Romantic relationships are a core feature of healthy young adult development, associated with overall well-being, high self-esteem, social adjustment (Bouchey, 2007; Davila et al., 2016; Kansky, 2018), and lower emotional distress and engagement of risky behaviors (Braithwaite et al., 2010). Evidence suggests that experiences of aggression in early interpersonal relationships, particularly in the family-of-origin, can influence behavioral interactions in subsequent romantic relationships (Cui et al., 2010). Consistent with theories about the development of early relationships, experiences in the family-of-origin are influential on future romantic functioning and are mediated by individual characteristics (e.g., cognitions) carried forward from the family-of-origin (Bryant & Conger, 2002). Prior work has tested this theory by examining the association between positive experiences in the family-of-origin on future romantic relationship functioning and found that positive cognitions (e.g., positive beliefs) partially mediated these relations (Masarik et al., 2013). Thus, identifying how negative experiences in early family relationships (i.e., aggression), specifically aggression perpetrated by a parent against their child or against the child's other parent, are associated with young adults’ cognitions related to current romantic relationships may be useful for identifying possible cognitive targets that are amenable to change. The present study investigates whether experiences of aggression in the family-of-origin impact young adults’ negative expectancies when facing a discussion about desired relationship changes with a romantic partner. We also investigate rejection sensitivity as a mechanism linking aggression exposure in the family-of-origin and negative expectancies about discussing relationship changes, and explore potential gender differences.

Expectancies within Romantic Relationships

Cognitions are known to influence future romantic relationship functioning, including the degree to which partners engage in adaptive coping strategies (e.g., cooperative problem-solving; Assad et al., 2007), and experience relationship satisfaction and commitment (Harel & Koslowsky, 2022; for review see: Lemay & Venaglia, 2016). One important type of relationship-relevant cognition is expectancies, defined broadly in the literature as anticipatory beliefs, which in the context of interpersonal relationships may include predictions about one’s own and one’s partner’s future behaviors and affect, as well as whether one’s interpersonal goals will be attained. Expectancies may vary in their valence, with some focusing on positive beliefs and others negative (Lemay & Venaglia, 2016). Prior work on expectancies has included expectancies of future partner behaviors (e.g, whether a partner will neglect chores or flirt with others; Joel et al., 2023), affect (e.g., partner mood, Schoebi et al., 2012; one’s own mood; McNulty & Karney, 2002), as well as anticipations of future relationship outcomes (e.g., future relationship satisfaction, Baker et al., 2017; and stability, McNulty & Karney, 2004). In the current study, we specifically examine expectancies about the task demands and one’s ability to cope in an upcoming, emotionally challenging couple conversation about desired relationship changes.

In romantic relationships, the valence of expectancies about a specific discussion may influence the behaviors during, and the outcomes of, that interaction (Bradbury & Fincham, 1991; Lemay & Venaglia, 2016). Expectancies likely influence individual and partner behaviors via a combination of reciprocity of anticipated behavior (i.e., self-fulfilling prophecies, Jones, 1986; Sanford, 2003; Schoebi et al., 2012) and temporal projection biases (i.e., continuity of one’s present state, Lemay et al., 2015). Positive expectancies have been linked with greater responsiveness to partner’s needs (Lemay et al., 2015), and increased closeness and positive interactions (Schoebi et al., 2012), whereas negative expectancies have been associated with more negative communication behaviors, such as teasing with hurtful intent, unresponsiveness, critical statements, and anger (Joel et al., 2023; Lemay et al., 2015; Sanford, 2006). Holding negative expectancies may be particularly salient in anticipation of high-conflict or particularly tense discussions, such that they may exacerbate negative individual or partner behaviors during the interaction. Despite robust evidence that expectancies may influence interpersonal interactions and relationship outcomes, little is known about how these expectancies develop (Daspe et al., 2022; Lemay & Venaglia, 2016).

Exposure to Aggression and Negative Expectancies

Expectancies likely develop in response to both state-like (e.g., current emotional state, Lemay et al., 2015) and trait-like (e.g., history of experiences to promote future safety and survival, Bryant & Conger, 2002; Roese & Sherman, 2007) factors. One important but previously untested trait-like influence on negative expectancies of an upcoming couple interaction is history of exposure to aggression in the family-of-origin. Several theories may help to explain the impact of exposure to aggression in the family-of-origin on subsequent expectancies regarding couple interactions. First, theories on intergenerational transmission of aggression posit that parental models of aggression increase the likelihood of similar behaviors being imitated in other relationships (Bandura, 1973). Parental modeling of aggression may also convey messages that aggression is typical, acceptable, and perhaps to be expected in close relationships, shaping expectancies to predict aggression victimization and perpetration in subsequent close relationships (Ehrensaft et al., 2003). Second, trauma-focused theories implicate impaired physiological-regulatory processes as a mechanism by which aggression exposure affects a person’s ability to develop healthy, committed romantic relationships (Courtois & Ford, 2013; van der Kolk, 2014); more frequently experiencing physiological fear cues may shape expectancies towards predicting threat in the interpersonal environment. Third, conflict sensitization theories suggest that exposure to intense interpersonal conflict is associated with increased emotional arousal, distress, avoidance, and schemas regarding future conflicts due to disrupted emotional security (Davies & Cummings, 1994; Davies et al., 1999; Grych & Fincham, 1990). That is, exposure to adverse relationship experiences leads to negative expectancies of emotionally intense discussions in future relationships via shaping cognitive scripts of interpersonal interactions. Although such expectancies may be adaptive in some high conflict situations to promote physical safety, they may be less useful as a long-term relationship strategy and, in fact, may be linked to social maladjustment in adulthood (Davies & Cummings, 1998; Davies et al., 1999).

Of course, not only distal experiences in the family-of-origin but also proximal experiences within romantic relationships are likely to influence relationship-related expectancies (Lemay & Venaglia, 2016). Through repeated exposures to aggressive romantic interactions, individuals may learn that emotionally vulnerable conversations are opportunities for dangerous escalations of conflict, leading to more negative expectancies for potentially conflictual interactions. Indeed, prior work has found that during romantic couples discussions of conflict, prior relationship aggression was associated with high hostility, frightening behaviors, low warmth and poor problem-solving (Gordis et al., 2005). Aggressive couples also tend to display greater reactive anger and escalations of negative communication compared to non-aggressive couples (Burman et al., 1993). With family-of-origin aggression associated with future romantic relationship aggression (Cui et al., 2010; for meta-analytic review see: Smith-Marek et al., 2015), it is difficult to distinguish the effects of growing up in an aggressive family from the associated risks of being victimized by an aggressive partner. Given our research constructs, the question to be explored here is whether aggression exposure in the family-of-origin, above and beyond aggression victimization in the current romantic relationship, specifically affects romantic partners’ expectancies.

Rejection Sensitivity as a Mediator

In line with conflict-sensitization theories, aggression exposure in the family-of-origin may be associated with negative expectancies about future romantic relationships through the development of generalized negative interpersonal schemas. One such schema may include rejection sensitivity–the tendency to “anxiously expect, readily perceive, and intensely react to rejection” across various interpersonal relationships (Downey & Feldman, 1996). Given parent-child relationships are expected to form a sense of safety (Bowlby, 1969) and parent-parent relationships serve as models (Davies & Cummings, 1994), when these relationships are riddled with adverse experiences (e.g., aggression), it may create a negative filter or schema for interpreting interpersonal interactions (Pietrzak et al., 2005). Rejection sensitivity functions as a cognitive-affective processing network (Downey & Feldman, 1996) and is thought to emerge in response to repeated experiences of rejection during childhood (Pietrzak et al., 2005), such as family violence, physical victimization, hostility, emotional neglect, harsh discipline, and conditional parental love (Downey et al., 1997; Downey et al., 1999; Ibrahim et al., 2015).

Romantic relationships during emerging adulthood likewise serve as salient and significant bonds in one’s life. Rejection sensitivity has also been linked to romantic relationship dysfunction, such as decreased relationship satisfaction and closeness, higher conflict, and the use of maladaptive communication behaviors (e.g., reacting with coldness, aggression, hostility, and jealousy; Downey & Feldman, 1996; Meehan et al., 2018; Murphy & Russell, 2018). In addition to examining links between rejection sensitivity and relationship difficulties, more recent work has found that persons who are sensitive to rejection have a cognitive bias to interpret ambiguous situations as negative (Normansell & Wisco, 2017). Further, evidence suggests that rejection sensitivity is linked to schema-congruent processing (Mor & Inbar, 2009). This may indicate that individuals who are sensitive to rejection have a negative schema for anticipating future events, including upcoming conversations with a romantic partner. That is, given the salience of romantic relationships, having a negative schema might be particularly activated, due to fears of losing one’s partner, rejection, or negative beliefs about getting one’s needs met (Mor & Inbar, 2009). In particular, romantic partner discussions about desired relationships changes that invite the potential for conflict or disagreement might particularly activate such negative thoughts or expectancies about such discussions (Downey & Feldman, 1996). Despite documented links between aggression exposure and rejection sensitivity, and also between rejection sensitivity and romantic relationship difficulties and negative cognitions, rejection sensitivity has yet to be tested as a mediator in the link between parent aggression exposure and negative expectancies of future romantic partner interactions.

The Present Study

The present study investigates whether family-of-origin aggression, specifically aggression from parents to the participant or aggression between the participant’s parents, is related to negative expectancies in young adults as they anticipate a laboratory-based discussion about desired relationship change with their romantic partner. We first hypothesize that parents’ aggression will be positively associated with one’s own negative expectancies (HO1). Second, toward the goal of identifying underlying mechanisms between exposure to parent aggression and negative expectancies in the current dating relationship, we test the hypothesis that rejection sensitivity (HO2) mediates the association between exposure to parent aggression and increased negative expectancies.

Finally, we explore the potential moderating role of gender in our pathways between parent aggression, rejection sensitivity, and negative expectancies. To date, the literature provides limited guidance on gender differences in associations between our main study variables, but does show some findings on baserates of the variables (e.g., higher rates of parental aggression for men, Straus & Douglas, 2019; higher rejection sensitivity for women, Maiolatesi et al., 2022; and mixed gender findings for negative expectancies, Daspe et al., 2022; McNulty & Karney, 2004). Additionally, gender differences have emerged in how rejection sensitivity differentially impacts men and women in romantic relationships. For example, Downey and Feldman (1996) found women’s rejection sensitivity associated with hostility and unsupportiveness towards their partner, and men's rejection sensitivity associated with jealous, suspicious, and controlling behaviors. Thus, given the possible relevance of gender in the variables of interest, this study will explore gender differences in our relationships among main study variables, but we do not propose a priori hypotheses.

Method

Participants

Participants for the present study came from a sample of 116 different-sex dating couples, drawn from a more extensive study on young adult dating relationships (Margolin et al., 2022), who completed both a pre-lab and in-lab component. Eligibility criteria included: (a) couples had to be in a dating relationship for at least two months; (b) both partners needed to be age 18 or older and one needed to be age 25 or younger; (c) at least one partner needed to reside in a two-parent family through age 10; and (d) both partners needed to be able to complete all study procedures in English. The analytic sample included only participants for whom we had complete data (n = 224; 114 women and 110 men). Most participants were recruited through online platforms and flyers in the Los Angeles community. Several participants (n = 27 individuals) had previously participated in a longitudinal study assessing family aggression and adolescent development and were recruited along with their romantic partner to participate in the present study. Participants from the longitudinal sample did not differ from newly recruited participants on main variables of interest. Couples were together, on average, for 30.08 months (SD = 23.78) and 43.3% of participants were cohabitating. Average age was 23.15 (SD = 2.87) for men and 22.11 (SD = 1.82) for women. The sample was ethnically/racially diverse (29% non-Hispanic White, 25% Hispanic/Latinx, 15.6% Multiracial, 14.7% African American/Black, 12.1% Asian, and 3.6% other).

Procedure

All study procedures were approved by the University of Southern California Institutional Review Board. Data procedures here included questionnaires from a pre-lab survey and an in-lab study visit. Prior to the in-lab visit, participants completed the online pre-lab Qualtrics survey that assessed: (a) background information; (b) family-of-origin exposure to aggression directly received from parents and witnessed between parents; and (c) rejection sensitivity. During the in-lab visit, participants completed a measure of aggression received from their current dating partner. The 4–5-hour in-lab visit included procedures outside the scope of the research questions here, including several samples of couple communication (for example, see Kazmierski et al., 2020; Margolin et al., 2022). Following the identification and priming of topics for a discussion about desired relationship change, participants rated their expectancies about the upcoming discussion, and then engaged in the 10-minute discussion.

Measures

Negative Expectancies

A 6-item adapted version of the Cognitive Appraisal Measure (Mendes et al., 2007), previously used by Daspe et al. 2022, assessed negative expectancies of the upcoming discussion task. On a 7 point scale from “Strongly Disagree” to “Strongly Agree”, partners indicated their agreement to the following statements: “I’m looking forward to this discussion”, “I’m dreading this discussion”, “I think I’ll do a good job of getting my points across in this discussion”, “I may have a hard time saying what I want to say in this discussion”, “Something good is likely to come out of this discussion”, and “I doubt this discussion will be useful.” To create a score of negative expectancies, we reverse coded the three positive items and averaged across all six items (α = .83).

Parent Aggression

Our measure of parent aggression assessed the following types of aggression: (a) received parent-to-child physical and emotional aggression (14 items) and (b) witnessed parent-to-parent physical and emotional aggression (14 items). The parent-to-child aggression items were adapted from the Parent-Child Conflict Tactics Scales (Straus, 1979; Straus et al., 1998). These items asked whether “at any time in your life, did a parent or stepparent do the following out of anger to you” (e.g., Insulted or shamed you in front of others; hit you with a hand or object). The parent-to-parent aggression items were modified from the Domestic Conflict Inventory (Margolin et al., 1998). These items asked whether “at any time in your life, did one of your parents and/or stepparents do the following to another parent/stepparent or romantic partner out of anger” (e.g., Destroyed possessions or property of parent or parent-like figure; left bruise or visible injury). For each item the respondents endorsed how often the aggression experience had occurred (0 = Never, 1 = Once, 2 = Twice, 3 = 3-5 Times, 4 = > 6 Times). We created a parent aggression score based on the mean across the 28 items (α = .95). In the present sample, 89.7% of participants (87.3% of men and 92.1% of women) endorsed parent aggression exposure.

Rejection Sensitivity

We used 7-items from the 9-item short version of the Adult Rejection Sensitivity Questionnaire (A-RSQ), which is a measure of general rejection sensitivity (Berenson et al., 2009; Downey & Feldman, 1996). Items assess the disposition to anxiously expect a possible rejection after making a request from various individuals (i.e., parent, boss, partner). For each item, the respondent provides two responses: (a) whether they are “anxious about making the request” (1 = “Very Unconcerned”, 6 = “Very Concerned”) to another individual (e.g., parent, friend, partner, and supervisor); and (b) whether they “expect the other person to comply with their request” (1 = “Very Unlikely”, 6 = “Very Likely”). Each item was calculated by multiplying the first item by the reverse of the second item. The resulting 7 scores were averaged to obtain an overall rejection sensitivity score (α = .71).

Covariates

Given the relevance of current dating aggression experiences on conversation expectancies, we controlled for perceived dating aggression with the current partner using 65-items from the How Partners and Friends Treat Each Other scale (Bennett et al., 2011). Partners reported on their dating aggression victimization across items that measure electronic aggressive behaviors (e.g., “Sent a mean, hurtful, or threatening email or text message”), sexual aggressive behaviors (e.g., “Touched me sexually or kissed me when I didn’t want it”), psychological behaviors (e.g., “Ridiculed or made fun of me in front of others”), and physically aggressive acts (e.g., “Pushed, grabbed, shoved, or shook me”). We created a total dating aggression score based on the mean across the 65 items (α = .96). In the present sample, 83.9% of participants (81.8% of men and 86% of women) endorsed dating aggression with their current partner. We also measured age, gender (dummy coded), race/ethnicity (treated as factor), couple cohabitation status (yes vs. no), and relationship length (in months) for potential inclusion as covariates.

Analytic Plan

Descriptive statistics and bivariate correlations were run in IBM SPSS Statistics version 28.0 (IBM Corp, 2021). Inferential statistical tests were run in R version 4.2.3 (R Core Team, 2023) and R Studio (RStudio Team, 2020). We tested our hypotheses using multilevel models with individuals (Level 1) nested within couple dyads (Level 2). Models testing Hypotheses 1 and 2 included dating aggression, age, gender, couple cohabitation status, and relationship length as covariates. Inclusion of the covariate race/ethnicity did not substantially alter our models and was removed for parsimony. While gender, couple cohabitation status, and relationship length were not significant, we kept these covariates due to their theoretical relevance to our constructs. To test HO1, we used a multilevel model with parent aggression as a predictor of negative expectancies. For HO2, we employed a mediation model based on the steps enumerated by Baron and Kenny (1986) within the lme4 package (Bates et al., 2015). First, we ran an equation testing Hypothesis 1, the total effect of parent aggression predicting negative expectancies (path c). Next, to test the indirect effect of parent aggression on expectancies through rejection sensitivity, we first ran an equation with aggression predicting rejection sensitivity (path a). We then modeled an equation in which negative expectancies were simultaneously predicted by both rejection sensitivity (path b) and aggression (path c’). To test Hypothesis 2, the indirect effect (path a*b), we utilized quasi-Bayesian approximation within the mediation package (Tingley et al., 2014). We generated 5,000 simulated samples to test significance and produced approximate confidence intervals for the indirect effect. To interpret the mediation effect size, we calculated the magnitude of the indirect to the total effect. For our exploratory aim investigating gender as a moderator, we added gender as an interaction term on paths a, b, and c in our mediation model.

Results

Descriptive Analyses and Bivariate Correlations

Table 1 presents bivariate correlations for the total analytic sample and descriptive statistics by men and women separately. Supplemental Table 1 provides bivariate correlations for women and men separately. For the total analytic sample, parent aggression was positively associated with dating aggression, rejection sensitivity, and negative expectancies. Dating aggression was positively associated with rejection sensitivity, age, and living together status. Rejection sensitivity was positively associated with negative expectancies and age. Age was positively associated with months together and living together status. For the main study variables of interest, paired sample t-tests matched by couples with complete data (n = 109 couples) show that women, compared to men, endorsed significantly higher rates of parent aggression, t(108) = 2.90, p = .005 and negative expectancies, t(108) = 2.13, p = .035.

Table 1 Descriptive Statistics and Bivariate Correlations Among All Study Variables
Full size table

Hypothesis Testing

Table 2 and Fig. 1 present the model testing rejection sensitivity as a mediator in the association between parent aggression and negative expectancies. In support of HO1, multilevel regression analyses showed that parent aggression was positively associated with negative expectancies for the upcoming relationship change discussion (path c, b = 0.184, SE = 0.092, p = 0.046, CI [0.006, 0.362]). Parent aggression was associated with higher rejection sensitivity (path a, b = 1.147, SE = 0.304, p < 0.001, CI [0.560, 1.745]). Higher rejection sensitivity was associated with higher negative expectancies, after adjusting for parent aggression (path b, b = 0.076, SE = 0.020, p < 0.001, CI [0.037, 0.114]). In support of HO2, parent aggression had a significant indirect effect on negative expectancies through rejection sensitivity (path a*b, b = 0.087, SE = 0.033, p < 0.001, CI [0.032, 0.160]). After adjusting for rejection sensitivity, the direct effect of parent aggression on negative expectancies was no longer was significant (path c’, b = 0.103, SE = 0.091, p = 0.269, CI [-0.078, 0.280]). This indirect effect accounted for 44% of the total effect of parent aggression on negative expectancies (PM = 0.442, p = 0.036, CI [0.108, 2.190]).

Table 2 Direct and Interactive Effects of Parent Aggression or Rejection Sensitivity and Gender on Negative Expectancies and Rejection Sensitivity
Full size table
Fig. 1
figure 1

Mediation model testing rejection sensitivity as a mediator for the association between exposure to parent aggression and negative expectancies. *p<0.05. ***p<0.001

Full size image

In regards to our exploratory analyses, gender did not significantly interact with our model pathways (path c, parent aggression x gender: b = 0.144, SE = 0.186, p = 0.438, CI [-0.214, 0.508]; path a, parent aggression x gender: b = 0.979, SE = 0.622, p = 0.117, CI [-0.223, 2.183]; path b, rejection sensitivity x gender: b = -0.028, SE = 0.037, p = 0.460, CI [-0.099, 0.044]). Given gender did not moderate any paths, we did not test further for moderated mediated effects.

Discussion

The overarching aim of this investigation was to examine the association between exposure to parent aggression and negative expectancies and to ascertain if rejection sensitivity is one potential mechanism of this process. Consistent with Hypothesis 1 of our study, results demonstrate that exposure to parent aggression was related to higher negative expectancies prior to engaging in a change discussion with a romantic partner, even after accounting for current dating aggression in the relationship. As anticipated, we found rejection sensitivity to be an underlying mechanism that explained the association between parent aggression and negative expectancies. Gender did not moderate these associations.

Romantic couples are often faced with navigating important, and at times challenging, discussions about their relationship. Our findings here reveal that romantic couples may enter conversations with preconceived notions with how these interactions will go. Although prior work has assessed expectancies of affect, behaviors, and relationship functioning (Lemay & Venaglia, 2016), our study goes beyond this work to assess expectancies of the task demands and one’s ability to cope prior to an upcoming change discussion with a romantic partner. As evidenced by previous work, couple conflict is highly influenced by the cognitions (e.g., expectancies) romantic partners bring into a conversation (Epstein et al., 1993). Further, negative expectancies have the potential to exacerbate negative communication behaviors and impair relationship functioning if partners do not have the skills necessary to navigate such conflict (Epstein et al., 1993). Thus, understanding the cognitions that couples bring into a discussion on relationship change and how these cognitions are shaped by prior aggressive experiences have the potential to be useful for preventing future conflict escalations.

Our data show that interactions in one generation can spillover into cognitions (i.e., negative expectancies) about future relationships with rejection sensitivity as an underlying mechanism. These findings align well to past work demonstrating rejection sensitivity as a mediator of the relationship between early adverse experiences and poor outcomes (e.g., such as avoidance-attachment behaviors; Feldman & Downey, 1994). Our findings extend this work to include cognitive outcomes (i.e., negative expectancies). Additionally, our work supports recent findings regarding rejection sensitivity being linked to negative cognitive processing (Normansell & Wisco, 2017) and reveals that individuals who are highly sensitive to rejection may possess negative schemas specific to future interactions with intimate partners.

We did not find gender differences in any of the three pathways tested here although we acknowledge that gender could emerge as a moderator with a larger sample size. Overall, women endorsed higher rates of negative expectancies compared to men, although in a study (Daspe et al., 2022) using the same measure with this sample prior to a non-conflictual discussion about planning a date, men reported more negative expectancies. We also found that women report more parental aggression, which supports some findings from a multi-national study indicating that women experience higher rates of physical aggression by mothers, but not other findings from this study showing that men experience more physical aggression by fathers or both parents (Straus & Douglas, 2019). As contrasted with direct experiences of parental aggression, our measure also includes the witnessing of parent-to-parent aggression. Additionally, our measure of parent aggression included items beyond physical aggression, such as psychological (e.g., “Insulted or shamed [you or parent/stepparent] in front of others”) as well as electronic aggressive items (e.g., “Sent [you or parent/stepparent] an insulting or threatening text, email, Tweet, etc.) that were not typically included in earlier assessments of aggression. Moreover, despite these gender differences in mean scores, associations between variables of interest did not vary by gender.

Limitations and Future Directions

This study contains several limitations that deserve note. First, the self-reports of exposure to parent aggression are retrospective, which may be distorted by memory recall. Second, the cross-sectional nature of our design precludes making any causal conclusions. While conceptually our variables represent different time points, additional research with a longitudinal design is warranted, that is, examining experiences of parent aggression during childhood and rejection sensitivity prior to romantic relationship initiation. Third and relatedly, our parent aggression measurement does not examine the timing (early childhood versus adolescence) or differentiate the type of aggressive experiences. Differentiating type and timing of aggression exposure may affect both rejection sensitivity and negative expectancies. Fourth, in using these items from a standard measure of rejection sensitivity, we assessed this construct in the context of various interpersonal relationships and a wide range of requests and circumstances rather than with specific reference to romantic couple interactions; more specificity to the ongoing relationship might hold more saliency for an upcoming discussion with that partner. Additionally, more recent work has adapted the rejection sensitivity measure to consider other emotional and behavioral reactions to a perceived or expected rejection, such as anger, hostility, and reactive aggression (Romero-Canyas et al., 2010; Zimmer-Gembeck & Nesdale, 2013), which also could be relevant to negative expectancies or future aggression perpetration. Fifth, the upcoming change discussion involved changes desired by both partners. Thus, negative expectancies could reflect concern about bringing up one’s own topics or topics that the partner might initiate. Sixth, although addressing our research questions with young adults offers insights into dating relationships, these results cannot be generalized to other developmental periods, to more long-term or continued relationships, or to gender- and sex-diverse couples, which is an important consideration for future investigations.

Implications

Despite these limitations, our study offers some insights into how experiences of aggression influence young adults’ expectancies of conversations about desired relationship change with their romantic partner and the role of rejection sensitivity in this link between family-of-origin aggression and expectancies. Prior work has documented that negative expectancies, generally, are associated with negative behaviors in romantic relationships (Joel et al., 2023). Additionally, evidence suggests that rejection sensitivity is linked to future aggressive victimization and perpetration (Gao et al., 2021). Thus, negative cognitions, such rejection sensitivity and negative expectancies may be key in determining the behavioral strategies used by romantic couples during difficult or challenging discussions. Further investigations would benefit from examining whether negative expectancies about the task demands and one’s ability to cope in an upcoming partner interaction translate to the use of one’s own and partner aggressive behaviors during the discussions themselves. Additionally, these negative expectancies may impact whether an individual is willing even to engage in such challenging discussions with their romantic dating partner. Although in some cases it may be understandable and normal to dread a difficult conversation with a partner, holding negative cognitions can become maladaptive if it results in avoiding important discussions. Additional research is needed to assess whether, indeed, rejection sensitivity and negative expectancies about the task demands and one’s ability to cope in an upcoming partner interaction, in combination, are associated with difficulty navigating challenging discussions with a partner or circumventing such discussions altogether.

Emerging work suggests that rejection sensitivity is a modifiable construct and that therapy techniques such as mindfulness (Peters et al., 2016) and interpersonal psychotherapy (Mellin, 2008) can be implemented to improve an individual’s sense of self-efficacy in navigating relationships with others. More specific to couple interventions, our study has identified two related cognitions–rejection sensitivity and negative expectancies–that may be potential foci in the context of couple and family therapy. In addition, in couple therapy, it may be important to discuss the legacy of aggression in the family-of-origin with respect to these cognitions and in combination with communication training.

Conclusion

This study investigates exposure to parent aggression as a risk factor for later negative expectancies in young adult dating couples. Our findings offer evidence that early adverse relationship experiences in the family-of-origin play a role in how young adults think about upcoming conversations with their romantic partner. Understanding these cognitions and how they develop may help inform researchers and clinicians as they support young adults to develop and maintain healthy romantic relationships.