded (ICC across coders = 0.65).
Results
We used the same dyadic analytic approach as in Study 1 to regress aggressive communication exhibited during the problem-solving discussions on individuals’ relative power score (relationship power), observations of partners’ behavioral avoidance (situational power), and the interaction between relative power and partners’ avoidance (relationship power x situational power). As in Study 1, we also modeled the main and interaction effects of individuals’ own and their partner’s SMD score (both mean-centered) to ensure that the effects of relationship power were not due to either partner having high or low SMD scores.
Table 4 (bottom section) presents the effects for men and women (first two columns) as well as the interaction effects testing whether the effects significantly differed across men and women (final column). As predicted, the interaction between relationship power and situational power was significant for men, but not women, and the 3-way interaction testing the gender difference was also significant (see effect in bold in final column). Shown in Figure 2, men low in relationship power demonstrated greater aggression when their partners avoidance was high and they faced low situational power (b = −.01, t = −3.38, p < .01, r = −.30), but not when their partners’ avoidance was low (b = .00, t = 0.95, p = .35, r = .09).
Figure 2. The interaction between relationship power and situational power on men’s aggressive communication exhibited within relationship problem discussions (Study 2).
Note: Low (versus high) relationship power was operationalized as experiencing more (versus less) rewards from the relationship than the partner. Low situational power was operationalized as the partner resisting influence by disengaging and avoiding whereas high situational power involved the partner being open to influence by engaging in problem solving.
Alternative Moderators and Explanations
As in Study 1, additional analyses supported that the results were not due to general relationship insecurity: men’s self-esteem, attachment anxiety and avoidance did not demonstrate the same effects as relationship power and the significant effects shown in Table 4 and Figure 2 remained when controlling for any of these variables. Relationship status did not vary across couples, who were all newlyweds within six months of their wedding. Age and cohabitation status before marriage also did not alter or modify the effects shown in Table 4 and Figure 2.
Discussion
Study 2 replicated the results of Study 1 using similar observational methods but different operationalizations of power (see Table 1). Men (but not women) who possessed lower relationship power exhibited greater aggressive communication during couples’ conflict discussions when their partners’ avoidance reduced their situational power.
STUDY 3
Study 3 extended Studies 1 and 2 in three major ways. First, to show that the effects of relative relationship power assessed in Studies 1 and 2 using the principle of least interest generalize to other ways in which relationship power has been assessed, we measured each partners’ perception of influence in their relationship using an established scale that has been applied in a variety of power-relevant contexts (Anderson, John, & Keltner, 2012). Second, instead of measuring aggression in a single, laboratory-based context, we assessed people’s daily levels of situational power and aggressive behavior for a 3-week period. These repeated daily assessments allowed us to examine aggressive responses to low situational power across couples’ more routine daily interactions, and to test within-person changes in aggressive responses as participants experienced varying levels of situational power on a day-to-day basis. The resulting within-person analysis is the strongest test of our contextual prediction by showing that a man who possesses low relationship power only responds aggressively when he confronts low situational power, but does not respond aggressively at other times.
Finally, in Study 3 we assessed low situational power in two ways that capture the two different ways relationship power has been previously conceptualized (see Table 1). Paralleling Studies 1 and 2, we first assessed low situational power arising from participants being unable to influence their partner in desired ways. Our second assessment of situational power captured the dependence central to foundational definitions and measures of relationship power. In particular, we assessed the extent to which participants needed support from their partner each day. When individuals need support, they are dependent on their partner for the fulfilment of key relatedness needs (i.e., help, comfort). This dependence places individuals in a one-down position (Kelley et al., 2003; Kelley & Thibaut, 1978) and also threatens masculine ideals of self-reliance, toughness, strength and independence (Levant et al., 1992; Mahalik et al., 2003; Thompson & Pleck, 1986; Vescio, et al., 2010). Indeed, being masculine involves rejecting feminine qualities, feelings and roles that involve dependence and so being in such feminine roles or positions threatens masculinity (Bosson et al., 2009; Dahl et al., 2015; Vescio, et al., 2010). Accordingly, men are more likely to find support from their partners threatening, especially if they feel less able to effectively cope with the issue and are therefore more dependent on their partners (e.g., Crockett & Neff, 2012). Thus, we predicted that low power men would respond aggressively in any context in which situational power (and thus masculinity) was undermined, including when unable to influence their partners and when dependent on their partners for support.
Method
Participants
Seventy-eight heterosexual couples (total N = 156) replied to advertisements distributed across a large university campus and associated organizations (e.g., health and recreation centers). Couples were involved in serious (44% married/cohabitating), long-term (M = 2.57 years, SD = 1.96) relationships. Participants were on average 22.44 years old (SD = 4.81), and 47% held a university degree. Participants were paid $45NZD.
Procedure
Participants completed baseline measures during an initial laboratory session. Over the following 21 days, both partners independently completed an end-of-day web-based record regarding their relationship experiences and behavior that day. On average, participants completed 19.3 diary entries (92%), for a total of 3,014 entries across the entire sample. Descriptive statistics for the diary measures are shown in Table 3.
Materials
Relationship Power
Participants completed the Sense of Power Scale with reference to their relationship (Anderson et al., 2012). This 8-item scale (α = .84) assesses individuals’ ability to make decisions (e.g., “if I want to, I get to make the decisions”, 1 = strongly disagree, 7 = strongly agree), influence the partner’s behavior or opinions (e.g., “even if I voice them, my views have little sway”, reverse-coded), and satisfy one’s own goals and desires (e.g., “even when I try, I am not able to get my way”, reverse-coded). Levels of perceived power did not differ across men and women (t = 1.51, p = .13; see Table 5).
Table 5.
Descriptive Statistics and Correlations between Perceptions of Relationship Power and Questionnaire Measures (Studies 3, 4 and 5)
| Questionnaire Measures | Study 3 | Study 4 | Study 5 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||||
| 1. | 2. | 3. | 4. | 1. | 2. | 3. | 4. | 1. | 2. | 3. | 4. | ||
| 1. Relationship Power | - | .20 | −.42* | −.17 | - | .17 | −.15 | −.27* | - | .26* | −.14 | −.22* | |
| 2. Self-Esteem | .04 | - | −.21 | −.15 | .25* | - | −.30* | −.32* | .20* | - | −.35* | −.29* | |
| 3. Attachment Anxiety | −.19 | −.38* | - | .13 | −.21* | −.39* | - | .09 | −.30* | −.26* | - | .50* | |
| 4. Attachment Avoidance | −.13 | −.12 | .10 | - | −.15 | −.24* | .16 | - | −.30* | −.12 | .46* | - | |
|
| |||||||||||||
| Men | Mean (SD) | 5.25 (0.75) | 5.39 (0.95) | 2.75 (1.00) | 2.80 (0.97) | 5.09 (0.91) | 5.10 (1.11) | 2.95 (0.99) | 2.78 (1.02) | 5.28 (0.91) | 3.50 (0.49) | 2.18 (0.97) | 2.25 (0.88) |
| Women | Mean (SD) | 5.48 (1.11) | 5.10 (1.17) | 3.23 (1.05) | 3.03 (1.11) | 5.36 (0.88) | 4.72 (1.05) | 3.19 (1.10) | 2.94 (1.02) | 5.42 (0.92) | 3.47 (0.50) | 2.20 (1.02) | 2.15 (0.94) |
Note. Correlations for men are above the diagonal, correlations for women are below the diagonal. All other measures were assessed on 1–7 scales except self-esteem in Study 2, which was assessed on a 1–4 scale. Relationship power was measured in each study by assessing participants’ perceptions of relationship power using the Sense of Power Scale (Anderson et al., 2012). See Table 2 for associations with relative relationship power.
p < .05.
Alternative Moderators
Participants completed the same scales used in Study 1 to assess self-esteem (α = .89), attachment anxiety (α = .80) and attachment avoidance (α = .75). Lower perceived power tended to be linked with greater attachment insecurity (see Table 5).
Daily Situational Power
To assess situational power, we measured two contexts: At the end of each day participants rated the degree to which he/she (1) was unable to influence his/her partner (“I could NOT get my partner to think, feel or behave the way I wanted him/her to”) and the degree to which he/she needed support (“I needed support from my partner”) that day (1 = not at all, 7 = extremely). These items were reverse-coded so that lower scores represent lower situational power (lower ability to influence the partner and needing more support) and higher scores represent higher situational power (greater ability to influence the partner and needing less support). Lower relationship power predicted lower reported ability to influence the partner for both men and women (B = .35, t = 4.22, p < .001; gender difference t = 1.08, p = .29), but was not associated with daily levels of support need (B = .03, t = .12, p =.80; gender difference t = .47, p = .64).
Daily Aggression
Participants rated five items used in prior research to assess destructive and hostile responses that are relevant to the day-to-day course of relationships (Hammond & Overall, 2013; Overall & Sibley, 2009, 2010). Based on the widely used and validated construct of accommodation (Rusbult, Verette, Whitney, Slovik, & Lipkus, 1991), this scale was developed to assess the degree to which individuals expressed destructive impulses to derogate or hurt the partner (“I was critical or unpleasant toward my partner,” “I acted in a way that could be hurtful to my partner”; 1 = not at all, 7 = extremely) rather than transform their destructive motivations to care for the partner and relationship (“I was willing to let my partner have things his/her way”, “I was forgiving toward my partner”, “I was affectionate and loving toward my partner”). The latter items were reverse-scored, and all items averaged (α = .76). Separate analyses of the items assessing expression and inhibition of aggressive responses produced the same pattern (see online supplemental materials [OSM]). Although these daily responses do not capture the overt hostility that we assessed during couples’ conflict discussions in Studies 1 and 2, these more routine responses are associated with those types of aggressive communication (Rusbult et al., 1991), and they predict more negative relationship outcomes for both partners (e.g., Wieselquist et al., 1999).
Results
We present two models focusing on our two assessments of situational power: ability to influence the partner (Model 1) and needing support from the partner (Model 2).
Model 1: Ability to Influence the Partner
Our analyses followed Kenny et al.’s (2006) and Bolger and Laurenceau’s (2013) recommendations for analyzing repeated measures dyadic data. Daily aggressive responses were modeled as a function of: (a) relationship power (mean-centered), (b) situational power (ability to influence the partner that day; person-centered), and (c) the interaction between situational power and relationship power. We also modeled average levels of situational power to isolate change in aggressive responses as daily situational power varied from typical levels of daily power (Bolger & Laurenceau, 2013; Bryk & Raudenbush, 2002). As in Studies 1 and 2, and following procedures by Kenny et al. (2006), we simultaneously estimated the effects for men and women (first two columns in Table 6) using a two-intercept model that controlled for the dyadic dependencies in the data, and we tested whether the differences across men and women were significant (see final column in Table 6) by pooling the effects across partners and modeling the main and interaction effects of gender (−1 women, 1 men). All analyses allowed the error variances to differ for men and women, errors were allowed to correlate within days and across dyad members, and the intercept was modeled as random.
Table 6.
The effects of Relationship Power and Situational Power on Daily Aggressive Responses toward Partner (Study 3)
| Power Variables | Main and Interaction Effects for Men | Main and Interaction Effects for Women | Interaction Effects testing Gender Differences | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| B | t | r | B | t | r | B | t | r | |
| Model 1: Ability to Influence the Partner | |||||||||
| Relationship Power | −.18 | −1.66 | −.19 | .06 | 1.08 | .12 | −.12 | −2.03* | .20 |
| Situational Power | −.18 | −13.23** | −.33 | −.25 | −16.09** | −.39 | .03 | 3.13* | .06 |
| Relationship Power × Situational Power | .06 | 2.66** | .29 | −.04 | −2.78** | −.31 | .05 | 3.78** | .33 |
| Model 2: Need of Partner Support | |||||||||
| Relationship Power | −.22 | −2.19* | −.24 | .01 | 0.22 | .03 | −.12 | −2.07* | .20 |
| Situational Power | −.07 | −5.05* | −.13 | −.07 | −4.88** | −.13 | .01 | 0.26 | .01 |
| Relationship Power × Situational Power | .05 | 2.58* | .28 | .01 | 0.10 | .01 | .02 | 2.10* | .19 |
Note. Effects that test our specific predictions are displayed in bold. The final column presents the interaction effects testing whether the main and interaction effects of relationship power and situational power significantly differ across men and women. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √(t2 / t2 + df). Between-person degrees of freedom were used for calculating the effect sizes for the interactions between relationship power and situational power (and associated gender interactions) shown in bold.
p < .05.
p < .01.
As shown in Table 6 (top section), the relationship power x situational power x gender interaction revealed that the interaction between relationship power and situational power significantly differed across men and women. The significant interaction between men’s relationship power and within-person daily fluctuations in situational power is shown in Figure 3. Low relationship power was associated with greater aggression on days when men experienced low situational power because they were unable to influence their partner in desired ways (b = −.26, t = −2.26, p = .03, r = −.25) but not on days when those same men were able to influence their partner (b = −.11, t = −0.95, p = .34, r = −.11). Unexpectedly, a significant interaction also emerged for women (see Table 6; top section) that was in the opposite direction than that of men (see OSM for more detail): lower relationship power predicted lower aggression on days of low situational power (b = .11, t = 1.83, p = .07, r = .21), but not on days of high situational power (b = .02, t = .23, p = .82, r = .03).
Figure 3. The interaction between relationship power and daily situational power (ability to influence the partner) on men’s daily aggressive communication (Study 3).
Note: Participants reported on their sense of power in their relationship (relationship power) and provided daily ratings of their ability to influence their partner in desired ways (situational power) across a 3-week period. Low and high values are indexed at 1 SD below and above the mean.
Model 2: Needing Support from the Partner
The results from analogous analyses testing whether men lower in relationship power responded to the dependence associated with needing support with greater daily aggression are shown in Table 6 (bottom section). A significant interaction between relationship power and within-person fluctuations in situational power emerged for men, but not women, and the gender difference was again significant (see final column). Shown in Figure 4, lower relationship power was associated with greater aggressive responses on days when men needed high levels of support (b = −.30, t = −2.79, p < .01, r = −.31) but not on days when men did not need support from their partners (b = −.15, t = −1.42, p = .16, r = −.16).
Figure 4. The interaction between relationship power and daily situational power (need of partner support) on men’s daily aggressive communication (Study 3).
Note: Participants reported on their sense of power in their relationship (relationship power) and provided daily ratings of the degree to which they needed support from their partner (low situational power) across a 3-week period. Low and high values are indexed at 1 SD below and above the mean.
Alternative moderators and explanations
As in Studies 1 and 2, rerunning Models 1 and 2 controlling for the main and interaction effects of self-esteem, attachment anxiety, attachment avoidance, age, relationship status (cohabiting/married versus not) or relationship length did not change the effects shown in Table 6 and Figures 3 and 4.
Discussion
Study 3 replicated and extended the results of Studies 1 and 2 using different methods and measures, including assessing individuals’ perceptions of relationship power, modeling within-person changes in daily aggressive responses to low situational power encountered across couples’ daily lives, and assessing an additional low situational power context—the degree to which individuals needed support and were thus dependent on their partner. As predicted, men who had lower relationship power reported greater daily aggression, but only on days when they experienced low situational power because they were unable to influence their partner or they needed support from their partner. As in Studies 1 and 2, women did not show these effects, and one effect suggested women low versus high in relationship power responded less aggressively when they were unable to influence their partner.
Study 4
An important contribution of Study 3 was showing that men low in relationship power exhibited greater daily aggression when they needed support and thus their heightened relational needs increased dependence on the partner. Such dependence for the fulfillment of relational needs is a central source of low power in relationship interactions (Kelley et al., 2003). Moreover, demonstrating the low power-aggression link in this situational context is important because it reveals that the aggressive responses enacted by low power men are motivated by masculinity related concerns rather than by instrumental efforts to coerce desired behaviors from partners. Aggressive responses are not effective ways to secure needed support from partners (e.g., Overall, Fletcher & Simpson, 2010; Pasch & Bradbury, 1998), but aggression does help to reduce dependence and reestablish power and control (Leary, Twenge & Quinlivan, 2006; Murray & Holmes, 2009; Overall & Sibley, 2009), and thus helps to restore masculinity (Bosson & Vandello, 2011; Vandello et al., 2008).
Our aim in Study 4 was to replicate the effect of low situational power arising from support need by examining aggressive responses emitted within couples’ support-relevant discussions. We also wanted to replicate the predicted relationship power x situational power interaction for men using the two different methods prior research has used to assess relationship power—relative power (as in Studies 1 and 2) and perceptions of power (as in Study 3)—within the same study. We assessed relative power using identical measures and procedures as in Study 1, and we also gathered participants’ perceptions of relationship power using the sense of power scale as in Study 3. We expected that both measures would produce the same effect: men lower in relative or perceived relationship power would exhibit greater aggression when they faced low situational power because they needed their partner.
To examine low power situations characterized by dependence and the need for support, participants identified their most important personal goal they were currently trying to achieve, and then discussed their goal-related strivings and progress with their partner. Following previously used measures to index greater support need (Collins & Feeney, 2000, 2004; Feeney, 2004; Crockett & Neff, 2012; Girme, Overall & Simpson, 2013; Girme, Overall, Simpson & Fletcher, 2015), we assessed participants’ distress during couples’ discussions and feelings of hopelessness regarding their personal goal. Greater distress and hopelessness signal a greater need for help in order to cope with and overcome the challenges associated with goal pursuit, and thus are associated with greater dependence on the partner’s support to alleviate distress and generate goal-related efficacy (e.g., Bar-Kalifa & Rafaeli, 2014; Girme et al., 2013; Cutrona et al., 2007). However, greater distress and hopelessness also threaten masculine ideals of self-reliance, strength and independence precisely because they indicate a need for support (Bem, 1974; Bem, 1981; Levant et al., 1992; Mahalik et al., 2003) and are feminine or ‘non-manly’ feelings and qualities (Bosson et al., 2009; Dahl et al., 2015; Vescio, et al., 2010). Thus, we expected men lower in relationship power would exhibit greater aggressive responses within couples’ discussions when they needed higher levels of support and this heightened dependence resulted in low situational power.
Method
Participants
One-hundred heterosexual couples (total N = 200 individuals) responded to paper and electronic advertisements distributed across a large university and associated organizations (e.g., health and recreation centers). Participants were involved in serious (13% married, 36% cohabiting, 47% serious dating relationships), long-term (M = 3.28 years, SD = 4.16) relationships, and were a mean age of 22.64 (SD = 6.51) years. Participants were paid NZ$80.
Procedure
After completing the questionnaires described below, participants identified and ranked in order of importance three current personal goals they were trying to achieve, which they were told they might discuss with their romantic partners. The top-ranked goal was selected for discussion. After a 5-minute warm-up discussion, each couple was video-recorded engaging in two 7-minute discussions about each person’s personal goal (order across male and female partners counterbalanced across the sample). Our analysis focused on the aggressive responses enacted by the person whose goal was discussed and therefore was in a position to need support from their partner. To assess support need, immediately after the discussion participants reported on their distress during the discussion and the degree to which they felt hopeless regarding their goal. To assess aggressive responses, objective coders rated the degree to which each individual displayed aggressive communication and participants reported on their anger toward their partner.
Materials
Relationship Power
We assessed relationship power in two ways that represent the different methods prior research has adopted to assess relationship power. First, as in Study 1, participants completed the IOS Scale (M = 5.25, SD = 1.24) and we scored each participant as low in relative power (−1) if they reported a higher IOS score than their partner, equal in power (0) if they reported an IOS score equal to their partner, or high in relative power (1) if they reported an IOS score lower than their partner. Men were scored as the low power partner in 40% of the couples, equal in 30%, and high in power in 30%. Second, as in Study 3, participants completed the Sense of Power Scale with reference to their relationship (α = .80; see Table 5). As in prior research (Caldwell & Peplau, 1984; Farrell et al., 2015; Felmlee, 1994), relative power and perceived power were only weakly associated (B = .16, t = 1.25, p =.22, r = .13; gender t = .22, p =.83), but we expected both low power based on relative dependence and low power based on perceived influence to have similar effects.
Alternative Moderators
Participants completed the scales used in Studies 1–3 to assess self-esteem (α = .87), attachment anxiety (α = .78), and attachment avoidance (α = .76). The pattern of associations with relative relationship power (see Table 2) and perceptions of relationship power (see Table 5) were similar to those found in Studies 1–3.
Situational Power
Using prior assessments of support need during couples’ discussions in prior research (e.g., Girme et al., 2013, 2015), immediately after the discussion, individuals whose goal was discussed reported how “stressed” and “upset” they felt during the discussion (1 = not at all, 7 = extremely; r = .74). An as additional indicator of need specific to the targeted goal, we asked participants to report the degree to which they felt ‘hopeless’, ‘discouraged’, ‘sad’ and ‘hopeful’ (reverse-scored) with regard to their goal (α = .82). These two measures of support need were highly correlated (r = .56), and so we combined them to construct an overall index of support need. As in Study 3, we then reverse-coded this overall score so that lower values represent lower situational power (needing more support). Lower perceptions of relationship power (B = .29, t = 3.25, p > .01; gender diff. t = −1.60, p = .11), but not relative relationship power (B = −.05, t = −.30, p =.77; gender diff. t = .05, p = .96), were associated with lower situational power (i.e., needing more support).
Aggressive Responses
Four independent coders rated the degree to which individuals exhibited aggressive communication when discussing their goal. The coding instructions and responses targeted were derived from established coding schedules of couples’ support behavior (Overall et al., 2010; Pasch & Bradbury, 1998; Sullivan et al., 2010) and captured the same aggressive behaviors assessed in Study 1, including the degree to which individuals (a) criticized and derogated the partner, (b) rejected or invalidated their partner’s input, and (c) expressed anger and hostility at the partner. Prior research has shown these behaviors predict lower partner support and reductions in both partners’ relationship wellbeing (Overall et al., 2010; Pasch & Bradbury, 1998; Sullivan et al., 2010). Men and women were coded in separate viewings (order counterbalanced). Inter-rater reliability was excellent (ICC = .94).
To supplement the observed ratings and ensure these aggressive responses were directed toward the partner, we followed prior assessments of aggression and aggressive motives (Finkel et al., 2012; Lemay, Overall & Clark, 2012) by also asking participants to report on the degree to which they felt antagonistic and aggressive feelings toward their partner during the discussion, including being “angry” and “annoyed” (1 = not at all, 7 = extremely; r = .77). These partner-oriented aggressive feelings are strongly linked with antagonistic motivations and hostile behavior (Fitness & Fletcher, 1993; Lemay et al., 2012; also see Fischer & Roseman, 2007). Accordingly, observer-coded aggressive communication was positively correlated with participants’ anger (r = .40), and we averaged these measures to index aggressive responses toward the partner during the discussion. Analyzing the two measures separately produced the same pattern of results (see OSM).
Results
To test our predictions, we present two models focusing on our two assessments of relationship power: relative relationship power as assessed in Studies 1 and 2 (Model 1) and perceptions of relationship power as assessed in Study 3 (Model 2).
Model 1: Relative Relationship Power
We adopted an identical dyadic model to the analyses of relative relationship power described in Study 1. As shown in the top of Table 7, the interaction between relationship and situational power was significant for men (first column), but not women (second column), and this gender difference was significant (final column). The significant interaction for men replicated the effects of daily support need in Study 3. Shown in Figure 5, lower relationship power was associated with greater aggressive responses toward the partner when men needed greater support and therefore had low situational power (b = −0.99, t = −3.53, p < .01, r = −.34) but not when men’s support need and dependence was low (b = .23, t = 1.04, p = .30, r = .10).
Table 7.
The effects of Relationship Power and Situational Power (Need of Partner Support) on Aggressive Responses during Couples’ Support Discussions (Study 4)
| Power Variables | Main and Interaction Effects for Men | Main and Interaction Effects for Women | Interaction Effects testing Gender Differences | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| B | t | r | B | t | r | B | t | r | |
| Model 1: Relative Relationship Power (as in Study 1) | |||||||||
| Relationship Power | −.38 | −2.09* | −.21 | .23 | 1.08 | .11 | −.30 | −1.81† | .18 |
| Situational Power | −.32 | −4.81** | −.45 | −.47 | −6.24** | −.54 | .08 | 1.66 | .14 |
| Relationship Power × Situational Power | .52 | 3.49** | .35 | −.14 | −.73 | −.08 | .33 | 2.63** | .21 |
| Model 2: Perceived Relationship Power (as in Study 3) | |||||||||
| Relationship Power | −.14 | −1.88† | −.19 | .05 | .47 | .05 | −.10 | −1.52 | .12 |
| Situational Power | −.22 | −3.40** | −.33 | −.50 | −6.29** | −.53 | .14 | 2.84** | .23 |
| Relationship Power × Situational Power | .16 | 2.71** | .27 | −.05 | −.66 | −.06 | .10 | 2.34** | .19 |
Note. Effects that test our specific predictions are displayed in bold. The final column presents the interaction effects testing whether the main and interaction effects of relationship power and situational power significantly differ across men and women. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √(t2 / t2 + df).
p < .07.
p < .05.
p < .01.
Figure 5. The interaction between relative relationship power and situational power (need of partner support) on men’s aggressive responses within couples’ discussions of men’s personal goals (Study 4).
Note: Relative relationship power was assessed as in Study 1 (incorporating the partner into one’s identity more versus less than the partner incorporated the individual into their identity). Aggressive responses were examined within the context of discussing an important personal goal with the partner. Low situational power represented greater need for support, which was indexed by greater levels of distress and goal-related hopelessness. Low and high values are indexed at 1 SD below and above the mean.
Model 2: Perceived Relationship Power
The results from dyadic analyses modeling perceptions of relationship power are shown in the bottom of Table 7. These results replicate the effects in Study 3 as well as those that emerged when modeling relative relationship power in this study. The interaction between relationship and situational power was significant for men (first column), but not women (second column), and this gender difference was significant (final column). The significant interaction for men is shown in Figure 6. Lower relationship power was associated with greater aggressive responses within couples’ discussion when men needed higher levels of support and therefore had low situational power (b = −0.32, t = −3.41, p < .01, r = −.33) but not when men’s support need and dependence was low (b = .04, t = .37, p = .71, r = .04).
Figure 6. The interaction between perceptions of relationship power and situational power (need of partner support) on men’s aggressive responses within couples’ discussions of men’s personal goals (Study 4).
Note: Perceptions of relationship power were assessed using the sense of power scale as in Study 3. Aggressive responses were examined within the context of discussing an important personal goal with the partner. Low situational power represented greater need for support, which was indexed by greater levels of distress and goal-related hopelessness. Low and high values are indexed at 1 SD below and above the mean.
Alternative moderators and explanations
As in Studies 1–3, the effects shown in Table 7 and Figures 5 and 6 remained when controlling for the main and interaction effects of self-esteem, attachment anxiety or attachment avoidance. Controlling for age, relationship status (cohabiting/married versus not) or relationship length also did not alter the effects.
Discussion
Study 4 replicated the effects of relationship power and situational power (support need) shown during daily life in Study 3 using methods that captured aggressive responses within couples’ support-relevant discussions of personal goals. The results also replicated using a dyadic index of relative relationship power (as in Study 1 and 2) as well as perceptions of relationship power (as in Study 3). The aggressive reactions shown by men low in relationship power when they are dependent and need support demonstrate that low power men do not aggress to coerce partners or “get their way”. Instead, aggression helps to reestablish power and self-reliance—core elements of masculinity (Bem, 1974; Bem, 1981; Levant et al., 1992; Mahalik et al., 2003). In Study 5, we more directly test the masculinity threat that we argue underlies men’s aggressive reactions to low relationship power.
Study 5
Study 5 was designed to test the theoretical mechanism underlying our hypothesis—that men with low relationship power respond with aggression when facing low situational power because low power threatens masculinity. To do this, we examined the links between perceptions of relationship power and daily situational power, feelings of manliness and aggressive responses across a 2-week period. To assess levels of situational power we gathered reports of partners’ willingness to change their own behavior, preferences, or goals to resolve relationship problems. As with partner avoidance in Study 2, partners’ unwillingness to change directly undercuts power because it conveys that the partner is (a) unwilling to be influenced or change for the good of the relationship and (b) prioritizes their own desires above what is needed to care for their relationship, which (c) reduces the degree to which individuals can sustain their connection and fulfill their own goals and needs.
Our predictions outlining the key role of masculinity underlying men’s aggressive responses to low power are shown in Figure 7. We predicted that men (but not women) low in relationship power should feel less manly on days when they faced lower situational power because their partner was unwilling to change (see Figure 7, Path A), and it is these drops in feelings of masculinity that should, in turn, predict aggressive responses by men (Path B).
Figure 7. The mediating role of feeling manly linking the interaction between relationship power and situational power to men’s daily aggressive behavior (Study 5).
Method
Participants
Participants were 117 newlywed couples (total N = 234 individuals) married for less than three months who participated in a broader study of marriage of 120 couples (3 couples could not be included in the current analyses due to husbands not completing diary reports [n=2] or the couple was lesbian [n=1]). On average, husbands were 31.91 (SD = 9.77) and wives were 29.93 (SD = 7.91) years old. The majority of husbands (70%) and wives (61%) were employed full time, and the majority of the sample (77%) identified as Caucasian.
Procedure
Couples completed a battery of surveys online, including the Sense of Power Scale, and attended a laboratory session at which they completed a variety of tasks beyond the scope of the current analyses. Over the following 14 days, both spouses independently completed an end-of-day web-based record regarding their relationship experiences and behavior that day. On average, participants completed 12.25 diary entries (88%), for a total of 2,854 entries across the entire sample. Descriptive statistics for the diary measures are shown in Table 3. Couples were paid $100 for completing the surveys and lab session and an additional $35 for completing all 28 diary records, or $1.00 per diary record if they failed to complete all 28.
Materials
Relationship Power
As in Studies 3 and 4, participants completed the Sense of Power Scale with reference to their relationship (α = .83).
Alternative Moderators
Participants completed the scales used in Study 2 to assess self-esteem, attachment anxiety, and attachment avoidance (αs > .87). Associations with relationship power were similar to those found in Studies 3 and 4 (see Table 5).
Daily Situational Power
At the end of each day, each partner rated “How willing are you to change your own behavior, preferences, or goals to resolve the problems that exist in your relationships?” (1 = not at all, 7 = very much). As described above, we indexed situational power according to the partners’ ratings of their willingness to change for the good of the relationship. Lower perceptions of relationship power were not significantly associated with partners’ reporting lower willingness to change for men (B = .13, t = 1.61, p = .11) or women (B = −.09, t = −0.92, p = .36; gender difference t = 1.77, p = .08).
Daily Feelings of Masculinity
To measure the degree to which low relationship and situational power threatened masculinity, each day participants also rated how much they felt “manly” today (1 = not at all, 7 = very much). Masculinity involves a set of qualities and characteristics that, although are considered appropriate to men, are not restricted to one biological sex (Bem, 1974, 1981; Mahalik et al., 2003). Thus, women vary in their feelings and presentation of masculinity (as men do in their femininity; Francis, 2010; Halberstam, 1998). Accordingly, despite the average ratings of feeling manly being lower for women than men (as is expected and typical), women still demonstrated reasonable variation in feelings of masculinity across persons and days (see bottom of Table 3).
Daily Aggression
On each day, individuals also responded to the question: “Did you do something today that you think your partner did not like?” When participants answered “yes” (on average 19% of the days completed for both men and women), they were then asked to describe what they did that their partner did not like. To determine whether described behaviors toward the partner were aggressive, two coders experienced in coding communication in couples’ conflict discussions independently classified whether each description represented daily aggression toward the partner (coded 1) or not (coded 0). The coding taxonomy used for coding followed those established schedules used to code aggressive communication in Studies 1, 2 and 4, including verbal aggression, hostility, derogating and blaming the partner, using threats, and rejecting and invalidating the partner. Examples of descriptions classified as aggressive involved yelling, fighting, criticizing, expressing anger, blaming, accusing, and snapping or saying hurtful things. Coder reliability was high (Cohen’s Kappa = .93, t = 21.82, p < .001), and final classification of the few instances of disagreement was determined by consensus. Days in which aggressive behavior was described constituted a total of 6.3% and 8.8% of days for men and women.
Results
All analyses followed the same dyadic modeling strategy described in Study 3 and as recommended by Kenny et al. (2006) and Bolger and Laurenceau (2013) to analyze repeated measures dyadic data, but we used the HLM 6.08 computer program (Bryk, Raudenbush, & Congdon, 2004) in order to model the categorical outcome of aggressive behavior (i.e., whether aggressive behavior occurred on a given day). We ran two nested analyses in order to examine the theoretical process outlined in Figure 7. Our first analysis (Model 1) tested Path A by examining whether men (but not women) low in relationship power felt less manly on days when they faced lower situational power. Our second analysis (Model 2) tested Path B by examining whether such drops in felt masculinity, in turn, predicted daily aggression by men (and not women). We then calculated indirect effects testing whether men (but not women) low in relationship power were more likely to be aggressive on days they encountered low situational power via drops in felt masculinity—the mediation chain shown in Figure 7 linking Path A (Model 1) and Path B (Model 2) together.
Model 1: Does low power reduce men’s feelings of masculinity?
To test the first path of our mediation model (Path A, Figure 7), feeling of manliness on a given day was modeled as a function of: (a) relationship power (mean-centered), (b) situational power (the partner’s person-centered reported willingness to change that day), and (c) the interaction between situational power and relationship power. As in Study 3, we also modeled average levels of partners’ reported willingness to change so that we were isolating the degree to which feelings of manliness changed as daily situational power varied from partners’ typical willingness to change (Bolger & Laurenceau, 2013; Bryk & Raudenbush, 2002). As in Studies 1 to 4, we simultaneously estimated the parameters for men and women using a two-intercept model that controlled for the dyadic dependencies in the data, and tested whether the differences across men and women were significant by pooling the effects across partners and modeling the main and interaction effects of gender (−1 women, 1 men). As shown in Table 8, the interaction between relationship power and within-person fluctuations in situational power significantly predicted men’s, but not women’s, feelings of manliness, and the gender difference was marginally significant. Figure 8 displays the significant interaction for men. Lower relationship power was associated with men reporting feeling less manly on days when situational power was low because partners were not willing to change for the relationship (b = .36, t = 2.08, p = .04, r = .19) but not on days of high situational power when partners were willing to change (b = .14, t = 0.82, p = .41, r = .08).
Table 8.
The effects of Relationship Power and Situational Power (Partners’ Willingness to Change) on Feeling Manly (Study 5)
| Power Variables | Main and Interaction Effects for Men | Main and Interaction Effects for Women | Interaction Effects testing Gender Differences | ||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| B | t | r | B | t | r | B | t | r | |
| Relationship Power | .25 | 1.49 | .14 | .10 | 1.42 | .13 | .06 | 0.71 | .05 |
| Situational Power | .05 | 1.25 | .12 | −.01 | −0.32 | .03 | .04 | 1.51 | .10 |
| Relationship Power × Situational Power | −.11 | −2.74** | .25 | .01 | 0.09 | .01 | −.05 | −1.69† | .11 |
Note. Effects that test our specific predictions are displayed in bold. The final column presents the interaction effects testing whether the main and interaction effects of relationship power and situational power significantly differ across men and women. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √(t2 / t2 + df). Between-person degrees of freedom were used for calculating the effect sizes for the interactions between relationship power and situational power (and associated gender interactions) shown in bold.
p < .10.
p < .05.
p < .01.
Figure 8. The interaction between relationship power and daily situational power (partner willingness to change) on men’s daily reports of feeling manly (Study 5).
Note: Perceptions of relationship power were assessed using the sense of power scale as in Study 3. Situational power was indexed by how willing their partner was to change their own behavior, preferences or goals each day across a 2-week period. Low and high values are indexed at 1 SD below and above the mean.
Model 2: Do threats to masculinity predict aggression toward partners?
To test the second path of our mediation model (Path B, Figure 7), we tested whether reductions in daily feelings of manliness (person-centered) predicted a greater likelihood that men (and not women) behaved aggressively toward their partner that day (0 = no aggression today, 1 = aggression today). We controlled for average levels of manliness in order to test whether within-person reductions in feelings of masculinity predicted aggression. We entered all of the predictors from Path A of our model (relationship power, situational power and the relationship x situational power interaction) to calculate the pathway and associated indirect effect between the effects of relationship x situational power on feelings of masculinity (Path A, Figure 7) and, in turn, the likelihood of aggressive responses toward the partner (Path B, Figure 7) as recommended by MacKinnon, Fritz, Williams, and Lockwood, 2007.
The results supported the mediation pathway outlined in Figure 7. As predicted, on days that men experienced reductions in feelings of masculinity they were more likely to report aggressive responses toward their partner (B = −.35, t = −3.42, p < .001, r = −.31). This was not true for women (B = −.004, t = −0.03, p = .98, r = −.003) and the gender difference was significant (B = −.19, t = −2.27, p = .02, r = −.21). Moreover, the indirect effect linking the relationship power x situational power effects on aggressive behavior via felt masculinity was significant for men (indirect effect = .04, 95% CI [.01, .08]) and not significant for women (indirect effect = .00, 95% CI [−.01, .01]).2 Thus, low relationship power predicted reduced feelings of manliness on days in which men faced low situational power, which in turn predicted a greater likelihood of men behaving aggressively toward the partner that day.
Alternative moderators and explanations
As in Studies 1–5, additional analyses revealed that these effects were not due to self-esteem, attachment anxiety and avoidance. Controlling for age and length of relationship prior to marriage also did not alter the effects.
Discussion
Study 5 extended the results of Studies 1–4 by testing the theoretical mechanism underlying men’s aggressive reactions to low relationship power. As predicted, men who had lower relationship power reported feeling less manly on days they faced low situational power because their partner was unwilling to change for the good of their relationship. These within-person reductions in manliness, in turn, predicted a great likelihood of behaving aggressively toward the partner that day. Women did not show these effects.
General Discussion
In the present research, we made a distinction between the power individuals generally possess in their relationship (relationship power) and the power they experience within power-relevant situations (situational power) in which people’s lack of influence or heightened dependence could produce aggressive responses to restore power. The results of five studies using different measures of power and aggression (see Table 1) supported our predictions. Low relationship power was associated with greater aggressive responses during relationship interactions, but (1) only when low levels of influence (Studies 1–3) or high levels of dependence (Studies 3–4) produced low situational power, and (2) only when low relationship and situational power were experienced by men. Study 5 also demonstrated that men respond aggressively to low power because low power threatens masculinity; men low in relationship power experienced within-person reductions in felt masculinity on days they had low situational power, and such drops in felt masculinity predicted a greater probability of men behaving aggressively toward their partner. Next, we discuss the importance of adopting a contextual perspective to understand when, why and for who power shapes interpersonal behavior, including psychological aggression in close relationships.
The Contextual Nature of Power
Although relationship power involves a relatively enduring state of greater dependence and lower influence across a specific relationship, our results illustrate that power is also an integral element of the specific context or situation that people are currently negotiating. Indeed, it is when situational power is undermined that lacking relationship power becomes consequential. Accordingly, low power men did not respond aggressively across all relationship interactions, but only responded aggressively in situations in which they experienced low levels of influence or high levels of dependence. In contrast, low situational power was not associated with aggression by men high in relationship power, which illustrates that moments of low power are relatively inconsequential for people who are not as heavily dependent on their partner or get their way the majority of the time.
This contextual pattern illustrates that failing to assess power dynamics within power-relevant situations in which people need to influence or depend on others to attain key goals and needs, and failing to use methods that capture variation in such situational power, will obscure the effects of power. This is likely one central reason why prior research relying on broad measures of relationship power has produced inconsistencies, particularly with regard to the links between relationship power and aggression. Low relationship power will instigate aggressive responses by men only (or predominantly) within situations in which their lack of influence and heightened dependence is consequential. In the absence of unsuccessful influence attempts and dependence on the partner, there may be little or no link between power and aggression. The tendency of general self-reports to gloss over these crucial contextual factors helps to explain the inconsistencies in prior studies, and this is likely to be the case for other important outcomes of power (e.g., compliance, sacrifice, empathy).
Indeed, the underlying implications of power-relevant situations are central to explaining why power has the effects it does. When people are unable to influence partners in desired ways or when they need support from their partners they are placed in a vulnerable, one-down position (Kelley et al., 2003; Thibaut & Kelley, 1959). Our results indicate that, because the possession and demonstration of power is central to “being a man” (Bosson & Vandello, 2011; Kimmel, 2008; Vescio, et al., 2010), these situations threaten masculinity and instigate aggressive responses by men to restore their masculinity. That non-conflictual dependence arising from needing support produced aggressive responses by low power men also supports that men’s aggression most likely represents efforts to repair power and restore masculinity rather than instrumental acts oriented toward “getting one’s way” and coercing change. Our final study also provided direct evidence that thwarting influence undermines men’s feelings of manliness, and such drops in masculinity activate greater aggressive responses toward partners to rebalance power and reestablish masculinity.
The importance of situational power in determining when people will enact power-restoring strategies is consistent with other research recognizing that the effects of power depend on context (e.g., Chen et al., 2001; Gordon & Chen, 2013). We examined the effects of power in a specific context—couples in committed, interdependent relationships. In contrast, many studies have examined the effects of power within hierarchical relationships that involve stable power differentials (e.g., boss-employee, leader-subordinate, parent-child). Similar effects of experiencing low relational power have emerged in those studies, but specifically for people who hold role-based positions of power, such as parents and leaders (e.g., Bugental & Lin, 2001). In contrast, people in subordinate roles cannot readily use aggression as a way to restore power because of the punishment people in high power positions can enact. Thus, the effects in the current studies may be most applicable to people whose role affords them the ability to aggress when they encounter low power situations.
In the types of committed relationships we investigated, both partners are in positions of power. In particular, partners low in relationship power do not occupy subordinate roles because even partners who have high relationship power are invested and dependent to some degree. Thus, men lower in power could afford to respond aggressively because, at least temporarily, their restoration of power would not be accompanied by dire consequences, such as the loss of their relationship. That high power partners were nonetheless invested is also likely central to why they were not threatened by any given instance of dependence or low influence and were thus able to respond in ways that prioritize relationship goals over defending power concerns (Karremans & Smith, 2010). However, when relative power is extremely divergent, and high power partners have little commitment and dependence, low power partners may not be able to aggress because the risks of retaliation and rejection are too high, and thus they will need to adopt other tactics to manage their lack of power.
The contextual nature of power, and the importance of assessing the effects of power in power-relevant situations, is also important in understanding how additional power-related variables are related to aggression. For example, unlike the possession of low versus high power (that we focused on in the current research), the motive to preserve or gain power has been consistently associated with aggressive responses in research examining both romantic (e.g., Dutton & Strachan, 1987; Kaura & Allen, 2004; Felson & Outlaw, 2007; Ronfeldt et al., 1998; Whitaker, 2013) and non-romantic (e.g., Case & Maner, 2014; Maner & Mead, 2010) contexts. This suggests that the links between low power and aggression shown in the current research will likely be exacerbated for men who have strong motives to sustain power (e.g., Hammond & Overall, 2013; Overall et al., 2011). However, the motive to attain power should be most strongly activated when people are unable to exercise desired influence or are dependent in some way. Thus, identifying when the links between power and aggression are subdued or intensified requires examining power within situations in which power matters.
Our contextual perspective also has implications regarding the measurement of relationship power. Our multiple methods of assessing relationship and situational power uniquely demonstrated that power based on dependence and influence produce similar effects. In doing so, we provide coherence across divergent approaches to power that enhances comparability across the diversity of prior studies. Interestingly, however, dependence-based and influence-based measures of power were not strongly correlated, which indicates that they may be distinct elements or bases of power. Both are central to the power-relevant concerns associated with masculinity, however, and thus both were associated with aggressive responses by men in the current studies. Nonetheless, given the contextual nature of power, dependence and influence may be differentially associated with other predictors and outcomes. Despite the convergence we demonstrated, future studies should consider what type of power is relevant to the contexts and constructs under investigation.
Power, Aggression and Gender: Caveats and Future Directions
The current studies identify power as one central factor in understanding the different conditions that predict men’s and women’s psychological aggression. Restoring power via aggression should be particularly relevant to men because demonstrating power is a key component of masculinity. In contrast, femininity does not involve asserting independence or power, but rather involves comfort with dependence and prioritizing relational wellbeing (Bosson & Vandello, 2011; Cross & Madson, 1997). We are not suggesting, however, that the important role of power as a predictor of men’s aggressive responses indicates that men will behave more aggressively in relationships compared to women. Indeed, consistent with prior research (Archer, 2004; Esquivel-Santoveña & Dixon, 2012; Straus et al., 1996), women tended to exhibit greater psychological aggression in relationship interactions compared to men in the current studies (see Table 2). Thus, our research identifies power as one important risk factor that is relevant to understanding men’s aggressive responses in relationships, but specific risk factors for women are important to consider in future research.
Additionally, our theoretical analysis and measurement focused on interpersonal power derived from relative dependence and the ability to influence partners within dyadic interactions. Power can also be analyzed at the societal level, such as economic and political inequalities, and societal power is also relevant to understanding gender differences in the links between power and aggression (Archer, 2006). There was some evidence in Study 3 that women lower in relationship power reacted less aggressively than women higher in power on days they were unable to influence their partners. Perhaps women lower in power inhibit their aggression because of nurturing gender-role expectations, associated social sanctions, and/or because women compared to men are more vulnerable due to economic dependence and this lower structural power reduces the degree to which women can respond aggressively. However, the reverse effect for women was only suggestive in one study. On the other hand, our samples consisted of relatively well-educated and satisfied couples embedded within relatively egalitarian cultural contexts. The inhibition of women’s aggression may emerge or be particularly pronounced when widespread social inequities impede women’s ability to aggress (e.g., economic dependence, lack of viable alternatives).
Gender differences are also likely to be more pronounced when assessing more severe forms of physical aggression and interpersonal violence (Archer, 2000; Archer, 2002), particularly within societies marked by stronger ideological support for the aggressive maintenance of men’s power (Archer, 2006). In the current research, we focused on more common forms of psychological aggression, such as criticizing and derogating partners. It is well established these types of aggressive responses have detrimental effects on partners’ health and wellbeing as well as the quality and stability of the relationship (see Gottman, 1998; Karney & Bradbury, 1995; Kiecolt-Glaser & Newton, 2001; Woodin, 2011). Not only are these forms of aggression as psychologically harmful as physical forms of aggression (Arriaga & Schkeryantz, 2015), they also precede physical aggression in marriage (Murphy & O’Leary, 1989) and mediate the links between relationship power and self-reported relationship violence (Babcock et al., 1993; Leonard & Senchak, 1996). Thus, our results highlight an important set of conditions that activate common forms of aggressive responses that undermine relationship functioning and can lead to more severe physical violence.
However, although it is important to assess naturally-occurring aggressive responses as couples negotiate actual power-relevant interactions, our ecologically-valid methods rely on correlational data that prevent causal conclusions. The reverse causal directions are less theoretically plausible. Although aggressive responses are likely to undermine influence by generating partner resistance (i.e., low situational power), it is unlikely this would be the case only for men—and not women—low in power, particularly given that demanding responses by women are more likely to elicit avoidance by male partners and women’s lower social power plays a role in this pattern (e.g., Christensen & Heavey, 1990; Eldridge & Christensen, 2002). The alternative pathway is even less plausible when considering aggressive responses within low power situations involving dependence. It is very unlikely that acts of aggression would lead low power men—and not women—to need more support, particularly given aggressive responses are implicated in the restoration of independence, control, and masculinity (e.g., Leary et al., 2006; Lemay et al., 2012; Vandello et al., 2008), all of which are undermined when people are dependent on their partner for support (e.g., Murray & Holmes, 2009; Mahalik et al., 2003). Finally, a reverse account of the mediating role of felt masculinity would suggest that aggressive responses predict drops in feeling manly, which is in direct contrast to masculinity involving the assertion of power, control and dominance.
The assessment of physical aggression, the additional moderating role of societal-level power, and the use of additional methods to strengthen causal conclusions, are all important to address in future research. However, by sampling well-functioning couples living in egalitarian societies, and by assessing frequent forms of aggressive responding in couples’ actual interactions, our studies illustrate the relevance of power threats for men in the course of typical day-to-day relationship life. The overall pattern also indicates that men’s difficulty with dependence will have a sweeping impact on their relationships and do so beyond conflictual interactions involving attempts to influence partners. Men who find dependence threatening will struggle to navigate a range of situations central to building and sustaining relationships (e.g., support, disclosure, emotional and physical intimacy). Moreover, recognizing the identity threat associated with men’s dependence and low influence may inform therapeutic attempts to prevent aggressive responses in relationships. In particular, aggression is only one way for men to embody and demonstrate masculinity. Helping men to generate more constructive ways of restoring masculinity, such as assertive but non-aggressive communication of needs and goals, should help reduce psychological (and perhaps physical) aggression in intimate relationships.
Conclusion
The current research demonstrates that understanding the effects of power requires assessing the degree to which people possess power across a particular social relationship (relationship power) as well as the degree to which people are experiencing power within specific power-relevant situations that involve the need to influence or depend on relationship partners to attain key goals and needs (situational power). Across five studies, and multiple methods of assessing power and psychological aggression, the results illustrate that low relationship power can produce aggressive responses, but only when low situational power caused by a lack of influence or heightened dependence impedes men’s ability to negotiate relationship interactions in ways that uphold masculine identities. Unfortunately, these results suggest the need to possess and demonstrate power to ‘be a man’ represents a significant risk factor for psychological aggression in relationships. More broadly, the results demonstrate that to fully understand when and why power is associated with any type of response requires differentiating between the level of relationship power people generally hold and the level of situational power people experience within specific power-relevant social interactions.
Supplementary Material
Acknowledgments
This research was supported by grant UoA FRDF 3626244 awarded to Nickola C. Overall and grants HD058314 and BCS-1251520 awarded to James K. McNulty.
Footnotes
Our investigations, and the majority of those reviewed in the following sections, focus on interdependent relationships involving highly committed couples. Although relative differences in dependence and influence occur in these types of close relationships, high power partners are still invested in their relationships and so aggressive responses by low power individuals may restore power without running the risk of losing their relationship, at least in the moment. However, when power differentials are very high, aggressive responses to assert power may be inhibited by the greater risk of partner punishment, rejection and abandonment faced by those lower in power. We consider these contexts and other important factors that may modify the connections between power and aggression in the general discussion.
The direct relationship power x situational power interaction effect on aggressive behavior was not significant for men (B = .07, t = 0.48, p = .63, r = .05) or women (B = .12, t = 0.89, p = .37, r = .09). We conducted a meta-analysis across studies to assess the relative strength and robustness of the relationship power x situational power interaction predicting aggressive responses. We focused specifically on situational power due to low influence as was assessed in Study 5 (although adding in the effects when assessing situations of dependence only strengthened the results). As expected, the aggression-inducing effect of low relationship power x low situational power was significant and robust for men (r = .22, z = 4.21, p < .001, 95% CI [.12, .32]) and not women (r = −.09, z = −1.29, p = .20, 95% CI [−.23, .05]).
Contributor Information
Nickola C. Overall, School of Psychology, University of Auckland
Matthew D. Hammond, School of Psychology, University of Auckland
James K. McNulty, Department of Psychology, Florida State University
Eli J. Finkel, Department of Psychology and the Kellogg School of Management, Northwestern University
References
- Agnew CR, Van Lange PAM, Rusbult CE, Langston CA. Cognitive interdependence: Commitment and the mental representation of close relationships. Journal of Personality and Social Psychology. 1998;74:939–954. [Google Scholar]
- Anderson C, Galinsky AD. Power, optimism, and risk-taking. European Journal of Social Psychology. 2006;36:511–536. [Google Scholar]
- Anderson C, John OP, Keltner D. The personal sense of power. Journal of Personality. 2012;80:313–344. doi: 10.1111/j.1467-6494.2011.00734.x. [DOI] [PubMed] [Google Scholar]
- Anderson C, Keltner D, John OP. Emotional convergence between people over time. Journal of Personality and Social Psychology. 2003;84:1054–1068. doi: 10.1037/0022-3514.84.5.1054. [DOI] [PubMed] [Google Scholar]
- Archer J. Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin. 2000;126:651–680. doi: 10.1037/0033-2909.126.5.651. [DOI] [PubMed] [Google Scholar]
- Archer J. Sex differences in physically aggressive acts between heterosexual partners: A meta-analytic review. Aggression and Violent Behavior: A Review Journal. 2002;7:313–351. [Google Scholar]
- Archer J. Sex differences in aggression in real-world settings: A meta-analytic review. Review of General Psychology. 2004;8:291–322. [Google Scholar]
- Archer J. Cross-cultural differences in physical aggression between partners: A social-role analysis. Personality and Social Psychology Review. 2006;10:133–153. doi: 10.1207/s15327957pspr1002_3. [DOI] [PubMed] [Google Scholar]
- Aron A, Aron EN, Smollan D. Inclusion of Other in the Self Scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology. 1992;63:596–612. [Google Scholar]
- Attridge M, Berscheid E, Simpson JA. Predicting relationship stability from both partners versus one. Journal of Personality and Social Psychology. 1995;69:254–268. [Google Scholar]
- Arriaga XB, Schkeryantz EL. Intimate relationships and personal distress: The invisible harm of psychological aggression. Personality and Social Psychology Bulletin. 2015;14:1332–1344. doi: 10.1177/0146167215594123. [DOI] [PubMed] [Google Scholar]
- Babcock JC, Waltz J, Jacobson NS, Gottman JM. Power and violence: The relation between communication patterns, power discrepancies, and domestic violence. Journal of Consulting and Clinical Psychology. 1993;61:40–50. doi: 10.1037//0022-006x.61.1.40. [DOI] [PubMed] [Google Scholar]
- Bar-Kalifa E, Rafaeli E. Disappointment’s sting is greater than help’s balm:Quasi-signal detection of daily support matching. Journal of Family Psychology. 2013;27:956–967. doi: 10.1037/a0034905. [DOI] [PubMed] [Google Scholar]
- Bem SL. The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology. 1974;42:155–162. [PubMed] [Google Scholar]
- Bem SL. Bem Sex Role Inventory: Professional manual. Palo Alto, CA: Consulting Psychologists Press; 1981. [Google Scholar]
- Bentley CG, Galliher RV, Ferguson TJ. Associations among aspects of interpersonal power and relationship functioning in adolescent romantic couples. Sex Roles. 2007;57:483–495. doi: 10.1007/s11199-007-9280-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Berman MI, Frazier PA. Relationship power and betrayal experience as predictors of reactions to infidelity. Personality and Social Psychology Bulletin. 2005;31:1617–1627. doi: 10.1177/0146167205277209. [DOI] [PubMed] [Google Scholar]
- Bornstein RF. The complex relationship between dependence and domestic violence. American Psychologist. 1996;61:595–606. doi: 10.1037/0003-066X.61.6.595. [DOI] [PubMed] [Google Scholar]
- Bosson JK, Vandello JA. Precarious manhood and its links to action and aggression. Current Directions in Psychological Science. 2011;20:82–86. [Google Scholar]
- Bosson JK, Vandello JA, Burnaford RM, Weaver JR, Wasti SA. Precarious manhood and displays of physical aggression. Personality and Social Psychology Bulletin. 2009;35:623–634. doi: 10.1177/0146167208331161. [DOI] [PubMed] [Google Scholar]
- Brennan KA, Clark CL, Shaver PR. Self-report measurement of adult attachment: An integrative overview. In: Simpson JA, Rholes WS, editors. Attachment theory and close relationships. New York: Guilford; 1998. pp. 394–428. [Google Scholar]
- Bryk AS, Raudenbush SW. Hierarchical linear models: Applications and data analysis methods. 2. Thousand Oaks, CA: Sage; 2002. [Google Scholar]
- Bryk AS, Raudenbush SW, Congdon RT. HLM: Hierarchical linear modeling with the HLM/2L and HLM/3L programs. Chicago: Scientific Software International; 2004. [Google Scholar]
- Bugental DB. Paradoxical power manifestations: Power assertion by the subjectively powerless. In: Guinote A, Vescio TK, editors. The social psychology of power. New York: Guilford Press; 2010. pp. 209–230. [Google Scholar]
- Bugental DB, Lin EK. The many faces of power: The strange case of Dr. Jekyll and Mr. Hyde. In: Lee-Chai AY, Bargh JA, editors. The use and abuse of power: Multiple perspectives on the causes of corruption. New York: Psychology Press; 2001. pp. 115–132. [Google Scholar]
- Burke LK, Follingstad DR. Violence in lesbian and gay relationships: Theory, prevalence, and correlational factors. Clinical Psychology Review. 1999;19:487–512. doi: 10.1016/s0272-7358(98)00054-3. [DOI] [PubMed] [Google Scholar]
- Buss DM, Schmitt DP. Sexual strategies theory: An evolutionary perspective on human mating. Psychological Review. 1993;100:204–232. doi: 10.1037/0033-295x.100.2.204. [DOI] [PubMed] [Google Scholar]
- Case CR, Maner JK. Divide and conquer: When and why leaders undermine the cohesive fabric of their group. Journal of Personality and Social Psychology. 2014:173. doi: 10.1037/a0038201. [DOI] [PubMed] [Google Scholar]
- Choi SYP, Ting K. Wife beating in South Africa: An imbalance theory of resources and power. Journal of Interpersonal Violence. 2008;23:834–852. doi: 10.1177/0886260507313951. [DOI] [PubMed] [Google Scholar]
- Chen S, Lee-Chai AY, Bargh JA. Relationship orientation as a moderator of the effects of social power. Journal of Personality and Social Psychology. 2001;80:173. doi: 10.1037/0022-3514.80.2.173. [DOI] [PubMed] [Google Scholar]
- Christensen A, Heavey CL. Gender and social structure in the demand/withdraw pattern of marital conflict. Journal of Personality and Social Psychology. 1990;59:73–81. doi: 10.1037//0022-3514.59.1.73. [DOI] [PubMed] [Google Scholar]
- Collins NL, Feeney BC. A safe haven: An attachment theory perspective on support seeking and caregiving in intimate relationships. Journal of Personality and Social Psychology. 2000;78:1053–1073. doi: 10.1037//0022-3514.78.6.1053. [DOI] [PubMed] [Google Scholar]
- Collins NL, Feeney BC. Working models of attachment shape perceptions of social support: Evidence from experimental and observational studies. Journal of Personality and Social Psychology. 2004;87(3):363–383. doi: 10.1037/0022-3514.87.3.363. [DOI] [PubMed] [Google Scholar]
- Crockett EE, Neff LA. When receiving help hurts: Gender differences in diurnal cortisol responses to spousal support. Social Psychological and Personality Science. 2012;4:190–197. [Google Scholar]
- Cross SE, Madson L. Model of the self: Self-construals and gender. Psychological Bulletin. 1997;122:5–37. doi: 10.1037/0033-2909.122.1.5. [DOI] [PubMed] [Google Scholar]
- Cutrona CE, Shaffer PA, Wesner KA, Gardner KA. Optimally matching support and perceived spousal sensitivity. Journal of Family Psychology. 2007;21(4):754–758. doi: 10.1037/0893-3200.21.4.754. [DOI] [PubMed] [Google Scholar]
- Dahl J, Vescio T, Weaver K. How threats to masculinity sequentially cause public discomfort, anger, and ideological dominance over women. Social Psychology. 2015;46:242–254. [Google Scholar]
- Diamond PM, Magaletta PR. The Short-Form Buss-Perry Aggression Questionnaire (BPAQ-SF): A validation study with federal officers. Assessment. 2006;13:227–240. doi: 10.1177/1073191106287666. [DOI] [PubMed] [Google Scholar]
- Dobash RE, Dobash RP. Violence against wives: A case against the patriarchy. New York, NY: Free Press; 1979. [Google Scholar]
- Dunbar NE, Burgoon JK. Perceptions of power and interactional dominance in interpersonal relationships. Journal of Social and Personal Relationships. 2005;22:231–257. [Google Scholar]
- Dutton DG. Intimate abusiveness. Clinical Psychology: Science and Practice. 1995;2:207–224. [Google Scholar]
- Dutton DG. Rethinking domestic violence. Vancouver: UCB Press; 2006. [Google Scholar]
- Dutton DG, Strachan CE. Motivational needs for power and spouse-specific assertiveness in assaultive and nonassaultive men. Violence and Victims. 1987;2:145–456. [PubMed] [Google Scholar]
- Eisler RM, Skidmore JR. Masculine gender role stress: Scale development and component factors in the appraisal of stressful situations. Behavior Modification. 1987;11:123–136. doi: 10.1177/01454455870112001. [DOI] [PubMed] [Google Scholar]
- Eldridge KA, Christensen A. Demand-withdraw communication during couple conflict: A review and analysis. In: Noller P, Feeney JA, editors. Understanding marriage: Developments in the study of couple interaction. New York, NY: Cambridge University Press; 2002. pp. 289–322. [Google Scholar]
- Esquivel-Santoveña EE, Dixon L. Investigating the true rate of physical intimate partner violence: A review of nationally representative surveys. Aggression and Violent Behavior. 2012;17:208–219. [Google Scholar]
- Farrell AK, Simpson JA, Rothman AJ. The relationship power inventory: Development and validation. Personal Relationships. 2015;22:387–413. [Google Scholar]
- Fast NJ, Chen S. When the boss feels inadequate: Power, incompetence and aggression. Psychological Science. 2009;20:1406–1413. doi: 10.1111/j.1467-9280.2009.02452.x. [DOI] [PubMed] [Google Scholar]
- Feeney BC. A secure base: Responsive support of goal strivings and explorations in adult intimate relationships. Journal of Personality and Social Psychology. 2004;87:631–648. doi: 10.1037/0022-3514.87.5.631. [DOI] [PubMed] [Google Scholar]
- Felmlee DH. Who's on top? Power in romantic relationships. Sex Roles. 1994;31:275–295. [Google Scholar]
- Felson RB, Outlaw MC. The control motive and marital violence. Violence and Victims. 2007;22:387–407. doi: 10.1891/088667007781553964. [DOI] [PubMed] [Google Scholar]
- Fletcher GJO, Simpson JA, Thomas G. The measurement of perceived relationship quality components: A confirmatory factor analytic approach. Personality and Social Psychology Bulletin. 2000;26:340–354. [Google Scholar]
- Finkel EJ, DeWall CN, Slotter EB, McNulty JK, Pond RS, Jr, Atkins DC. Using I3 theory to clarify when dispositional aggressiveness predicts intimate partner violence perpetration. Journal of Personality and Social Psychology. 2012;102:533–549. doi: 10.1037/a0025651. [DOI] [PubMed] [Google Scholar]
- Fischer AH, Roseman IJ. Beat them or ban them: The characteristics and social functions of anger and contempt. Journal of Personality and Social Psychology. 2007;93:103–115. doi: 10.1037/0022-3514.93.1.103. [DOI] [PubMed] [Google Scholar]
- Fitness J, Fletcher GJ. Love, hate, anger, and jealousy in close relationships: A prototype and cognitive appraisal analysis. Journal of Personality and Social Psychology. 1993;65:942–958. doi: 10.1037//0022-3514.65.5.942. [DOI] [PubMed] [Google Scholar]
- Fraley RC, Waller NG, Brennan KA. An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology. 2000;78:350–365. doi: 10.1037//0022-3514.78.2.350. [DOI] [PubMed] [Google Scholar]
- Francis B. Re/theorising gender: Female masculinity and male femininity in the classroom? Gender and Education. 2010;22:477–490. [Google Scholar]
- Galinsky AD, Gruenfeld DH, Magee JC. From power to action. Journal of Personality and Social Psychology. 2003;85:453–466. doi: 10.1037/0022-3514.85.3.453. [DOI] [PubMed] [Google Scholar]
- Galinsky AD, Magee JC, Inesi ME, Gruenfeld DH. Power and perspectives not taken. Psychological science. 2006;17:1068–1074. doi: 10.1111/j.1467-9280.2006.01824.x. [DOI] [PubMed] [Google Scholar]
- Gelles RJ, Straus MA. Determinants of violence in the family: Toward a theoretical integration. In: Burr WR, Hill R, Nye FI, Reiss IL, editors. Contemporary Theories about the Family. New York: The Free Press; 1979. pp. 549–581. [Google Scholar]
- Georgesen JC, Harris MJ. Why’s my boss always holding me down?: A meta-analysis of power effects on performance evaluation. Personality and Social Psychology Review. 1998;2:184–195. doi: 10.1207/s15327957pspr0203_3. [DOI] [PubMed] [Google Scholar]
- Girme YU, Overall NC, Simpson JA. When visibility matters: Short-term versus long-term costs and benefits of visible and invisible support. Personality and Social Psychology Bulletin. 2013;39:1441–1454. doi: 10.1177/0146167213497802. [DOI] [PubMed] [Google Scholar]
- Girme YU, Overall NC, Simpson JA, Fletcher GJO. “All or nothing”:Attachment avoidance and the curvilinear effects of partner support. Journal of Personality and Social Psychology. 2015;108:450–475. doi: 10.1037/a0038866. [DOI] [PubMed] [Google Scholar]
- Goetz SMM, Shattuck KS, Miller RM, Campbell JA, Lozoya E, Weisfeld GE, Carre JM. Social status moderates the relationship between facial structure and aggression. Psychological Science. 2013;24:2329–2334. doi: 10.1177/0956797613493294. [DOI] [PubMed] [Google Scholar]
- Gordon AM, Chen S. Does power help or hurt? The moderating role of self-other focus on power and perspective taking in romantic relationships. Personality and Social Psychology Bulletin. 2013;39:1097–1110. doi: 10.1177/0146167213490031. [DOI] [PubMed] [Google Scholar]
- Gottman JM. Psychology and the study of marital processes. Annual Review of Psychology. 1998;49:169–197. doi: 10.1146/annurev.psych.49.1.169. [DOI] [PubMed] [Google Scholar]
- Gray-Little B, Burks N. Power and satisfaction in marriage: A review and critique. Psychological Bulletin. 1983;93:513–538. [Google Scholar]
- Halberstam J. Female masculinity. Duke University Press; 1998. [Google Scholar]
- Hammond MD, Overall NC. Men’s hostile sexism and biased perceptions of intimate partners: Fostering dissatisfaction and negative behavior in close relationships. Personality and Social Psychology Bulletin. 2013;39:1585–1599. doi: 10.1177/0146167213499026. [DOI] [PubMed] [Google Scholar]
- Heavey CL, Christensen A, Malamuth NM. The longitudinal impact of demand and withdrawal during marital conflict. Journal of Consulting and Clinical Psychology. 1995;63:797–801. doi: 10.1037//0022-006x.63.5.797. [DOI] [PubMed] [Google Scholar]
- Hotaling GT, Sugarman DB. An analysis of risk markers in husband to wife violence: The current state of knowledge. Violence and Victims. 1986;1:101–124. [PubMed] [Google Scholar]
- Holtzworth-Munroe A, Bates L, Smutzler N, Sandin E. A brief review of the research on husband violence. Aggression and Violent Behavior. 1997;2:65–99. [Google Scholar]
- Huston TL. Power. In: Kelley HH, et al., editors. Close relationships. New York: W. H. Freeman; 1983. pp. 169–219. [Google Scholar]
- Inesi ME, Gruenfeld DH, Galinsky AD. How power corrupts relationships: Cynical attributions for others’ generous acts. Journal of Experimental Social Psychology. 2012;48:795–803. [Google Scholar]
- Karremans JC, Smith PK. Having the power to forgive: When the experience of power increases interpersonal forgiveness. Personality and Social Psychology Bulletin. 2010;36:1010–1023. doi: 10.1177/0146167210376761. [DOI] [PubMed] [Google Scholar]
- Karney BR, Bradbury TN. The longitudinal course of marital quality and stability: A review of theory, method and research. Psychological Bulletin. 1995;118:3–34. doi: 10.1037/0033-2909.118.1.3. [DOI] [PubMed] [Google Scholar]
- Kaura SA, Allen CM. Dissatisfaction with relationship power and dating violence perpetration by men and women. Journal of Interpersonal Violence. 2004;19:576–588. doi: 10.1177/0886260504262966. [DOI] [PubMed] [Google Scholar]
- Kelley HH, Holmes JG, Kerr NL, Reis HT, Rusbult CE, Van Lange PAM. An atlas of interpersonal situations. Cambridge, United Kingdom: Cambridge University Press; 2003. [Google Scholar]
- Kelley HH, Thibaut JW. Interpersonal relations: A theory of interdependence. New York: Wiley; 1978. [Google Scholar]
- Keltner D, Gruenfeld DH, Anderson C. Power, approach and inhibition. Psychological Review. 2003;110:265–284. doi: 10.1037/0033-295x.110.2.265. [DOI] [PubMed] [Google Scholar]
- Kenny DA, Cook WL. Partner effects in relationship research: Conceptual issues, analytic difficulties, and illustrations. Personal Relationships. 1999;6:433–448. [Google Scholar]
- Kenny DA, Kashy DA, Cook WL. Dyadic data analysis. New York, NY: Guilford Press; 2006. [Google Scholar]
- Kiecolt-Glaser JK, Newton TL. Marriage and health: His and hers. Psychological Bulletin. 2001;127:427–503. doi: 10.1037/0033-2909.127.4.472. [DOI] [PubMed] [Google Scholar]
- Kifer Y, Heller D, Perunovic WQE, Galinsky AD. The good life of the powerful: The experience of power and authenticity enhances subjective well-being. Psychological Science. 2013;24:280–288. doi: 10.1177/0956797612450891. [DOI] [PubMed] [Google Scholar]
- Kimmel M. Guyland: The perilous world where boys become men. New York: HarperCollins; 2008. [Google Scholar]
- Kuehn MM, Chen S, Gordon AM. Having a thicker skin: Social power buffers the negative effects of social rejection. Social Psychological and Personality Science. 2015;6:701–709. [Google Scholar]
- Leary MR, Twenge JM, Quinlivan E. Interpersonal rejection as a determinant of anger and aggression. Personality and Social Psychology Review. 2006;10:111–132. doi: 10.1207/s15327957pspr1002_2. [DOI] [PubMed] [Google Scholar]
- Lemay EP, Jr, Overall NC, Clark MS. Experiences and interpersonal consequences of hurt feelings and anger. Journal of Personality and Social Psychology. 2012;103:982–1006. doi: 10.1037/a0030064. [DOI] [PubMed] [Google Scholar]
- Leonard KE, Senchak M. Prospective prediction of husband marital aggression within newlywed couples. Journal of Abnormal Psychology, 1996. 1996;3:369–380. doi: 10.1037//0021-843x.105.3.369. [DOI] [PubMed] [Google Scholar]
- Levant RF, Hirsch LS, Celentano E, Cozza TM, Hill S, MacEachern M, … Schnedeker J. The male role: An investigation of contemporary norms. Journal of Mental Health Counseling. 1992;14:325–337. [Google Scholar]
- Lipsey MW, Wilson DB. Practical meta-analysis. London: Sage; 2001. [Google Scholar]
- Loving TJ, Heffner KL, Kiecolt-Glaser JK, Glaser R, Malarkey WB. Stress hormone changes and marital conflict: Spouses’ relative power makes a difference. Journal of Marriage and Family. 2004;66:594–611. [Google Scholar]
- Maass A, Cadinu M, Guarnieri G, Grasselli A. Sexual harassment under social identity threat: The computer harassment paradigm. Journal of Personality and Social Psychology. 2003;85:853–870. doi: 10.1037/0022-3514.85.5.853. [DOI] [PubMed] [Google Scholar]
- Magdol L, Moffitt TE, Caspi PA, Newman DL, Fagan J, Silva PA. Gender differences in partner violence in a birth cohort of 21-year-olds: Bridging the gap between clinical and epidemiological approaches. Journal of Consulting and Clinical Psychology. 1997;65:68–78. doi: 10.1037//0022-006x.65.1.68. [DOI] [PubMed] [Google Scholar]
- Mahalik JR, Locke BD, Ludlow LH, Diemer MA, Scott RP, Gottfried M, Freitas G. Development of the Conformity to Masculine Norms Inventory. Psychology of Men & Masculinity. 2003;4:3–25. [Google Scholar]
- Maner JK, Mead NL. The essential tension between leadership and power: When leaders sacrifice group goals for the sake of self-interest. Journal of Personality and Social Psychology. 2010;99:482–497. doi: 10.1037/a0018559. [DOI] [PubMed] [Google Scholar]
- McNulty JK, Russell VM. When “negative” behaviors are positive: A contextual analysis of the long-term effects of interpersonal communication. Journal of Personality and Social Psychology. 2010;98:587–604. doi: 10.1037/a0017479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mead NL, Maner JK. On keeping your enemies close: Powerful leaders seek proximity to ingroup power threats. Journal of Personality and Social Psychology. 2012;102:576–591. doi: 10.1037/a0025755. [DOI] [PubMed] [Google Scholar]
- Medeiros RA, Straus MA. Risk factors for physical violence between dating partners: Implications for gender-inclusive prevention and treatment of family violence. Family interventions in domestic violence: A handbook of gender-inclusive theory and treatment. 2006:59–85. [Google Scholar]
- Moore TM, Stuart GL. A review of the literature on masculinity and partner violence. Psychology of Men and Masculinity. 2005;6:46–61. [Google Scholar]
- Murphy CM, Meyer SL, O’Leary KD. Dependency characteristics of partner assaultive men. Journal of Abnormal Psychology. 1994;103:729–735. doi: 10.1037//0021-843x.103.4.729. [DOI] [PubMed] [Google Scholar]
- Murphy C, O’Leary KD. Psychological aggression predicts physical aggression in early marriage. Journal of Consulting and Clinical Psychology. 1989;57:579–582. doi: 10.1037//0022-006x.57.5.579. [DOI] [PubMed] [Google Scholar]
- Murray SL, Holmes JG. The architecture of interdependent minds: A motivation-management theory of mutual responsiveness. Psychological Review. 2009;116:908–928. doi: 10.1037/a0017015. [DOI] [PubMed] [Google Scholar]
- O’Leary KD, Tintle N, Bromet E. Risk factors for physical violence against partners in the US. Psychology of Violence. 2014;4:65–77. doi: 10.1037/a0034537. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Oriña MM, Collins WA, Simpson JA, Salvatore JE, Haydon KC, Kim JS. Developmental and dyadic perspectives on commitment in adult romantic relationships. Psychological Science. 2011;22:908–915. doi: 10.1177/0956797611410573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Osgood CE, Suci GJ, Tannenbaum PH. The measurement of meaning. Urbana, IL: University of Illinois; 1957. [Google Scholar]
- Overall NC, Fletcher GJO, Simpson JA. Helping each other grow: Romantic partner support, self-improvement and relationship quality. Personality and Social Psychology Bulletin. 2010;36:1496–1513. doi: 10.1177/0146167210383045. [DOI] [PubMed] [Google Scholar]
- Overall NC, Fletcher GJ, Simpson JA, Sibley CG. Regulating partners in intimate relationships: the costs and benefits of different communication strategies. Journal of Personality and Social Psychology. 2009;96:620–639. doi: 10.1037/a0012961. [DOI] [PubMed] [Google Scholar]
- Overall NC, Sibley CG. Attachment and dependence regulation within daily interactions with romantic partners. Personal Relationships. 2009;16:239–261. [Google Scholar]
- Overall NC, Sibley CG. Convergent and discriminant validity of the Accommodation Scale: Evidence from three diary studies. Personality and Individual Differences. 2010;48:299–304. [Google Scholar]
- Overall NC, Sibley CG, Tan R. The costs and benefits of sexism: resistance to influence during relationship conflict. Journal of Personality and Social Psychology. 2011;101:271–290. doi: 10.1037/a0022727. [DOI] [PubMed] [Google Scholar]
- Overall NC, Simpson JA. Regulation processes in close relationships. In: Simpson JA, Campbell L, editors. The Oxford handbook of close relationships. New York: Oxford University Press; 2013. pp. 427–451. [Google Scholar]
- Pasch L, Bradbury TN. Social support, conflict and the development of marital dysfunction. Journal of Consulting and Clinical Psychology. 1998;66:219–230. doi: 10.1037//0022-006x.66.2.219. [DOI] [PubMed] [Google Scholar]
- Peplau AL, Campbell SM. The balance of power in dating and marriage. In: Freeman J, editor. Women: A Feminist Perspective. 4. Mountain View, CA: Mayfield Publishing Co; 1989. [Google Scholar]
- Righetti F, Luchies LB, Van Gils S, Slotter EB, Witcher B, Kumashiro M. The prosocial vs. proself power holder: How power influences sacrifice in romantic relationships. Personality and Social Psychology Bulletin. 2015;41:779–790. doi: 10.1177/0146167215579054. [DOI] [PubMed] [Google Scholar]
- Rogers WS, Bidwell J, Wilson L. Perception of and satisfaction with relationship power, sex, and attachment styles: A couple level analysis. Journal of Family Violence. 2005;20:241–251. [Google Scholar]
- Ronfeldt HM, Kimerling R, Arias I. Satisfaction with relationship power and the perpetration of dating violence. Journal of Marriage and Family. 1998;60:70–78. [Google Scholar]
- Rosenberg M. Society and the adolescent self-image. Princeton, NJ: Princeton University Press; 1965. [Google Scholar]
- Rosenthal R, Rosnow RL. Essentials of behavioral research: Methods and data analysis. 3. New York, NY: McGraw-Hill; 2007. [Google Scholar]
- Rusbult CE. A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology. 1983;45:101–117. [Google Scholar]
- Rusbult CE, Arriaga XE, Agnew CR. Interdependence in close relationships. In: Fletcher GJO, Clark MS, editors. Blackwell Handbook of Social Psychology: Interpersonal Processes. 2001. pp. 359–387. [Google Scholar]
- Rusbult CE, Van Lange PAM. Interdependence, interaction and relationships. Annual Review of Psychology. 2003;54:351–75. doi: 10.1146/annurev.psych.54.101601.145059. [DOI] [PubMed] [Google Scholar]
- Rusbult CE, Verette J, Whitney GA, Slovik LF, Lipkus I. Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal of Personality and Social Psychology. 1991;60:53–78. [Google Scholar]
- Sagrestano LM, Heavey CL, Christensen A. Perceived power and physical violence in marital conflict. Journal of Social Issues. 1999;55:65–79. [Google Scholar]
- Shaver PR, Segev MA, Mikulincer MA. A behavioral systems perspective on power and aggression. In: Shaver PR, Mikulincer M, editors. Human aggression and violence: Causes, manifestations, and consequences. Washington, DC: American Psychological Association; 2011. [Google Scholar]
- Sillars AL, Coletti SF, Parry D, Rogers MA. Coding verbal conflict tactics: Nonverbal and perceptual correlates of the “avoidance-distributive-integrative” distinction. Human Communication Research. 1982;9:83–95. [Google Scholar]
- Simpson JA, Farrell A, Oriña MM, Rothman AJ. Power and social influence in relationships. In: Simpson JA, Dovidio JF, editors. The handbook of personality and social psychology: Interpersonal relations and group processes. Washington, DC: American Psychological Association; 2013. [Google Scholar]
- Simpson JA, Rholes WS, Phillips D. Conflict in close relationships: An attachment perspective. Journal of Personality and Social Psychology. 1996;71:899–914. doi: 10.1037//0022-3514.71.5.899. [DOI] [PubMed] [Google Scholar]
- Sprecher S, Felmlee D. The balance of power in romantic heterosexual couples over time from “his” and “her” perspectives. Sex Roles. 1997;37:361–379. [Google Scholar]
- Sprecher S, Schmeeckle M, Felmlee D. The principle of least interest: Inequality in emotional involvement in romantic relationships. Journal of Family Issues. 2006;27:1255–1280. [Google Scholar]
- Straus MA. Measuring Intrafamily conflict and violence: The Conflict Tactics (CT) Scale. Journal of Marriage and Family. 1979;41:75–88. [Google Scholar]
- Straus MA. Future research on gender symmetry in physical assaults on partners. Violence Against Women. 2006;12:1086–1097. doi: 10.1177/1077801206293335. [DOI] [PubMed] [Google Scholar]
- Straus MA, Gelles RJ, editors. Physical violence in American families: Risk factors and adaptations to violence in 8, 145 families. New Brunswick, NJ: Transaction; 1990. [Google Scholar]
- Straus MA, Hamby SL, Boney-McCoy S, Sugarman DB. The Revised Conflict Tactics Scales (CTS2): Development and preliminary psychometric data. Journal of Family Issues. 1996;17:283–316. [Google Scholar]
- Sullivan KT, Pasch LA, Johnson MD, Bradbury TN. Social support, problem solving, and the longitudinal course of newlywed marriage. Journal of Personality and Social Psychology. 2010;98:631–644. doi: 10.1037/a0017578. [DOI] [PubMed] [Google Scholar]
- Thibaut JW, Kelley HH. The social psychology of groups. New York: Wiley; 1959. [Google Scholar]
- Thompson EH, Pleck JH. The structure of male role norms. American Behavioral Scientist. 1986;29:531–543. [Google Scholar]
- Van Willigen M, Drentea P. Benefits of equitable relationships: The impact of sense of fairness, household division of labor, and decision making power on perceived social support. Sex Roles. 2001;44:571–597. [Google Scholar]
- Vandello JA, Bosson JK, Cohen D, Burnaford RM, Weaver JR. Precarious Manhood. Journal of Personality and Social Psychology. 2008;95:1325–1339. doi: 10.1037/a0012453. [DOI] [PubMed] [Google Scholar]
- Vescio TK, Schlenker KA, Lenes JG. Power and sexism. In: Guinote A, Vescio TK, editors. The social psychology of power. New York: Guilford Press; 2010. pp. 363–380. [Google Scholar]
- Vissing YM, Straus MA. Verbal aggression by parents and psychosocial problems of children. Child Abuse and Neglect. 1991;15:223–238. doi: 10.1016/0145-2134(91)90067-n. [DOI] [PubMed] [Google Scholar]
- Waller WW, Hill R. The family, a dynamic interpretation. New York: Dryden Press; 1951. [Google Scholar]
- Weaver KS, Vescio TK. The justification of social inequality in response to masculinity threats. Sex Roles. 2015;72:521–535. [Google Scholar]
- Wieselquist J, Rusbult CE, Foster CA, Agnew CR. Commitment, pro-relationship behavior, and trust in close relationships. Journal of Personality and Social Psychology. 1999;77:942–966. doi: 10.1037//0022-3514.77.5.942. [DOI] [PubMed] [Google Scholar]
- Whitaker MP. Centrality of control-seeking in men’s intimate partner violence perpetration. Prevention Science. 2013;14:513–523. doi: 10.1007/s11121-012-0332-z. [DOI] [PubMed] [Google Scholar]
- Woodin EM. A two-dimensional approach to relationship conflict: Meta-analytic findings. Journal of Family Psychology. 2011;3:325–335. doi: 10.1037/a0023791. [DOI] [PubMed] [Google Scholar]
- Worchel S, Arnold SE, Harrison W. Aggression and power restoration: The effects of identifiability and timing on aggressive behavior. Journal of Experimental Social Psychology. 1978;14:43–52. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.