Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Couple Communication, Conflict, and Repair

Abstract. Couple communication, conflict, and repair constitute an interdependent sequence of relationship processes through which partners interpret disagreements, influence one another, regulate affect, negotiate consequential change, and restore coordination following disruption. This critical narrative research monograph integrates the history of behavioral marital research, attachment and attribution perspectives, observational interaction coding, physiological and diary methods, dyadic longitudinal modeling, meta-analytic evidence, and trials of couple interventions. It examines demand–withdraw dynamics, conflict escalation, contextual benefits and costs of directness, post-conflict recovery, the distinction between reciprocal conflict and coercive abuse, and the limits of prediction from single observed conversations. Strongest evidence supports modest prospective associations between communication and later relationship functioning, heterogeneous effects under socioeconomic constraint, and a distinct temporal role for repair following disagreements. The synthesis specifies experimental and measurement limitations, separates theory from demonstrated mechanisms, and develops falsifiable research priorities for human relationships and technologically mediated interaction. This is a source-identified critical narrative synthesis, not an original experiment or registered systematic review.

Keywords: couple communication; romantic relationship conflict; conflict repair; demand–withdraw; marital interaction; conflict escalation; attributions in marriage; post-conflict recovery; emotion regulation; dyadic methods; relationship satisfaction; social and economic constraints; couple therapy; longitudinal relationship research; human–artificial mediation

The Scientific Object of Couple Communication, Conflict, and Repair

Couple communication comprises the observable and interpreted exchanges through which partners convey information, register concerns, regulate accessibility, negotiate decisions, and respond to one another over time. Conflict is a sequence of incompatible goals, expectations, perceptions, or actions that becomes relevant within a continuing relationship. Repair refers to situated attempts and achieved processes that restore workable coordination, understanding, or emotional connection after a disruption. These constructs overlap without being interchangeable: a relationship can contain disagreement alongside mutual respect, and an apology can occur without restoring trust. The principal research question concerns how sequences of interaction combine with prior expectations and external constraints to shape subsequent relational functioning Finkel et al. (2017) Karney & Bradbury (1995).

The unit of analysis varies with the question. A single utterance may express a complaint; a turn-by-turn sequence can show whether a partner validates, contradicts, withdraws, or redirects; repeated encounters can establish habitual cycles; and longitudinal trajectories reveal changes in satisfaction, safety, commitment, and relationship stability. A scientific account therefore separates speech content from vocal affect, nonverbal behavior, physiological response, subjective meaning, and the social context in which a dispute occurs. Coders can observe interruption and withdrawal but usually cannot infer hostile motive or felt vulnerability without additional evidence. A repair attempt must be distinguished from a repair outcome, because acceptance, timing, and later behavior determine whether an initiative becomes effective Gottman (1993) Overall & McNulty (2017).

This monograph conducts a critical narrative synthesis of primary interaction studies, prospective dyadic research, experimental emotion-regulation paradigms, longitudinal reviews, quantitative syntheses, and evidence concerning couple treatment. Its argument proceeds from conceptual and historical foundations to conflict organization, affect regulation, repair, variation across contexts, methodological inference, and ethical boundaries. It evaluates observational associations against experimental manipulation and prospective evidence rather than assigning identical evidentiary status to every finding. No original couples were recruited, no new effects were estimated, and no systematic-review search protocol or pooled meta-analysis was performed. The synthesis treats its explanatory integration as a map for future testing, grounded in identified research programs Kanter et al. (2022) Johnson et al. (2022) Finkel et al. (2017).

From Marital Interaction Research to Contemporary Relationship Science

Early laboratory studies of marital behavior brought a methodological change to the study of intimate relationships: researchers recorded couples discussing consequential disagreements and quantified temporal patterns rather than relying exclusively on retrospective global satisfaction reports. Levenson and Gottman's investigation of physiological linkage and affective exchange established an influential research program joining behavioral observation with autonomic measurements during interaction. Subsequent longitudinal studies asked whether features of exchanges preceded marital instability. Such designs made interaction an empirically inspectable process while raising questions about laboratory representativeness, selection effects, and the extent to which observed correlations support prediction for new populations Levenson & Gottman (1983) Gottman & Levenson (1992).

The 1990s consolidated rival explanatory accounts. Bradbury and Fincham reviewed how partners' causal and responsibility attributions could organize responses to negative events, while Christensen and Heavey demonstrated that demand–withdraw behavior depended on who requested change and how conflict was socially structured. Karney and Bradbury then integrated vulnerabilities, stress, and adaptive processes into a developmental account of marriage. These traditions located communication within the histories and circumstances of two interdependent people. A partner's ostensibly negative move might reflect distress, limited resources, anticipated rejection, issue ownership, or an immediate effort to end escalating conflict Bradbury & Fincham (1990) Christensen & Heavey (1990) Karney & Bradbury (1995).

Later research widened its methodological and demographic scope through prospective datasets, behavioral coding, daily diaries, partner-specific analyses, and more explicit scrutiny of generalizability. Noller and colleagues followed 33 couples from premarital assessment into the first two years of marriage, observing links running in both directions between communication and satisfaction; subsequent work expanded these designs Noller et al. (1994). Quantitative reviews now permit assessment of the average association between communication and later satisfaction rather than reliance on memorable individual laboratory studies. Research on socioeconomic disadvantage has additionally challenged the assumption that identical verbal strategies serve every couple equally. The development of couple science therefore reflects a continuing movement from description of specific conversational acts toward explanations that include timing, motivations, structural constraints, and changing social environments Kanter et al. (2022) Ross et al. (2019) Overall & McNulty (2017).

Definitions, Operational Boundaries, and Construct Distinctions

Communication quality is typically operationalized through coded positive and negative behavior, self-reported experiences of responsiveness, verbal support, hostility, engagement, or partner withdrawal. Conflict frequency measures how often disputes arise, whereas conflict severity concerns the magnitude or consequences of incompatible positions; neither variable by itself captures how partners proceed after disagreement. Relationship satisfaction is a person's evaluation of a partnership, and stability refers to its persistence or dissolution. These outcomes need separate measurement because distressed relationships sometimes persist, and relationships can end despite periods of effective communication. The validity of a claim depends on whether its chosen indicators represent the specified construct Karney & Bradbury (1995) Kanter et al. (2022).

Repair has at least three distinct operational meanings. Initiation denotes a behavior intended to reduce interpersonal trouble, such as acknowledging harm, expressing responsibility, clarifying intent, or inviting a renewed conversation. Reception concerns whether the partner perceives the initiative as responsive, sincere, safe, and relevant. Recovery refers to a subsequent change in interactional behavior, affect, or relational evaluation across minutes, days, or longer periods. Studies of conflict recovery and reconciliation show why these phases warrant independent observation: the presence of conciliatory words alone cannot establish restoration of emotional closeness Salvatore et al. (2011) Wilder et al. (2024).

An essential boundary distinguishes reciprocal disagreements from patterns of intimidation, threats, control, or violence. Ordinary conflict research often assumes both partners can voice interests and negotiate consequences. Coercive domination changes the meaning of avoidance, compliance, silence, and apparent agreement because a partner's behavior may be organized around risk and protection. Ethical and scientific evaluation must therefore assess safety and power before interpreting a couple's exchange as symmetric communication failure. Population research on intimate partner violence and the broader literature on contextual differences indicate why a single universal conflict score can conceal substantively different processes World Health Organization (2026) Ross et al. (2019).

Theoretical Accounts of Relational Interaction

The vulnerability–stress–adaptation model understands marital development through interactions among enduring personal characteristics, stressful conditions, and adaptive processes. Communication is one potential adaptive process through which stress and vulnerabilities become consequential, although it is also shaped by them. This bidirectional formulation prevents the inference that an observed argument is necessarily the original source of later dissatisfaction. For example, economic pressure may intensify conflict over household responsibilities while also constraining the time and security available for repair. Prospective studies must measure stress exposures and vulnerabilities to distinguish their effects from within-couple interaction Karney & Bradbury (1995) Ross et al. (2019).

Attributional accounts examine how a partner explains an unwelcome action. A delayed response might be interpreted as temporary distraction, enduring indifference, purposeful disrespect, or an inevitable consequence of external demands. These interpretations differ in causal stability, locus, intentionality, and blame. Attribution theory predicts that construing negative behavior as deliberate and enduring may intensify defensive or retaliatory responses, yet these cognitions can also represent accurate recognition of persistent harmful conduct. The empirical problem is to measure interpretations independently of both preceding events and later behavior, rather than circularly describing conflict as the product of 'negative thinking' Bradbury & Fincham (1990) Caughlin et al. (2000).

Attachment perspectives place the negotiation of closeness and distance in the context of expectations about availability and rejection. Threatening disagreements can activate concerns about abandonment or vulnerability, while withdrawal can offer short-term control of emotional exposure. Interpersonal regulation theories add that the effect of an utterance depends on the response it receives; respectful directness may provide necessary information, whereas contemptuous escalation may undermine capacity to negotiate. These accounts identify overlapping mechanisms requiring discriminating tests: stress appraisal, attachment expectation, partner influence, and issue-level incentives must be measured rather than collapsed into a global communication style Simpson et al. (1996) Overall et al. (2009) Overall et al. (2013).

Couples as Dynamic and Interdependent Systems

A couple is a dyad with interdependent responses: one partner's behavior changes the probabilities of the other's subsequent behavior, and the resulting pattern can alter the first partner's next action. Consequently, individual measures of 'communication skill' may fail to represent relational dynamics. A demanding statement becomes part of a demand–withdraw sequence only through its relationship to a partner's distancing response; a pause may be cooperative self-regulation when negotiated but rejecting disengagement when prolonged without explanation. Interaction coding must preserve temporal order and distinguish patterns from isolated moves Christensen & Heavey (1990) Heavey et al. (1993).

Reciprocity is neither automatically constructive nor uniformly destructive. Mutual validation can help coordinate problem solving, yet mutual escalation can rapidly intensify a disagreement through increasingly hostile responses. Apparent behavioral symmetry can also mask unequal power, resources, and consequences. Researchers can analyze transition probabilities between coded states, the duration of affective episodes, the onset of interruption, or the time required to recover baseline interaction. Such models recognize that a dyad's average negativity may conceal consequential short events and differences in sequence structure Gottman (1993) Levenson & Gottman (1983).

Interdependence introduces important inferential difficulties. When both partners report declining satisfaction, their scores are correlated because they share circumstances and influence one another. Treating those observations as independent inflates apparent precision. Dyadic methods should separate individual-level actor associations from partner-level associations, test reciprocal change, and account for nested observations. Prospective reassessment further requires distinguishing stable between-couple differences from within-couple fluctuations. This matters because couples who generally communicate well may remain happier overall even when temporary improvements in communication do not generate large subsequent changes Johnson et al. (2022) Kanter et al. (2022).

Conflict Episodes and the Organization of Disagreement

Conflict episodes begin when a discrepancy becomes interactionally consequential. A disagreement about time, affection, parenting, finances, or domestic labor can unfold through different courses depending on how the issue is introduced, the partner's interpretation, and whether both people perceive an opportunity to influence the outcome. A useful episode definition specifies an onset, an issue, a sequence of responses, and a criterion for temporary closure. These features allow researchers to distinguish an unresolved disagreement that is calmly postponed from an escalating exchange in which one or both partners experience threat or humiliation Gottman (1993) Karney & Bradbury (1995).

An episode's apparent subject does not necessarily identify its central stakes. Discussions over household routines may involve fairness, autonomy, fatigue, or repeated experiences of being overlooked. However, researchers must avoid substituting an interpretive story for observed evidence. A study can assess each partner's account of the underlying concern before the discussion, code behavior during the exchange, and compare later descriptions of what happened. Divergence among these measures is itself informative: partners may differ in their appraisal even while they agree about observable actions. Such disagreements in interpretation can affect the sequence that follows Bradbury & Fincham (1990) Finkel et al. (2017).

Identifying resolution also requires care. An explicit agreement may settle a practical decision while leaving emotional hurt unaddressed. A conciliatory gesture may ease immediate tension while a substantive problem persists. Conversely, disagreement can remain while the partners recover sufficient responsiveness to continue cooperating. Research outcomes should therefore include task settlement, perceived justice, mutual understanding, emotional recovery, and behavioral follow-through. This multidimensional approach makes it possible to investigate which types of closure predict later outcomes under differing material and relational conditions Salvatore et al. (2011) Wilder et al. (2024).

Demand–Withdraw as a Sequential Interaction Pattern

Demand–withdraw describes an interaction in which one partner pressures, criticizes, requests change, or pursues discussion while the other becomes silent, evasive, disengaged, or otherwise withdraws. The sequence is consequential because pursuit can intensify when a response is withheld, whereas withdrawal can increase when pursuit feels intrusive or overwhelming. Christensen and Heavey's original research challenged a simple view of demand and withdrawal as fixed sex-linked attributes: the person whose concern or desired change was at issue influenced the distribution of behavior. Issue ownership and structural position are therefore necessary variables in explanations of the pattern Christensen & Heavey (1990).

Replication and prospective work clarified both the pattern's relevance and its methodological complexity. Heavey and colleagues reproduced relationships among conflict structure and demand–withdraw behavior, and later examined the prospective significance of observed sequences. These studies relied on defined laboratory contexts and partner assessments; observed patterns should not be treated as universal responses to all conflict or deterministic markers of breakup. At a functional level, withdrawal may avert immediate escalation, preserve attention, communicate disengagement, or protect someone from anticipated hostility. Its meaning depends on duration, antecedents, safety, and opportunities for subsequent reengagement Heavey et al. (1993) Heavey et al. (1995).

Contemporary contextual research highlights the need to avoid classifying every instance of withdrawal as equally maladaptive. Ross and colleagues analyzed economically diverse longitudinal samples and reported that behaviors commonly regarded as ineffective among more advantaged couples could function differently under socioeconomic disadvantage. Resource scarcity changes the perceived feasibility of addressing certain problems and the risks of persistent confrontation. This evidence invites preregistered tests of whether issue solvability, perceived control, and external constraints moderate the association of withdrawal with later functioning. It does not imply that coercive silencing or chronic dismissiveness becomes harmless when resources are limited Ross et al. (2019) Overall & McNulty (2017).

Escalation, Reciprocity, and Negative Affect

Escalation refers to a transition toward greater intensity, hostility, physiological mobilization, or interpersonal threat within an episode. It may involve mutual criticism, defensive counterarguments, contemptuous expressions, rapid interruptions, or increasingly global accusations. Gottman's interaction research examined engagement, validation, avoidance, and volatile patterns and challenged the simplistic assumption that all observable negative emotion predicts instability in the same way. An intense but mutually engaged debate can have a different course from persistent contempt or withdrawal accompanied by punitive control. Emotional intensity must therefore be interpreted alongside its function and its temporal organization Gottman (1993).

Behavioral reciprocity involves contingent changes in partners' actions. A critical statement may elicit defensiveness; an escalating reply may feed further criticism. Yet a similar initial statement can evoke acknowledgment or curiosity in another relational context. Researchers can investigate these possibilities through sequential coding and estimates of whether particular responses occur more often than baseline expectations. The question is whether response patterns add information beyond each partner's overall rates of behavior and the characteristics of the issue. A cross-sectional association between two negative behaviors cannot identify who initiated the exchange or how the pattern was maintained Levenson & Gottman (1983) Gottman & Levenson (1992).

The language of prediction requires unusually precise qualification. Longitudinal laboratory studies showed that measured affective and physiological patterns were associated with later relationship outcomes within the observed samples, but predictive accuracy can decline when tested in independent populations or under different conditions. A model trained on selected marriages, a specific observational task, and historically situated cohorts is not automatically a general-purpose divorce test. The proper research agenda combines preregistered coding, independently evaluated prediction, calibration for absolute risk, and inclusion of contextual factors that might account for the observed signal Gottman & Levenson (1992) Gottman & Levenson (2000) Kanter et al. (2022).

Criticism, Blame, and Attributional Feedback

Attributions connect an event to a presumed cause, a judgment of responsibility, and an expectation about whether the same conduct will recur. When a partner repeatedly experiences unmet agreements, attributing the behavior to an enduring unwillingness to cooperate can shape both emotional reactions and demands for change. Attributional models predict that global and blame-oriented interpretations often correspond with distress, but their interpretation requires knowledge of the actual event history. What seems like a negative attribution may accurately describe a repeated pattern. A valid study should therefore examine external evidence, partner reports, and temporal order instead of coding all suspicious interpretations as cognitive error Bradbury & Fincham (1990).

Trait characteristics can influence conflict partly through their effects on perception and interpersonal behavior. In research on anxiety, interpersonal negativity, and marital satisfaction, Caughlin and colleagues examined how dispositional vulnerabilities and communication processes relate to relationship evaluations. Their findings exemplify a central problem: people enter interaction with histories and tendencies that influence both what they notice and how they respond. If personal vulnerability is omitted, communication may appear to explain changes attributable partly to a stable predisposition. Repeated observation and careful covariate selection strengthen the test of genuinely interactional mechanisms Caughlin et al. (2000) Karney & Bradbury (1995).

The practical distinction between a complaint and a demeaning identity judgment also has scientific value. A complaint specifies an event and its consequence; a global attack attributes the problem to a partner's character. These categories can be coded independently from the speaker's intention and the listener's reaction. Yet there is no invariant threshold at which direct criticism is automatically harmful: when a substantive issue is repeatedly ignored, direct opposition may facilitate necessary change. Comparative research should test whether specificity, feasibility, perceived responsiveness, and the partner's capacity to respond explain why directness has different outcomes Overall et al. (2009) Overall & McNulty (2017).

Relationship Power, Issue Ownership, and Structural Constraints

Couples differ in who controls resources, whose needs routinely determine priorities, and which partner can safely disengage from a difficult discussion. These differences structure both the issues over which conflict occurs and the consequences of particular communication strategies. Research on demand–withdraw identified social structure and issue ownership as explanations for asymmetries that otherwise appeared to arise from sex alone. A partner requesting change may have little institutional power to secure it; the other partner may experience pressure as a threat to valued routines. Behavioral asymmetry should therefore be analyzed together with who stands to gain or lose from the proposed change Christensen & Heavey (1990) Heavey et al. (1993).

Material circumstances further shape the available solution space. If a couple cannot readily alter employment schedules, housing conditions, debt obligations, or caregiving demands, repeatedly confronting an insoluble immediate problem can carry costs that differ from those in financially secure relationships. The evidence on socioeconomic moderation suggests that some apparently negative tactics can have context-dependent functions. This does not license generalization from class categories to individual households; within-group variation is substantial, and causal mechanisms remain to be tested. Researchers should measure resources, stress exposure, and realistic options for change rather than substituting broad socioeconomic labels for actual constraint Ross et al. (2019) Karney & Bradbury (1995).

A more fundamental boundary appears when power is exercised through threats, intimidation, isolation, or violence. In such settings, agreement can reflect constrained choice and withdrawal can be a safety strategy. Mutual communication exercises may then fail to address the central problem or expose the more vulnerable person to retaliation. Scientific coding must distinguish a reciprocal dispute about incompatible preferences from a pattern of controlling behavior. Ethical research requires confidential assessment, safe recruitment and response procedures, and analysis that never assigns equal responsibility for coercive acts to both partners merely because the unit of recruitment is a couple World Health Organization (2026) Finkel et al. (2017).

Affective Arousal and Physiological Coordination

Conflict changes attention, appraisal, autonomic activity, and the ability to integrate a partner's meaning. Levenson and Gottman's laboratory study examined physiological linkage alongside affective exchange, establishing that relationship interactions may involve coordinated bodily responses rather than independent emotional events. Linkage alone does not prove emotional agreement or mutually beneficial regulation: two people can become physiologically synchronized while escalating each other's distress. Interpretation depends on the direction of change, the behavior surrounding that change, and the measurement interval. Physiology is therefore an additional channel of evidence, not a direct readout of relational health or unspoken intention Levenson & Gottman (1983).

During high-intensity conflict, people may attend more strongly to threat cues and have fewer resources available for considering alternative explanations. Theoretical models describe feedback among bodily activation, expressive behavior, and partner response. Yet identifying this feedback causally requires more than a correlation between heart rate and criticism. Investigators need baseline measures, temporal resolution, control conditions, and information about the physical or situational demands of the task. Laboratory conflict may elevate physiological measures because participants are being observed or asked to address a consequential issue, irrespective of their ordinary household interactions Gottman & Levenson (1992) Ben-Naim et al. (2013).

Affective coordination can be assessed through continuous behavioral ratings, physiological time series, and self-reported experience collected from both partners. The challenge is to align their different temporal scales. A momentary behavioral intervention may affect subjective emotion after a delay, while self-report at the end of the discussion may summarize the most memorable episode rather than every intervening shift. Statistical methods should test lagged relationships and account for shared task effects. The general scientific principle is that physiological coupling becomes informative only when associated with carefully specified behavioral transitions and independently defined outcomes Levenson & Gottman (1983) Salvatore et al. (2011).

Attachment, Threat, and Defensive Communication

Attachment orientations shape expectations about proximity, responsiveness, and vulnerability when an intimate relationship feels threatened. Simpson, Rholes, and Phillips studied conflict from an attachment perspective, linking differences in attachment orientation to affective and behavioral responses during relationship disagreements. Such findings support a conditional account: a partner's complaint can have different meanings depending on prior expectations about accessibility and rejection. However, self-reported attachment anxiety or avoidance should not be mistaken for immutable relationship types. People can respond differently across partners, conflicts, and periods of development Simpson et al. (1996) Finkel et al. (2017).

Defensive distancing can reduce immediate exposure to perceived criticism but also deprive a partner of information needed to resolve a shared problem. Overall, Simpson, and Struthers investigated whether partners could soften avoidant individuals' defensive responses during discussions of conflict. Their work supports attention to the interactional fit between a partner's vulnerabilities and the way concerns are expressed. It would be overgeneralization to conclude that a single 'soft' phrase universally resolves avoidance: relationship history, issue importance, and the actual behavioral response remain crucial. Effective explanation locates sensitivity within the dyad rather than labeling one person defective Overall et al. (2013).

Attachment accounts intersect with attribution and contextual stress models without replacing them. Repeated withholding may be interpreted as abandonment because of prior experience, but it may also be an accurate description of a partner's pattern; a desire for distance can reflect self-protection in an unsafe situation. Research should include independent assessment of partner conduct, power, and externally imposed constraints when evaluating attachment-based explanations. A useful experimental design manipulates the manner of approaching a difficult issue while observing responses from both partners and following the effects beyond the immediate exchange Simpson et al. (1996) Overall et al. (2009) World Health Organization (2026).

Emotion Regulation, Suppression, and Interpersonal Influence

Emotion regulation during disagreement includes changing attention, reappraising a partner's conduct, choosing when to speak, withholding an impulsive response, or seeking support. These strategies vary in their immediate and downstream consequences. In an experimental study of couples discussing conflict, Ben-Naim and colleagues examined emotion regulation in relation to attachment orientation and physiological arousal. Such evidence helps identify pathways by which a regulation strategy affects conflict experience under specified conditions. Experimental manipulation improves causal identification of a short-term response, while still leaving open its generality to ordinary disagreements and long-term relationship quality Ben-Naim et al. (2013).

Suppression exemplifies a construct that requires functional differentiation. Temporarily postponing an angry reply while preparing a more constructive response may aid negotiation; persistently hiding a serious concern can preserve a harmful pattern. Similarly, direct expression of negative emotion can communicate an important boundary or can invite retaliation, depending on how it is delivered and received. Overall and McNulty argue that beneficial conflict communication depends partly on the circumstances and on whether change is possible. This view predicts interactions between strategy, issue severity, partner responsiveness, and relationship history rather than one universal ranking of emotional tactics Overall & McNulty (2017) Overall et al. (2009).

Dyadic regulation adds a partner-mediated route to individual regulation. A listener may reduce distress through acknowledgment, practical help, or a credible willingness to change; conversely, invalidation may heighten threat even when the speaker tries to calm down. Meta-analytic research on dyadic coping links shared stress-management processes with relationship satisfaction, but it does not demonstrate that every association is caused by coping behavior. Self-report common-method effects and preexisting relationship quality may contribute. Experiments, daily designs, and partner reports are complementary ways to test mechanisms without conflating these different causal propositions Falconier et al. (2015) Johnson et al. (2022).

Repair Attempts, Reception, and Consequential Change

Repair can involve acknowledging the partner's experience, accepting responsibility, correcting an inaccurate assertion, making a practical concession, asking for clarification, or returning to a suspended conversation under safer conditions. The same behavior can carry different meanings in different relationships. An apology may be a meaningful act of accountability when it is specific and followed by changed conduct; it can also function as a ritual that terminates discussion without addressing the underlying harm. Consequently, researchers should code both the content of attempted repair and its subsequent uptake, rather than designating all conciliatory moves as successful reconciliation Wilder et al. (2024) Salvatore et al. (2011).

Recent diary research on romantic partners' efforts to reconcile examined how repair strategies are judged in ongoing relationships. Wilder, Prager, and Garapati investigated the helpfulness of actions taken after disagreement, including the importance of listening and partners' perceptions. This work makes post-conflict repair an object of measurement in its own right. Nevertheless, perceptions of helpfulness are not interchangeable with independently observed behavioral change, reduction in recurrence, or durable increases in trust. Dyadic diaries can reduce retrospective bias but remain vulnerable to selective reporting and to the influence of the day's prior relationship climate Wilder et al. (2024).

The distinction between making peace and correcting a problem is fundamental. Emotional recovery can occur before a substantive decision is reached, and practical agreement can occur while one partner remains hurt. An adequate study should measure when a repair attempt occurs, whether it is recognized, whether both partners regain conversational coordination, and whether the relevant commitment is honored later. These dimensions permit evaluation of a sequence from conflict to short-term recovery and subsequent change. No fixed threshold of apologizing or positive gestures can substitute for evidence that the relationship's actual problem has been addressed Salvatore et al. (2011) Overall et al. (2009).

Recovery Trajectories After Disagreement

Post-conflict recovery is a temporal phenomenon. In research following romantic partners beyond a conflict episode, Salvatore and colleagues investigated the ability to recover affective coordination and its developmental correlates. Their approach separates behavior during disagreement from what happens after the dispute has ostensibly ended. Recovery may involve renewed warmth, willingness to engage in a different task, reduced defensiveness, and restoration of ordinary interaction. This separation matters because couples with similar conflict intensity can differ in whether antagonism persists, is quickly resolved, or resurfaces during a subsequent exchange Salvatore et al. (2011).

Recovery can unfold across several horizons. In minutes, partners may shift from anger to ordinary conversation; over a day, they may reconsider responsibility or return to a practical issue; across weeks, repeated changes can alter expectations about whether future difficulties will be manageable. These timescales should be specified before analysis. Laboratory observations offer fine temporal detail but limited ecological breadth. Daily diaries capture routine interactions but often provide less precise coding of utterances. Longitudinal follow-up can assess change but is exposed to intervening life events and partnership selection. A cumulative account requires linkage among these distinct measures Wilder et al. (2024) Karney & Bradbury (1995).

Successful recovery is not necessarily a quick return to positive affect. In the presence of betrayal, serious neglect, or coercive conduct, sustained caution and slower restoration of trust can be appropriate responses. Scientific measurement must avoid defining the vulnerable partner's immediate forgiveness as the benchmark for relational health. A more defensible operationalization includes safety, voluntary participation, accountability, and behavioral follow-through. These indicators make repair a potentially consequential process while preserving a distinction between emotional soothing, conflict avoidance, and actual alteration of circumstances that produced harm World Health Organization (2026) Wilder et al. (2024) Overall & McNulty (2017).

What Longitudinal Evidence Establishes About Communication

Longitudinal studies address whether observed or reported communication at one occasion predicts later relationship outcomes. Kanter and colleagues synthesized dyadic longitudinal research to evaluate the claim that communication is a major independent cause of subsequent satisfaction and dissolution. Their quantitative findings support associations, with average prospective effects considerably smaller than many broad popular claims suggest. Differences in measurement, observation timing, and the inclusion of baseline adjustment all affect interpretation. A small average association can still be scientifically informative, especially when recurring behaviors affect many people, but it cannot by itself establish a large treatment effect from teaching a particular conversational technique Kanter et al. (2022).

Within-couple and between-couple relationships answer different questions. A couple whose members generally communicate more constructively than those in another couple may also report greater satisfaction; this difference can reflect shared history, resources, temperament, or other unobserved characteristics. A more demanding test asks whether the same couple becomes subsequently more satisfied after its communication improves relative to its own usual level. Johnson and colleagues examined within-couple longitudinal associations, contributing evidence that contemporaneous links are often clearer than prospective changes. This distinction limits simple causal stories inferred from cross-sectional correlations and supports designs able to separate stable selection from dynamic influence Johnson et al. (2022).

Prospective evidence can nevertheless clarify temporal ordering and bound explanatory claims. If negative communication precedes later distress after accounting for earlier distress, it is consistent with an influence hypothesis; reverse causality, omitted stressors, and bidirectional change may remain plausible. If associations largely disappear after appropriate adjustment, researchers should revise the expected causal importance of the measured construct rather than dismiss the evidence. Stronger conclusions will require interventions with verified process changes, intensive longitudinal observation, and explicit tests of whether improvement in one process precedes improvement in another Kanter et al. (2022) Karney & Bradbury (1995).

Observational Coding, Validity, and Generalizability

Behavioral coding converts recorded interaction into analytically specified categories such as criticism, positive affect, withdrawal, validation, defensiveness, and engagement. The resulting data have methodological value only if observers can consistently apply the categories and the recorded encounter samples a meaningful domain of behavior. Coders should receive training, remain masked to later outcomes when possible, and demonstrate reliability for both categories and episode boundaries. Researchers should report whether speaking turns, time intervals, continuous ratings, or whole-discussion judgments were used, because aggregation affects what a 'negative communication' score represents Gottman (1993) Christensen & Heavey (1990).

Laboratory conflict tasks improve standardization by asking each couple to discuss a selected disagreement under similar recording conditions. They can generate rich evidence about order, timing, posture, vocal affect, and physiological activation. However, participant awareness of observation, unequal topic severity, task duration, and sample selection constrain generalization to home interactions. A problem chosen by one partner may elicit a different pursuit pattern from a problem chosen by the other. Researchers should document how topics were selected, the perceived importance of each issue, and whether the coding scheme preserves reciprocal sequences instead of treating partners as independent speakers Heavey et al. (1993) Levenson & Gottman (1983).

Ecological validity is improved through complementary naturalistic recordings and paired daily diaries, although each adds its own limitations. Diaries reduce the interval between event and report but may miss low-salience interactions or overrepresent emotionally intense episodes. Naturalistic recordings are sensitive to consent, selective recording, and privacy. In either method, reports from both partners improve assessment of divergent experiences. Cross-method convergence should be treated as a research question rather than presumed: a global survey of communication quality, a ten-minute laboratory discussion, and an evening diary may index different components of a relational process Wilder et al. (2024) Johnson et al. (2022).

Dyadic Analysis, Causality, and Statistical Precision

Statistical dependency is intrinsic to couple research. Partners share environments, influence one another, and may be jointly selected into a sample based on relationship status. Appropriate models accommodate repeated measurements and correlated responses. Actor and partner associations can be estimated separately, as can lagged effects at different timescales, but each parameter needs a clear substantive definition. An actor association links an individual's report or behavior with that individual's outcome, whereas a partner association links one person's characteristics with the other's outcome. Neither is automatically causal in observational data Kanter et al. (2022) Johnson et al. (2022).

Analyses of change are complicated by measurement error and regression to the mean. People who report unusually high conflict on one assessment may look better at a later assessment simply because initial extreme scores contained transient noise. Couples who remain in a multiyear study can also differ systematically from those who separate or decline participation. A credible longitudinal analysis describes attrition, missing-data assumptions, prior levels of outcome variables, and whether nonindependence was addressed. Reporting uncertainty intervals and plausible effect magnitudes is more useful than converting an isolated significant p value into a categorical claim about relationship functioning Karney & Bradbury (1995) Kanter et al. (2022).

Experimental evidence contributes a different type of inference. Researchers can manipulate aspects of expression, appraisal, or immediate emotion regulation and observe short-term responses, as in conflict regulation experiments. Randomization supports causal interpretation of the manipulated condition within that setting, although it does not automatically establish durable effects on satisfaction. Couple intervention trials can test broader packages of change, but identifying the active mediator requires measuring actual communication shifts and showing that they account for later outcomes. Combining randomized intervention, observational process measurement, and repeated partner reports provides a stronger cumulative approach Ben-Naim et al. (2013) Christensen et al. (2010).

Development, Aging, and Long-Term Change

Couples change as individual capacities, shared routines, and external demands evolve. Conflict during the formation of a household may concern expectations that have not yet stabilized, while later conflicts can involve employment, health, caregiving, and changes in autonomy. Longitudinal accounts of marriage place adaptive processes within stress and developmental time. This perspective predicts that the same observable behavior can be differently consequential at different stages of a relationship. A pause may allow deliberation in one period and signal long-standing avoidance in another; researchers must test such interpretations rather than relying on age stereotypes Karney & Bradbury (1995) Finkel et al. (2017).

Holley, Haase, and Levenson examined age-related changes in demand–withdraw communication, using repeated observations to study how couple behavior changes across later adulthood. Evidence of age trends should be interpreted in light of cohort differences, relationship duration, and selective survival of partnerships. Couples observed after many years together represent a selected population of continuing relationships, which may differ from those dissolved earlier. Researchers should distinguish chronological age, years partnered, historical period, and the influence of physical limitations or changing social roles on how partners manage disagreement Holley et al. (2013).

Developmental approaches to repair also examine the origins of regulatory capacities and interpersonal expectations. Salvatore and colleagues used post-conflict recovery to connect developmental histories with adult couple processes. Such associations suggest that regulation in intimate partnerships has a history, yet they do not entail that childhood relationships fix future outcomes. Subsequent partners, contexts, learning, and substantive changes in safety can alter recurring patterns. A rigorous lifespan model should test both continuity and change while avoiding explanations that attribute current difficulties solely to an individual's early experience Salvatore et al. (2011) Simpson et al. (1996).

Children, Family Systems, and the Reach of Couple Conflict

Conflict between caregivers can be relevant to children through emotional security, stress exposure, modeling of interaction, and changes in parenting cooperation. A meta-analysis by van Eldik and colleagues examined associations between the interparental relationship and children's maladjustment and responses to conflict. Such synthesis supports studying couple interaction as part of a wider family system. However, parental disagreement varies considerably in form, frequency, severity, and resolution. Researchers should distinguish ordinary disagreements that are managed without intimidation from chronic hostility, threats, and violent exchanges, rather than treating all exposure as one uniform risk van Eldik et al. (2020).

Family-level effects can follow indirect pathways. Disruptive couple conflict may change caregiver availability, generate inconsistent routines, or increase demands on a child to regulate adult emotions. At the same time, family economic circumstances, caregiver mental health, and environmental stressors can contribute to both parental conflict and child outcomes. Observational associations therefore support multiple plausible causal directions. Longitudinal designs with independent child assessments and carefully modeled confounders help separate mechanisms, while intervention studies can ask whether changes in adult behavior coincide with improvements in children's functioning van Eldik et al. (2020) Karney & Bradbury (1995).

Repair within a caregiving household has implications beyond whether two partners feel reconciled. Children may observe adults acknowledge mistakes, address concrete problems, and resume cooperative coordination; alternatively, they may observe repeated hostility followed by assurances without behavioral change. Direct empirical tests are needed before claiming that a particular apology teaches a specific developmental lesson. The scientific objective is to study observable exposure, perceived security, and downstream family outcomes while recognizing that children's reactions are shaped by age, previous history, social support, and the severity and consequences of the dispute van Eldik et al. (2020) Salvatore et al. (2011).

Cultural Context, Socioeconomic Conditions, and Relational Diversity

Research based on married heterosexual couples in North American laboratories provides valuable evidence about a defined population and setting. It cannot establish that identical conflict norms apply across cultural communities, nonmarital partnerships, sexual and gender identities, or historically different arrangements of household labor. What counts as respectful disagreement, an acceptable pause, or a public expression of distress may vary by context. The appropriate scientific response is to specify the population and behavioral task, examine measurement equivalence, and recruit diverse samples with sufficient representation to estimate heterogeneity rather than merely mentioning diversity in a limitation statement Finkel et al. (2017) Kanter et al. (2022).

Socioeconomic context is particularly consequential because communication and resource allocation are intertwined. Ross and colleagues documented that the longitudinal implications of ostensibly maladaptive communication differed across levels of economic disadvantage. One interpretation is that avoiding repeated confrontation may reduce immediate costs when a partner has little control over the material problem being discussed. That account remains a mechanism to test rather than a universal behavioral prescription. Researchers should measure constrained choices, household instability, issue resolvability, and partner perceptions directly, and should test whether the observed moderation replicates in distinct samples Ross et al. (2019) Overall & McNulty (2017).

Historical relationship research also needs to account for the legal and institutional positions of partners. Demands for change have different stakes when employment, caregiving, housing, or exit options are unequally distributed. A pattern described as one partner's avoidance may arise partly from an unequal ability to impose the status quo, while apparent acquiescence may reflect dependence or risk. Comparative methods can integrate narrative interviews with independently coded interactions and prospective outcomes to determine which interpretation best fits the evidence. This scientific commitment to context strengthens construct validity without abandoning careful behavioral measurement Christensen & Heavey (1990) Karney & Bradbury (1995) World Health Organization (2026).

Intervention Evidence and the Scope of Couple Treatment

Couple interventions may target constructive communication, cognitive interpretation, behavioral exchange, emotional accessibility, problem solving, or acceptance of persistent differences. A review by Lebow and colleagues synthesized research on treating couple distress, establishing an evidence base for structured approaches while drawing attention to methodological differences across trials. Intervention research tests whether a treatment package improves measured outcomes under study conditions; it does not necessarily establish that one conversational technique is the active ingredient. Baseline relationship severity, treatment engagement, therapist effects, and follow-up attrition influence the interpretation of apparent benefit Lebow et al. (2012).

Christensen and colleagues followed participants from a randomized clinical trial comparing traditional behavioral couple therapy with integrative behavioral couple therapy. The multiyear follow-up contributes stronger evidence about treatment trajectories than a single uncontrolled post-treatment satisfaction score. Still, comparisons between treatments should be interpreted using the actual randomized allocation, assessment timing, attrition, and chosen outcomes. A trial of a complex therapeutic program cannot be converted into proof that all couples should use one language formula or a fixed number of repair gestures. Process measures and mediation tests are needed to identify which changes account for results Christensen et al. (2010) Lebow et al. (2012).

Research on ordinary couple communication must preserve a safety boundary. Interventions centered on joint disclosure and reciprocity can be inappropriate where a person faces coercion, intimidation, or violence. An apparently neutral exercise that asks both people to acknowledge equivalent responsibility may obscure asymmetric harm. The scientific and ethical requirement is to distinguish those contexts before assessing intervention applicability. Public educational discussion of conflict research may explain studied processes but should not present an individualized treatment plan or imply that better communication alone resolves violent control World Health Organization (2026) Overall & McNulty (2017).

Reproducibility, Measurement Disagreement, and Competing Explanations

Relationship science faces a recurrent challenge: measures that share an appealing label often operationalize different phenomena. Global perceptions of communication, coded criticism in a single conversation, partner-rated responsiveness, and the frequency of daily disagreements are related constructs with distinct sources of error. A meta-analysis can pool effects only when their operational differences are addressed. Researchers should disclose coding manuals, interrater reliability, scoring rules, analysis decisions, and the full range of sampled interactions. Open data may require careful privacy protections because recordings and dyadic responses contain sensitive interpersonal information Kanter et al. (2022) Gottman (1993).

Independent replication is especially important for vivid claims about predicting dissolution from a brief conversation. Results can be sensitive to sample composition, how couples are recruited, which outcomes define dissolution, the observation task, and how predictive thresholds are selected. External validation should evaluate both discrimination and calibration, provide uncertainty intervals, and examine whether models perform similarly across populations. Prediction and causation remain separate questions: a behavior can help forecast dissatisfaction even if modifying that behavior does not change the outcome. Conversely, an intervention may help a subset of couples despite a modest aggregate prospective association Gottman & Levenson (2000) Kanter et al. (2022).

Competing explanations should be incorporated explicitly rather than acknowledged only after analysis. One model treats hostile exchanges as a driver of dissatisfaction; another treats dissatisfaction as a cause of negative interpretations and conduct; a third treats material stress or enduring vulnerabilities as common causes. All three can generate similar observational correlations. Intensive longitudinal measurement, randomization of feasible components, and transparent estimation of within-couple change can help adjudicate among them. The evidence to date favors a plural account in which reciprocal influence is plausible but the average magnitude and the distribution of causal effects remain important empirical questions Johnson et al. (2022) Karney & Bradbury (1995) Bradbury & Fincham (1990).

Technological Mediation and Testable Questions in the Artificial Era

Digital communication changes the observable conditions of couple disagreement. Text messaging delays or removes vocal and facial cues, enables editing before sending, and creates persistent records of statements. Synchronous video introduces different cues and coordination constraints. The implications for repair depend on what partners infer from latency, acknowledgment, and the perceived credibility of a response. Findings from face-to-face laboratory discussions cannot simply be transferred to asynchronous exchanges. Methodological continuity requires comparable definitions of disagreement, repair initiation, reception, and subsequent outcome across communication channels Finkel et al. (2017) Overall & McNulty (2017).

Artificial systems may become additional participants in the communicative environment by suggesting wording, summarizing disagreements, or helping a partner prepare a conversation. Such involvement raises measurable hypotheses about perceived agency, disclosure, epistemic trust, fairness, and whether suggestions foster understanding or merely more persuasive influence. Controlled studies could compare unassisted preparation with different forms of automated support while examining both partners' experiences and longer-term follow-through. No inference that an artificial assistant possesses subjective emotion is required to investigate its observable effects as a mediating system Finkel et al. (2017) Kanter et al. (2022).

The Artificial Era offers a broader question about how relational coordination changes when communication resources become technically redistributed. A system that proposes an apology may influence its wording, timing, and sincerity as perceived by the recipient; a mediator that receives only one partner's narrative may reproduce that person's attributions while presenting them as neutral. Research designs should therefore document who controls the system, what information it receives, which outputs reach the other partner, and how outcomes are measured. Ethical evaluation must consider confidentiality, dependency, differential power, and the possibility of amplifying coercion. These are prospective research questions, not established benefits of AI-mediated couple repair Bradbury & Fincham (1990) World Health Organization (2026).

Conclusion

Couple communication, conflict, and repair are interconnected processes organized through partner interdependence, histories of interpretation, issue ownership, affect regulation, and the opportunities available for consequential change. The historical literature identified patterns of engagement, escalation, avoidance, and demand–withdraw; later research established that these patterns are sensitive to context and cannot be treated as fixed indicators of character or sex. Experiments illuminate immediate regulation mechanisms, longitudinal studies examine temporal associations, and observational coding reveals within-episode sequences. Each method addresses a different level of explanation and offers different strengths of causal inference Christensen & Heavey (1990) Ben-Naim et al. (2013) Karney & Bradbury (1995).

The strongest quantitative synthesis supports meaningful but generally modest average prospective links between communication and later relationship functioning. This requires proportionate claims about prediction and intervention: improving an observed behavior is a testable pathway rather than a guaranteed route to satisfaction or stability. Repair is best conceptualized as a process involving an initiative, its perception by the other partner, short-term emotional recovery, and behavior that addresses the substantive problem. Socioeconomic conditions, attachment-related sensitivities, developmental history, family responsibilities, and relationship safety can alter the consequences of superficially similar acts Kanter et al. (2022) Johnson et al. (2022) Salvatore et al. (2011) Ross et al. (2019).

A cumulative research program should preregister temporally explicit definitions, combine recordings and both partners' reports, test independent replication, and compare model predictions under different contextual constraints. It should distinguish partner influence from common stressors, communication from coercion, and perceived reconciliation from demonstrated changes in behavior. Intervention research should identify active mechanisms and measure durable outcomes, including safety and autonomy. The resulting scientific framework understands conflict as a dynamic relational event and repair as a verifiable trajectory of restored coordination, accountability, and adaptation. Those principles remain open to rigorous testing as technological mediation changes the environments in which human partnerships operate Lebow et al. (2012) Wilder et al. (2024) World Health Organization (2026).

References

  • Ben-Naim, S., Hirschberger, G., Ein-Dor, T., & Mikulincer, M. (2013). An experimental study of emotion regulation during relationship conflict interactions: The moderating role of attachment orientations. Emotion, 13(3), 506–519. https://doi.org/10.1037/a0031473
  • Bradbury, T. N., & Fincham, F. D. (1990). Attributions in marriage: Review and critique. Psychological Bulletin, 107(1), 3–33. https://doi.org/10.1037/0033-2909.107.1.3
  • Caughlin, J. P., Huston, T. L., & Houts, R. M. (2000). How does personality matter in marriage? An examination of trait anxiety, interpersonal negativity, and marital satisfaction. Journal of Personality and Social Psychology, 78(2), 326–336. https://doi.org/10.1037/0022-3514.78.2.326
  • Christensen, A., & Heavey, C. L. (1990). Gender and social structure in the demand/withdraw pattern of marital conflict. Journal of Personality and Social Psychology, 59(1), 73–81. https://doi.org/10.1037/0022-3514.59.1.73
  • Christensen, A., Atkins, D. C., Baucom, B., & Yi, J. (2010). Marital status and satisfaction five years following a randomized clinical trial comparing traditional versus integrative behavioral couple therapy. Journal of Consulting and Clinical Psychology, 78(2), 225–235. https://doi.org/10.1037/a0018132
  • Falconier, M. K., Jackson, J. B., Hilpert, P., & Bodenmann, G. (2015). Dyadic coping and relationship satisfaction: A meta-analysis. Clinical Psychology Review, 42, 28–46. https://doi.org/10.1016/j.cpr.2015.07.002
  • Finkel, E. J., Simpson, J. A., & Eastwick, P. W. (2017). The psychology of close relationships: Fourteen core principles. Annual Review of Psychology, 68, 383–411. https://doi.org/10.1146/annurev-psych-010416-044038
  • Gottman, J. M. (1993). The roles of conflict engagement, escalation, and avoidance in marital interaction: A longitudinal view of five types of couples. Journal of Consulting and Clinical Psychology, 61(1), 6–15. https://doi.org/10.1037/0022-006X.61.1.6
  • Gottman, J. M., & Levenson, R. W. (1992). Marital processes predictive of later dissolution: Behavior, physiology, and health. Journal of Personality and Social Psychology, 63(2), 221–233. https://doi.org/10.1037/0022-3514.63.2.221
  • Gottman, J. M., & Levenson, R. W. (2000). The timing of divorce: Predicting when a couple will divorce over a 14-year period. Journal of Marriage and Family, 62(3), 737–745. https://doi.org/10.1111/j.1741-3737.2000.00737.x
  • Heavey, C. L., Layne, C., & Christensen, A. (1993). Gender and conflict structure in marital interaction: A replication and extension. Journal of Consulting and Clinical Psychology, 61(1), 16–27. https://doi.org/10.1037/0022-006X.61.1.16
  • Heavey, C. L., Christensen, A., & Malamuth, N. M. (1995). The longitudinal impact of demand and withdrawal during marital conflict. Journal of Consulting and Clinical Psychology, 63(5), 797–801. https://doi.org/10.1037/0022-006X.63.5.797
  • Holley, S. R., Haase, C. M., & Levenson, R. W. (2013). Age-related changes in demand-withdraw communication behaviors. Journal of Marriage and Family, 75(4), 822–836. https://doi.org/10.1111/jomf.12051
  • Johnson, M. D., Lavner, J. A., Mund, M., Zemp, M., Stanley, S. M., Neyer, F. J., Impett, E. A., Rhoades, G. K., Bodenmann, G., Weidmann, R., Bühler, J. L., Burriss, R. P., Wünsche, J., & Grob, A. (2022). Within-couple associations between communication and relationship satisfaction over time. Personality and Social Psychology Bulletin, 48(4), 534–549. https://doi.org/10.1177/01461672211016920
  • Kanter, J. B., Lavner, J. A., Lannin, D. G., Hilgard, J., & Monk, J. K. (2022). Does couple communication predict later relationship quality and dissolution? A meta-analysis. Journal of Marriage and Family, 84(2), 533–551. https://doi.org/10.1111/jomf.12804
  • Karney, B. R., & Bradbury, T. N. (1995). The longitudinal course of marital quality and stability: A review of theory, method, and research. Psychological Bulletin, 118(1), 3–34. https://doi.org/10.1037/0033-2909.118.1.3
  • Lebow, J. L., Chambers, A. L., Christensen, A., & Johnson, S. M. (2012). Research on the treatment of couple distress. Journal of Marital and Family Therapy, 38(1), 145–168. https://doi.org/10.1111/j.1752-0606.2011.00249.x
  • Levenson, R. W., & Gottman, J. M. (1983). Marital interaction: Physiological linkage and affective exchange. Journal of Personality and Social Psychology, 45(3), 587–597. https://doi.org/10.1037/0022-3514.45.3.587
  • Noller, P., Feeney, J. A., Bonnell, D., & Callan, V. J. (1994). A longitudinal study of conflict in early marriage. Journal of Social and Personal Relationships, 11(2), 233–252. https://doi.org/10.1177/0265407594112005
  • Overall, N. C., Fletcher, G. J. O., Simpson, J. A., & Sibley, C. G. (2009). Regulating partners in intimate relationships: The costs and benefits of different communication strategies. Journal of Personality and Social Psychology, 96(3), 620–639. https://doi.org/10.1037/a0012961
  • Overall, N. C., Simpson, J. A., & Struthers, H. (2013). Buffering attachment-related avoidance: Softening emotional and behavioral defenses during conflict discussions. Journal of Personality and Social Psychology, 104(5), 854–871. https://doi.org/10.1037/a0031798
  • Overall, N. C., & McNulty, J. K. (2017). What type of communication during conflict is beneficial for intimate relationships? Current Opinion in Psychology, 13, 1–5. https://doi.org/10.1016/j.copsyc.2016.03.002
  • Ross, J. M., Karney, B. R., Nguyen, T. P., & Bradbury, T. N. (2019). Communication that is maladaptive for middle-class couples is adaptive for socioeconomically disadvantaged couples. Journal of Personality and Social Psychology, 116(4), 582–597. https://doi.org/10.1037/pspi0000158
  • Salvatore, J. E., Kuo, S. I.-C., Steele, R. D., Simpson, J. A., & Collins, W. A. (2011). Recovering from conflict in romantic relationships: A developmental perspective. Psychological Science, 22(3), 376–383. https://doi.org/10.1177/0956797610397055
  • Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conflict in close relationships: An attachment perspective. Journal of Personality and Social Psychology, 71(5), 899–914. https://doi.org/10.1037/0022-3514.71.5.899
  • van Eldik, W. M., de Haan, A. D., Parry, L. Q., Davies, P. T., Luijk, M. P. C. M., Arends, L. R., & Prinzie, P. (2020). The interparental relationship: Meta-analytic associations with children's maladjustment and responses to interparental conflict. Psychological Bulletin, 146(7), 553–594. https://doi.org/10.1037/bul0000233
  • Wilder, S., Prager, K. J., & Garapati, S. (2024). But I said I'm sorry: Helpfulness of romantic couples' efforts to reconcile after conflict. Personal Relationships, 31(3), 836–854. https://doi.org/10.1111/pere.12565
  • World Health Organization. (2026, June 18). Violence against women. https://www.who.int/news-room/fact-sheets/detail/violence-against-women