Abstract
We were interested in how people in a romantic relationship would perceive the intelligence of their partners who have high or low trait anger. Specifically, we referred to the tension between compassion (low anger) and competence (high intelligence) in mate choice. Some evolutionary theories suggest that mating might be considered a bargaining process between these two higher-order attributes. Our study involved 148 heterosexual couples in romantic relationships. We measured the relationship between relationship satisfaction, trait anger, objective intelligence, self-assessed intelligence, and subjectively assessed partners' intelligence. We found that angrier men were less satisfied in their romantic relationship than those men who were less angry, and their partners were also less satisfied in the relationship. Additionally, women perceived angrier men as less intelligent, an effect that remained after controlling for men's objective intelligence. Lastly, we found that women's perception of their partner's intelligence mediated the link between men's anger and relationship satisfaction for both sexes. Our findings suggest that both anger and intelligence play important roles in romantic relationship functioning, consistent with evolutionary theories that emphasize the value of competence (i.e., intelligence) and compassion (i.e., low anger) in romantic partners. Furthermore, our study highlights the importance of women's perception of their partner's intelligence in determining the quality of the relationship.
Statement of Relevance
Traits such as kindness, anger, or intelligence, are considered crucial when finding a romantic partner. These traits may also influence relationship satisfaction. In the current study, we found that anger and the perception of partners' intelligence affected satisfaction. Angrier men were less satisfied in their romantic relationship than those men who were less angry, and their partners were also less satisfied in the relationship. Additionally, women perceived angrier men as less intelligent, which impacted relationship satisfaction of both partners. Our findings highlight the importance of traits from the area of competence (intelligence) and kindness (anger) for relationship satisfaction.
The emotion of anger serves many social functions as it regulates interpersonal interactions and helps to challenge goal blockages (Harmon-Jones & Harmon-Jones, 2016). Anger is one of the most intense emotions that can be experienced in a romantic relationship (Fischer & Roseman, 2007; Lazarus, 1991). Both state and trait anger are associated with diminished relationship quality (Mackenzie et al., 2014; Renshaw et al., 2010). This could arise from the broader social consequences of anger. For example, anger increases the likelihood of aggressive behavior and violence (Berkowitz, 1993). It is associated with disagreeableness (Wilkowski & Robinson, 2010), more marital conflict, and less marital satisfaction (Ali & Naylor, 2013; Baron et al., 2007; Mackenzie et al., 2014; Smith et al., 1990, 2010). Anger is a precursor for relationship problems and even breakups (Gottman, 1993; Roberts, 2000).
Anger is often considered a “hot” process as opposed to “cool” processes, such as analytical thinking, so much so that it is considered a destructive force and not easily tempered (Wilkowski & Robinson, 2010). Correspondingly, anger has an adverse effect on cognitive functioning. For instance, angrier people have narrower memory capacity (Threadgill & Gable, 2019) and selective attention toward anger-related semantic cues (Parrott et al., 2005; Wilkowski & Robinson, 2010). Additionally, angrier people not only report more effortful control (Rothbart et al., 2000) and less dispositional self-control (Tangney et al., 2004) but also display less cognitive inhibition and worse fluid intelligence (Zajenkowski & Zajenkowska, 2015). Lastly, anger leads to the underestimation of risk (Lerner & Keltner, 2001) and the overestimation of one's abilities (Zajenkowski & Gignac, 2018). Because anger is associated with impulsive rather than reflective processes, it is more likely for people to associate anger with being “hot-headed” rather than “cool-headed.”
By contrast, intelligence is defined as a general mental capability that involves the ability to reason, plan, solve problems, think abstractly, and broader capability for comprehending our surroundings (Gottfredson, 1997). Furthermore, intelligence is an important factor determining many real-life outcomes, such as educational achievement, occupational, and economic success as virtually all activities require some reasoning (Gottfredson, 1997). Thus, intelligence might be more likely associated with being “cool-headed.”
In the current study, we were interested in how people in a romantic relationship would perceive the intelligence of their partners who have high or low trait anger. Specifically, we referred to the tension between compassion (e.g., low anger) and competence (e.g., high intelligence) in mate choice (Jonason & Antoon, 2019). Theoretically, mating might be considered as a process of bargaining between two higher-order attributes: competence and compassion of partners (March & Jonason, 2023). Competence as a myriad of desired attributes such as education, level of income, and intelligence represents an ability to invest in a relationship, to accrue necessary resources for survival (Jonason & Thomas, 2022). On the other hand, compassion, represented by kindness and agreeableness (or low anger), reflects the potential partners' willingness to share owned resources, to build favorable and peace space in romantic relationship (March & Jonason, 2023). Widely understood competence and compassion might be regarded as an internally balanced construct of partners' value. The role of this value in shaping mating preferences has received considerable attention (March & Jonason, 2023). However, less is known how competence and compassion operate within relationships and how these qualities are linked to relationship outcomes.
People have various preferences in who they would like to pair-up with romantically. However, both men and women most value compassion in their partners. Specifically, they prefer partners who are kind and understanding is (Buss & Barnes, 1986; Gignac et al., 2018). The description of this attribute suggests that it might be considered as an opposition to anger (Buss & Barnes, 1986). On the other hand, intelligence, associated with competence, is ranked as second (Gignac et al., 2018) or third (after exciting personality; Buss & Barnes, 1986) desired characteristics of a romantic partner. For women, intelligence is an especially attractive attribute of men (Gignac et al., 2018; Jonason & Thomas, 2022). Additionally, women value, more than men, potential partners who have good earning capacity and graduated college, which can be considered as a derivative of intelligence (Jonason et al., 2012, 2019). The importance of intelligence in human mating is further supported by research suggesting strong evidence for assortative mating for intelligence (r ≈ .40; Escorial & Martín-Buro, 2012; Gignac & Zajenkowski, 2019) and is greater for intelligence than for other behavioral (≈.10) or even physical characteristics (e.g., height; ≈.20; Plomin & Deary, 2015). However, the process behind that mating is not exactly known yet. It has been shown that intelligence is sexually attractive on its own (Gignac et al., 2018; Jonason et al., 2019) However, intelligence, as a representation of valuable heritable attributes, offers unarguably evolutionary advantages (Barkow, 1989; Miller, 2000) and social ones (Jonason et al., 2019).
The Current Study
In the current study, in a sample of heterosexual couples, we investigated the association between trait anger, objectively measured intelligence, as well as subjectively assessed intelligence—evaluated by both the participants themselves and their romantic partners. Additionally, we examined the relationship satisfaction of both partners. First, considering the findings showing aversive consequences of anger for relationships, we expected that anger would predict less relationship satisfaction in both partners (H1), as anger might be considered the opposite of kindness/agreeableness (i.e., low compassion; Jonason & March, 2023). Furthermore, because of the bargain between compassion and competence, people with higher levels of anger (low compassion) might be perceived by their partners as less intelligent (low competence) even if their intelligence is objectively not that low. Thus, we expected that angrier men will be perceived by their partners as less intelligent (H2). The question remains as to what extent this relationship holds when objective intelligence is controlled. Objective intelligence must be controlled, to maintain that assessments of partners' subjective intelligence are not biased by their objective intelligence. Lastly, we expected that the bargain between competence and compassion might influence the relationship quality. We hypothesized that partners' views of intelligence explains (mediate) the relationship between anger and relationship satisfaction. Because of the partners' lower estimation of intelligence, more anger could be associated with less relationship satisfaction (H3).
Method
Participants
The study involved 148 romantic heterosexual couples. Women were aged from 18 to 74 (M = 27.63, SD = 10.22, 1 missing case) and men were aged from 18 to 80 (M = 29.30, SD = 10.76, 2 missing cases). This sample size allows us to detect a modest correlation (r = .23; Gignac & Szodorai, 2016) with α of 0.05 and power of 0.80 (power analysis based on the R package “pwr”).
Measures
Trait anger was measured by the Polish adaptation (Bąk, 2016) of the 10-item trait anger subscale from the State-Trait Anger Expression Inventory-2 (Spielberger, 1999). Participants report how frequently they experience angry feelings over time (1 = almost never; 5 = almost always) with items such as “I am a hothead person.” A reanalysis of the scale's internal consistency was confirmed for this data separately for men (Cronbach's α = .84) and women (α = .86). A trait anger index was calculated from the mean.
Objectively assessed intelligence was measured as fluid intelligence with the Raven Advanced Progressive Matrices (Raven et al., 1983). The test contains 36 items that include a three-by-three matrix of figural patterns with a missing bottom-right pattern. The goal is to discover the rule between patterns in rows and columns and choose the proper missing pattern from eight provided options. The chosen pattern should match the rule in each matrix. Sum of correct answers is the intelligence score. A reanalysis of the scale's internal consistency was confirmed for this data separately for men (α = .90) and for women (α = .88). We chose Raven's test as it is considered one of the most substantial individual indicators of general intelligence (Gignac, 2015).
Subjectively assessed intelligence (Zajenkowski et al., 2016) was measured by asking participants to assess their and their partners' intelligence. Participants were informed, that “People differ with respect to their intelligence and can have a low, average, or high level.” and were asked to indicate where their and their partner's intelligence can be placed compared to other people on a 1- to 25-point rating scale.
Relationship satisfaction was measured by the 10-item Relationship Assessment Scale (Hendrick, 1998). Participants were asked to what extend (1 = low to 5 = high) they agree with the question such as “How well does your partner meet your needs?.” A reanalysis of the scale's internal consistency was confirmed for this data separately for men (Cronbach's α = 0.74) and for women (Cronbach's α = 0.76).
Procedure
Participants were recruited through popular social media platforms (e.g., Facebook) by posting announcements in various discussion groups. Participants gave their informed consent and were presented with the general aim of the study. Each partner was tested individually during an in-person meeting with the researcher. The partners were studied independently to ensure they did not influence each other's results. Each session lasted approximately 45 min. Participants were asked about the duration of the current relationship, the number of “breakups” in this relationship, and the number of previous relationships. The current study is a part of a larger project (Zajenkowski & Gignac, 2021). The study was approved by an institutional ethics committee and conducted from January to December 2019.
Data Analysis Strategy
To test if anger would predict less relationship satisfaction in both partners (Hypothesis 1) and if angrier men would be perceived by their partners as less intelligent (Hypothesis 2), we ran a correlation analysis between all variables of interest. In this part, we checked the relationship between each participant's romantic relationship satisfaction and their partner's trait of anger. We also analyzed the relationship between participants’ estimation of their partners’ intelligence and their partners’ trait anger.
In the next step, we conducted a regression analysis to examine if anger predicts how women estimate their partners’ intelligence while controlling for the objective intelligence of their partners. In the regression analysis, we included objective intelligence as a predictor of the estimation of their partner's intelligence. Then, we added partners’ anger to the model to check if the explained variance of the subjective estimation of their partners’ intelligence significantly increased.
To test Hypothesis 3, we checked if partners’ views of intelligence explain the relationship between anger and relationship satisfaction, we ran two mediation analyses using the PROCESS macro for SPSS (Hayes, 2015). We tested two models examining how women's perception of their partners’ intelligence may account for lower relationship satisfaction in both men and women. In both models, we used women's estimation of their partner's intelligence as an indirect effect accounting for the relation between men's trait anger and either men's relationship satisfaction or women's relationship satisfaction.
Results
Table 1 contains descriptive statistics for all measured variables, while Table 2 presents descriptive statistics in the sex division and sex differences in each variable. Table 3 contains the correlations in men and women. Men reported significantly lower levels of trait anger and higher subjective intelligence than women. The most consistent findings concerned men's trait anger and our hypotheses were supported only in men. 1 First, it was negatively associated with relationship satisfaction of both sexes, supporting H1. Second, in line with H2, women's estimation of their partners' intelligence was negatively correlated with their partners' trait anger. Men's anger was also positively correlated with the number of breakups.
Descriptive Statistics of All Measured Variables.
Note. Duration of current relationships was measured in months;
Descriptive Statistics of All Variables for Men and Women with T-Test Results Comparing Their Scores and Effect Size Estimations.
Note. Duration of current relationships was measured in months; * p < .05.
Correlations of All Variables in the Women Group, Men Group and Between the Women and the Men Group.
*p < .05; **p < .01.
Next, we conducted two regression analyses (Table 4) with women's estimations of partner's intelligence as an outcome variable and two predictors: men's objectively measured intelligence (Step 1) and men's anger (Step 2). The model with men's objectively measured intelligence was significant (F[1, 146] = 12.91, p < .001) and explained 8% of women's estimation of their partner's intelligence. The second model was also significant (F[2, 145] = 8.96, p < .001); men's anger accounted for an additional 2% of the variance in women's estimation of partners' intelligence. Thus, the association between men's anger and their partners' estimation of intelligence remained significant even after controlling for objective intelligence.
Regression Models with Men's Objective Intelligence and Trait Anger as Predictors of Women's Partners Intelligence Estimation.
Steiger's Z Test Comparing Correlations of Variables Between Woman and Men Group (Lower Triangular) and Comparing Correlations of Variables Between Men × Women and Women × Men.
Lastly, we examined to what extent the women's perception of their partners' intelligence may account for the lower men's and women's relationship satisfaction. We tested two models with men's anger as a predictor, men's and women's relationship satisfaction as an outcome and women's estimation of partners' intelligence as a mediator. Then we tested the indirect effects. In the first model (see Figure 1) explaining women's relationship satisfaction, the indirect effect through women's estimation of partner's intelligence was significant (β = −.06, p < .001, 95% CI [−0.13, −0.01]), while the direct effect of men's anger was not significant (β = −.09, p = .095, 95% CI [−0.20, 0.02]). In the second model (Figure 2) with men's relationship satisfaction as the outcome, the indirect effect through women's estimation of their partner's intelligence was significant (β = −.03, p = .006, Cl 95% [−0.07, −0.00]), and the direct effect of men's anger remained also significant (β = −.11, p = .047, 95% CI [−0.22, −0.01]). Thus, women's perception of their partners' intelligence accounts for the link between men's anger and relationship satisfaction in both sexes. This result supports H3, however, only with respect to men's anger.

Model linking men's trait anger, women's relationship satisfaction, and women's estimation of partners' IQ.

Model linking men's trait anger, men's relationship satisfaction, and women's estimation of partners' IQ.
Discussion
We examined the association between trait anger, self-perceived intelligence, objectively measured intelligence, and relationship satisfaction within heterosexual couples. We suggested that anger and intelligence might be indicators of broader constructs of compassion and competence, important for relationship functioning (March & Jonason, 2023). We hypothesized that trait anger would be correlated with partners' intelligence assessment and relationship satisfaction. The results were consistent with these expectations, however, significant effects of anger were observed only in men. Although we found that men reported lower trait anger than women, the meta-analytic review of literature about anger covering self-reports, observations, peer reports, and teacher reports (Archer, 2004) suggest rather that there are no sex differences in anger. Additionally, that men's anger might be more destructive and lead to aggressive behavior (Archer, 2004; Bettencourt & Miller, 1996). Other studies revealed that the aggressiveness of men might be driven by their increased revenge motivation (Wilkowski et al., 2012). Additionally, men are less likely to forgive and have suboptimal communication skills in a relationship (Fincham & Beach, 2002). These aggressive tendencies may lead to further relationship difficulties and breakups (Gottman, 1993; Roberts, 2000), which is consistent with our results indicating that men's anger was positively correlated with the number of breakups in the relationship.
Second, men's anger was negatively correlated with the assessment of their intelligence made by their partners. Anger is regarded as a “hot” attribute that leads to lower cognitive ability (Wilkowski & Robinson, 2010). Consequently, those high in trait anger might be perceived as less intelligent by their partners because intelligence is considered a “cool” attribute (Jonason & Hughes, 2021). Additionally, our findings fit evolutionary theory, whereby people equally value competence and compassion in their partners (Buss & Barnes, 1986; Jonason & Antoon, 2019). Thus, low compassion (i.e., high anger) is expected to go along with low competence (i.e., high intelligence). Interestingly, in our study, this pattern was observed for men's anger and women's perception of their partners. This is consistent with the findings showing that women were less interested in partners who were disagreeable and less educated than men (Jonason & Antoon, 2019). High levels of compassion and competence might be especially important for women as they incur much higher costs while mating in situations of choosing the wrong partner (Haselton & Buss, 2000; Koehn & Jonason, 2021; Jonason et al., 2015). Women are at a greater risk not only as a victim of date rape, sexual assault, and sexual aggression (Rickert & Wiemann, 1998) but also of difficulties in finding a partner ready to parent another man's child (Buss, 2016), possible fertility problems caused by sexually transmitted infections (Bowden et al., 2002), or as a single parent exposed to physical illness and mental health difficulties in challenging alone parenting (Stack & Meredith, 2018). Finally, compassion might be an indicator of good parenting in a relationship (Farrelly, 2013) and the willingness to share resources (Phillips et al., 2008; Van Vugt & Iredale, 2013). In contrast, men not only benefit more in short-term mating but also pay fewer costs than women do after choosing possibly the wrong partner (Schmitt et al., 2001).
Lastly, our findings shed light on the association between a partner's anger and relationship satisfaction. Specifically, we found that men's anger was negatively correlated not only with women's relationship satisfaction but also with their own relationship satisfaction. This is consistent with previous findings showing that anger leads to relationship difficulties and reduces relationship satisfaction (Ali & Naylor, 2013; Baron et al., 2007; Mackenzie et al., 2014; Smith et al., 1990, 2010). Interestingly, we found that the association between anger and satisfaction (of both men and women) was mediated by how women estimated their partners’ intelligence. Perceived intelligence might be especially important for men as previous findings indicate that men tend to rate their intelligence higher than women do (Szymanowicz & Furnham, 2011). Moreover, studies examining estimations from individuals' family members support this notion. Typically, male family members (such as grandfathers, fathers, and brothers) are perceived to possess higher general intelligence compared to their female counterparts and parents commonly rate their sons' IQ as higher than that of their daughters (Furnham, 2001). This effect is known as “male hubris, female humility” and suggests that perceptions of intelligence are influenced by gender norms (Furnham, 2001). Thus, men who are perceived as less intelligent may experience lower levels of self-worth, which, in turn, leads to reduced relationship satisfaction.
Limitations and Conclusion
Despite yielding novel findings about the role of compassion and competence in romantic relationships, our study was not without limitations. First, our study was correlational and tested only partners in existing relationships. Future investigations might explore the relationship between anger and perceived intelligence in mating (e.g., among people seeking for a partner). Additionally, anger was measured using only a self-report questionnaire. It would be worth measuring behavioral indicators of anger in future studies.
Second, we measured only one indicator of compassion—anger. However, compassion is regarded as a complex trait that can be also operationalized by various attributes. Future research could explore how other aspects of compassion are associated with intelligence and, further, with romantic relationship satisfaction. Additionally, emotional intelligence seems to be a construct that involves both compassion and competence. Previous research indicates that emotional intelligence is associated with higher general intelligence (Schulte et al., 2004) and lower anger/aggression (García-Sancho et al., 2014). Thus, future studies might examine its role in the context of intelligence/anger among couples in romantic relationships.
Additionally, future studies might consider the antecedents of anger as our study indicated that anger had negative effects on relationship functioning. For instance, there is evidence that anger is negatively associated with socioeconomic status, both when estimated using objective measures (Rahman et al., 2014) and when subjectively assessed by participants (Greitemeyer & Sagioglou, 2018) It has been suggested that low SES leads to higher anger and aggressiveness, which, in turn, may increase social rejection (e.g., by a romantic partner). Lastly, we did not ask participants whether they had children. Having children might be important for the compassion/competence association.
Our study supports the notion that compassion and competence complement each other and are valued attributes of partners in a romantic relationship. Specifically, we found that men's anger was negatively associated with the way the partners' perceived their intelligence, and that these traits collectively explained relationship satisfaction in romantic relationships. These findings have implications for understanding the complex interplay between personality traits and relationship outcomes.
Footnotes
Declaration of Conflicting Interests
The author(s) declared that there were no conflicts of interest with respect to the authorship or the publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
