Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Interdependence, Investment, and Relationship Commitment

Abstract. Relationship commitment arises within situations of interdependence in which present outcomes, perceived alternatives, accumulated investments, and future expectations shape the motivation to maintain a partnership. This critical narrative monograph traces the conceptual foundations of social exchange and interdependence theory, examines the formation and psychometric operationalization of the Investment Model, and integrates original experimental research, longitudinal couple studies, meta-analyses, and evidence on accommodation, sacrifice, and dissolution. It distinguishes satisfaction, dependence, voluntary dedication, and constraints, explains why observed persistence alone cannot establish relationship quality or safety, and evaluates life-course, cultural, economic, and institutional moderators. Its synthesis identifies strong correlational support for the central predictors of commitment, substantial limitations to causal inference and individualized forecasting, and testable extensions for technologically mediated and human–artificial relations. The review is a critical narrative integration rather than an original experiment, registered systematic review, or new meta-analysis.

Keywords: interdependence theory; Investment Model; relationship commitment; relationship satisfaction; quality of alternatives; investment size; relationship dissolution; cognitive interdependence; dyadic longitudinal research; relationship maintenance; commitment versus constraint; relationship safety; relationship decision-making; human–artificial relations

The Research Object and Its Explanatory Boundaries

Interdependence, investment, and relationship commitment constitute a connected but analytically differentiated field of close-relationship science. Interdependence describes the degree and form in which one person's outcomes depend on another person's choices, available resources, and repeated interaction. Investment denotes resources tied to a relationship whose value would be diminished or lost through its dissolution. Commitment concerns a person's orientation toward maintaining a particular relationship through time. The research problem is to explain how patterns of outcome dependence, subjective evaluations, alternatives, and accumulated commitments produce persistence, constructive conduct, and sometimes costly continuation. These constructs become scientifically useful when independently measured and connected through falsifiable predictions rather than treated as alternative names for love Rusbult (1980) Rusbult & Van Lange (2003) Le & Agnew (2003).

The principal unit of analysis may be an individual, a dyad, a repeated interaction, or a social environment. Individual perceptions of alternatives can diverge from objective opportunity, partners can hold unequal investments, and shared commitments can persist through changing satisfaction. Researchers therefore distinguish actor-level reports from partner-level reports, momentary preferences from longitudinal intentions, and consensual dedication from externally imposed constraints. A person can remain in an unhappy union because of caregiving duties, housing arrangements, social expectations, or restricted resources; another can be satisfied yet foresee little continuity. Observed persistence alone cannot establish security, mutuality, fairness, or freely chosen dedication Rhoades et al. (2010) Karney & Bradbury (1995).

This monograph is a critical narrative synthesis of foundational theory, original experiments, prospective surveys, dyadic studies, measurement research, and published quantitative syntheses. The guiding questions concern the origin and explanatory structure of interdependence theory, the distinct roles of satisfaction, alternatives and investments, the formation of a commitment orientation, the behavioral effects and limitations of commitment, and the conditions under which people stay or leave. It also evaluates cultural, institutional and technological changes that may alter the availability of alternatives and the structure of investments. No original participants were recruited, no new pooled effect was calculated, and no claim of systematic-review completeness is made. Claims are evaluated according to the available design, measurement, and reproducibility information Finkel et al. (2017) Tran et al. (2019).

From Social Exchange to the Interdependence Framework

The historical trajectory begins with formal approaches to interpersonal exchange and their distinction between a person's absolute outcomes and the outcomes available in alternative circumstances. Thibaut and Kelley presented their account of group interaction in 1959, and Kelley and Thibaut developed the interdependence approach in a 1978 theoretical volume. In this tradition, interaction is evaluated relative to a comparison level reflecting expectations and a comparison level for alternatives reflecting what the person believes could be obtained elsewhere. These comparisons answer different questions: whether current participation is satisfying and whether continuation seems preferable to departure. The distinction remains central because evaluations of satisfaction and dependence can move independently Kelley & Thibaut (1978) Rusbult & Van Lange (2003).

Interdependence theory investigates the structure of situations in which outcomes are jointly determined. A choice that is immediately attractive to one participant may impose costs on the other, while a collectively beneficial choice may require restraint. The framework identifies patterns of mutual dependence, corresponding and conflicting interests, relative power, temporal dependence, and how information about partners is acquired. Rusbult and Van Lange's 2003 synthesis described six dimensions of interdependence situations and placed transformation of motivation at the center of interaction: immediate self-interest can be reorganized around future consequences, partner welfare, and joint goals. These dimensions describe possible interaction structures rather than fixed personality types Rusbult & Van Lange (2003).

Formal exchange language offers useful comparative discipline. The same behavioral act can have different meanings when it occurs under different options, obligations, and power arrangements. Remaining after disagreement may indicate shared purpose, low mobility, fear of retaliation, or concern about dependents. Observational equivalence creates an identification problem: conduct is not a transparent measure of the mechanism producing it. Scientific progress therefore requires separate estimates of expected outcomes, feasible options, perceived risks, internal commitment, and partner responses. Treating interpersonal exchange as a whole system also prevents a narrow interpretation in which only immediately visible rewards and costs count Finkel et al. (2017) Rhoades et al. (2010).

Satisfaction, Comparison Levels, and Perceived Alternatives

Satisfaction is an evaluative judgment about experienced relationship outcomes relative to standards, hopes, or expectations. Within exchange-based models, rewards and costs acquire meaning through those standards rather than through universal values. The prospect of leaving introduces a different reference point: the anticipated quality of the best available alternative, which may involve another partner, living independently, spending more time with friends, or reorganizing one's life. The scientific model predicts that satisfaction and alternative quality can independently contribute to the motivation to maintain a union. A favorable current relationship can still appear replaceable, and an unfavorable one can remain difficult to exit when available alternatives are poor Rusbult (1980) Drigotas & Rusbult (1992).

Perceived alternatives have informational and structural components. Individuals must imagine plausible future experiences, estimate uncertainty, and assess accessible material and social resources. The quality of an imagined alternative is not a direct inventory of available people; remaining single may be an attractive option, and external opportunities can be constrained by childcare, income, mobility, disability, law, or threats. Two respondents with identical reported dissatisfaction may accordingly show different levels of dependence. Since alternatives are partly forecasts, their reports can also change after conflict, social comparison, or a major transition without any objective change in opportunities Rusbult & Van Lange (2003) Finkel et al. (2017).

The measurement implication is straightforward: satisfaction scales cannot substitute for alternative-quality scales. The 1998 Investment Model Scale explicitly separated commitment, satisfaction, alternative quality, and investment size, and its validation studies examined whether these variables formed distinguishable factors. Prospective prediction requires particular care when evaluations, alternatives, and intention to continue are collected in the same questionnaire; similar wording and the same respondent may inflate their apparent association. Whenever feasible, independent reports and later behavioral outcomes should supplement contemporaneous evaluations Rusbult et al. (1998) Le & Agnew (2003).

Rusbult's Investment Model and Its Early Empirical Tests

Caryl E. Rusbult's 1980 paper organized three hypothesized bases of relationship commitment: satisfaction, quality of alternatives, and investment size. The model treats investment as a resource attached to an existing association that would lose some value if the relationship ended. Shared history, intimate knowledge, integrated friendship networks, financial arrangements, time, and jointly maintained projects may all become salient investments, although their psychological meaning differs across individuals. In the model's conventional directional account, greater satisfaction and investment and poorer perceived alternatives are associated with greater commitment. Commitment is thereby distinguished from simple attraction to the present relationship Rusbult (1980).

The 1980 article reported two initial tests, and Rusbult's 1983 longitudinal study followed people in relatively new heterosexual dating relationships at repeated intervals. The latter reported that commitment changed with satisfaction, alternatives, and investments while showing how development and deterioration could be studied as processes rather than static group differences. Its small undergraduate sample and historical setting limit population generalization, and the prospective observations alone cannot establish that each predictor was experimentally manipulated. Nevertheless, the design was important because it moved the research program toward within-relationship change and termination outcomes Rusbult (1983).

The investment model is a theory of commitment formation and dependence, and each proposed path deserves separate scrutiny. Investment can represent freely chosen joint development or accumulated obstacles to departure; alternatives can reflect genuinely available possibilities or biased estimates; and satisfaction can reflect present rewards while failing to capture safety. The three predictors are consequently informative but not exhaustive explanations of why a relationship continues. An adequate contemporary application also considers power asymmetry, constraints imposed by institutions, and differences between the partners' interpretations of the same shared resource Rhoades et al. (2010) Stanley et al. (2006).

Dependence, Dedication, and Constraint as Distinct Dimensions

Dependence is a structural or perceived relation between the outcomes obtained within the present union and those anticipated outside it. Dedication refers to a personally endorsed orientation toward the partner and shared future. Constraint refers to features that make an exit difficult, costly, or consequential independently of present relational satisfaction. The distinctions permit a high-dependence, low-dedication configuration: someone may wish to leave yet be held by finances, housing, legal entanglement, social pressure, or fear. They also permit strong dedication without large material entanglements. A single global commitment score can conceal these divergent motivational situations Rhoades et al. (2010) Stanley et al. (2006).

Stanley, Rhoades, and Markman proposed the possibility that relationship transitions can create inertia when material and social constraints grow before a deliberate decision about shared commitment. Their 2006 discussion of sliding versus deciding offers a conceptual account of cohabitation transitions rather than evidence that all cohabitation has one causal effect. Rhoades and colleagues then studied dedication and three constraint dimensions in 1,184 unmarried adults and found that the different facets supplied distinct information about relationship adjustment and subsequent stability. In that prospective study, greater felt constraint, an experience of being trapped, differed from other types of constraint in its association with continuation Stanley et al. (2006) Rhoades et al. (2010).

This distinction has direct consequences for ethical interpretation. Constraint can facilitate valued long-term cooperation when partners freely endorse a commitment; it can also help preserve an unequal or dangerous arrangement. Continuing a relationship is therefore a descriptive outcome to be explained, not an automatic criterion of well-being. Studies that collapse voluntary dedication, material costs, and felt entrapment into one indicator risk mistaking reduced exit capacity for relational health. Research involving abuse or coercive control requires direct assessment of safety, threat, access to assistance, and the practical feasibility of leaving Rhoades et al. (2010) Dutton & Painter (1993).

Operational Measurement of the Investment Model

The Investment Model Scale developed by Rusbult, Martz, and Agnew in 1998 established an operational framework for four distinct constructs: satisfaction level, perceived quality of alternatives, investment size, and commitment level. In three validation studies, internal-consistency assessments and factor analyses supported a four-factor organization, and longitudinal evidence linked early measurements to later adjustment and relationship status. The instrument offers a reproducible route from conceptual variables to observable survey responses. A scale's validation in particular populations, however, is not a guarantee that its factor structure or item meaning will remain invariant across age groups, relationship types, languages, and social environments Rusbult et al. (1998).

Reliability and construct validity answer different questions. High interitem consistency indicates covariation among answers within an item set, while discriminant validity asks whether commitment can be empirically distinguished from satisfaction, attachment security, or felt obligation. Criterion validity asks whether the measure predicts behavior or outcomes external to itself. Researchers should examine response distributions, item wording, reverse-coded items, measurement invariance, temporal stability, and the consequences of treating ordinal responses as approximately continuous. It is also informative to ask whether a person's low perceived alternatives score reflects genuine resource scarcity, low desire for change, cultural norms, or simply an inability to imagine the future Rusbult et al. (1998) Finkel et al. (2017).

When partners are simultaneously assessed, actor and partner reports create additional possibilities. A dyadic design can test whether one partner's dedication predicts the other's confidence or persistence after accounting for that other person's own reports. Repeated measurement can ask whether within-person changes in alternatives precede changes in commitment. Behavioral records, observations of problem solving, and later status reports provide partial protection against same-method bias. Reports should specify their prediction horizon and the meaning of dissolution: a temporary separation, a canceled marriage plan, a permanent breakup, and legal divorce represent different outcomes Karney & Bradbury (1995) Le et al. (2010).

Meta-Analytic Evidence for the Three Core Determinants

The 2003 meta-analysis by Benjamin Le and Christopher Agnew combined 52 studies, 60 independent samples, and 11,582 participants. Satisfaction, alternative quality, and investment size each showed associations with commitment; the three constructs jointly accounted for nearly two-thirds of the variance in commitment across the reported models. Commitment also predicted breakup. This synthesis was pivotal because it moved the Investment Model beyond single-sample demonstrations and assessed generalization across relational and some nonrelational contexts. Reported relations were stronger in relationship domains, a finding consistent with the psychological specificity of enduring close bonds Le & Agnew (2003).

An updated synthesis by Tran and colleagues drew on 202 independent samples involving 50,427 participants collected between 1980 and early 2016. The abstract reports aggregate correlations between commitment and satisfaction of r = .65, investments of r = .53, and quality of alternatives of r = −.43. These are summary associations across distinct samples and instruments, not experimentally established causal coefficients. The absolute ordering indicates that reported satisfaction is, on average, the strongest correlational predictor among the three; it does not imply that individual participants always prioritize satisfaction, or that intervention on one variable would reproduce the corresponding correlation as a causal effect Tran et al. (2019).

Both syntheses are constrained by the studies available for inclusion and the operational definitions those studies chose. Shared-response methods, selective retention in longitudinal cohorts, historical changes in partnership norms, and correlated predictor variables may complicate interpretation. Meta-analytic heterogeneity should inform rather than disappear behind a single mean coefficient. The evidence strongly supports the usefulness of the model as a predictive account of perceived commitment, while causal mechanisms and applications to populations under coercive or structurally unequal conditions require other designs and independent outcome measures Le & Agnew (2003) Tran et al. (2019).

Cognitive Interdependence and the Representation of a Shared Future

Commitment is associated with cognitive organization as well as explicit intentions. Agnew, Van Lange, Rusbult, and Langston's 1998 study investigated cognitive interdependence, a representation of the self in relation to a continuing pair. They assessed such indicators as spontaneous plural-pronoun use, perceived unity between self and partner, and the centrality of the relationship to identity. Their cross-sectional and two-wave longitudinal evidence connected commitment to this shared representation and suggested reciprocal changes over time. The result supports a dynamic account in which a person's imagined future, identity and routine decisions become partly organized around a relational unit Agnew et al. (1998).

The pronoun marker is useful because it approaches commitment through ordinary language rather than only through a direct question about intended persistence. Its meaning nevertheless depends on linguistic conventions, context, the topic of conversation, and conscious impression management. Saying we does not establish a harmonious partnership, and saying I does not establish its absence. Self–other overlap is similarly multidimensional: joint planning may coexist with personal boundaries, asymmetric influence, or different beliefs about future obligations. Evidence gains force when convergent indicators show that the shared relationship is represented and enacted in plans, attention, and decisions Agnew et al. (1998) Finkel et al. (2017).

Cognitive interdependence offers a possible feedback process. Greater commitment can increase accessibility of joint goals, while frequent joint planning can reinforce an expectation of continuity. Such reciprocal relations are plausible and consistent with repeated-measures associations, but temporal feedback estimates remain vulnerable to unmeasured stable traits and changing contexts. Future tests should distinguish within-person changes from between-couple differences and measure whether shared representation improves adaptation, suppresses individuality, or varies according to the quality of partner responsiveness Agnew et al. (1998) Joel et al. (2020).

Transformation of Motivation in Repeated Interaction

Interdependence theory separates immediate preference from transformed preference. A partner may initially favor an action because it brings an immediate benefit, then reconsider after including the other's welfare, long-term reciprocity, or the future of the relationship. This transformation is particularly relevant when partners' interests temporarily diverge. It can involve delaying an impulse, making an explicit joint decision, or forgoing a resource in response to a valued partner's need. The transformation process is a theoretical account of how social motives enter ordinary action, and its empirical content is tested through responses to conflict, behavioral choices, and repeated interactions Rusbult & Van Lange (2003) Van Lange et al. (1997).

The opportunity for prosocial choice arises from the structure of a situation. A cooperative act that entails no sacrifice provides limited evidence about a person's willingness to give up an immediate benefit. Conversely, willingness to incur costs under pressure may indicate asymmetric dependence rather than freely expressed care. The scientific question is therefore how participants understand their options and motives. Experimental scenarios can isolate selected contingencies, but ecological observation is needed to identify how concrete histories, caregiving responsibilities, and recurring conflicts shape decisions. The distinction between personal intention and a relationship's actual quality remains central Rusbult et al. (1991) Impett et al. (2005).

Transformation also connects momentary choices to longer developmental trajectories. Repeated constructive choices can alter expectations of reliability, while repeated destructive responses may reduce confidence in the partner's future cooperation. Models that incorporate both actors' motives can make specific predictions about how behavioral sequences accumulate into trust or resentment. These predictions require temporal ordering, partner reports, and behaviorally grounded outcomes rather than assuming that all instances of compromise are beneficial Wieselquist et al. (1999) Overall & McNulty (2017).

Accommodation, Sacrifice, and Observable Maintenance

Accommodation concerns responding constructively, or refraining from an immediately destructive response, when a partner behaves in a hurtful or inconsiderate way. Rusbult and colleagues' 1991 research reported six studies examining this process and linked accommodation to commitment, satisfaction, alternatives, and investments. The studies advanced a behavioral bridge between dependence-based commitment and subsequent interaction. Accommodation may take the form of problem solving, restraint, patience, and constructive engagement, although the value of any response depends on whether the situation is safe and changeable Rusbult et al. (1991).

Van Lange and colleagues' 1997 six-study investigation linked willingness to sacrifice to commitment and examined associations with adjustment and persistence using survey, simulation, and longitudinal methods. The reported role of commitment provides evidence for a maintenance pathway, but willingness to sacrifice cannot be identified with relationship quality by definition. Impett, Gable, and Peplau's 2005 daily-life research showed that approach-motivated sacrifice and avoidance-motivated sacrifice have differing associations with well-being and relationship functioning. Thus motivation, reciprocity, and context determine the interpretive value of the same outwardly costly act Van Lange et al. (1997) Impett et al. (2005).

Wieselquist and colleagues' 1999 longitudinal studies extended the pathway by examining commitment, perceived prorelationship acts, and trust. In their account, observable behavior signals a partner's willingness to consider shared outcomes, and that signal may support subsequent trust and dependence. This framework helps specify how commitment can affect another person's experience through behavior rather than mere self-description. Yet trust should not be inferred automatically from self-sacrifice, particularly when one partner has limited alternatives or perceives a threat. A scientific assessment must observe mutual responsiveness, costs, and the partner's independent evaluation Wieselquist et al. (1999) Overall & McNulty (2017).

Relationship Dissolution as a Prospective Outcome

Relationship dissolution can test commitment theories only when the outcome is defined prospectively and its observation is sufficiently reliable. A breakup event is not simply the inverse of satisfaction: partners may dissolve a satisfying relationship after relocation, incompatibility of future goals, or changing circumstances, while a dissatisfied partnership may persist because of shared caregiving or limited resources. In studies with repeated reports, investigators must specify whether they predict an individual's desire to leave, an announced decision, a temporary separation, or an actual termination confirmed by later follow-up. These outcomes differ in their time scales and dependence on both partners' actions Karney & Bradbury (1995) Le et al. (2010).

Le and colleagues' 2010 meta-analysis integrated evidence from 137 studies and 37,761 participants in nonmarital relationships. Commitment, love, inclusion of other in the self, and dependence were among the stronger predictors of dissolution, while satisfaction, alternatives, and investments were more modest direct predictors. This pattern is compatible with the argument that commitment serves as a relatively proximal summary of multiple pressures and motivations, although the design of the contributing studies limits causal claims. The quantitative synthesis also emphasized relational and social-network factors, showing that dissolution cannot be modeled only as a private comparison between two partners Le et al. (2010).

The meaning of prediction must be specified. A statistically associated variable may offer limited improvement when added to a well-calibrated multivariable forecast, and predictive accuracy depends on breakup prevalence, follow-up interval, reporting losses, and population selection. Couples recruited in stable periods differ from couples actively contemplating separation. Future work should compare nested models, report uncertainty and calibration, preregister outcomes, and validate findings in independent cohorts. Especially important is accounting for the fact that breakup decisions are dyadic events even when measured through one partner's survey Le et al. (2010) Joel et al. (2018).

Ambivalence and the Decision to Stay or Leave

Stay–leave deliberation involves more than computing a net score of positive and negative features. People can simultaneously perceive strong reasons for continuing and strong reasons for ending a relationship. Joel, MacDonald, and Page-Gould's research on the structure of stay–leave reasons identified multiple categories of motivations, including intimacy and investments on the staying side and conflict or violations of trust on the leaving side. Ambivalence is therefore a meaningful psychological state with distinct reasons and time horizons rather than mere survey indecision Joel et al. (2018).

Commitment can also reflect concern for the other person's welfare. Joel, Impett, Spielmann, and MacDonald's 2018 study examined how concern for a partner's dependence and anticipated distress enters choices about ending a relationship. Such findings extend narrow self-interest accounts by treating a breakup as a decision that affects someone to whom the decision maker is connected. Their significance lies in identifying other-oriented motives, not in establishing an obligation to remain. Concern may facilitate compassionate discussion or prolong a difficult transition, and its role must be assessed alongside personal safety and autonomy Joel et al. (2018).

Longitudinal decisions may have several stages: questioning, discussing, preparing, leaving, reconsidering, and reorganizing practical ties. Alternative expectations, commitment, and available resources can change at each stage. An evidence-based framework therefore separates the determinants of breakup contemplation from the practical execution of leaving and the eventual durability of separation. Research that codes only the final binary status discards important information about repeated decisions, shared obligations, and post-dissolution contact Joel et al. (2018) Rhoades et al. (2010).

Dedication, Material Constraints, and the Inertia of Transitions

Material integration develops through mundane choices such as sharing rent, moving, caring for dependents, merging finances, and arranging work schedules. These commitments can stabilize meaningful cooperation, but they can also raise the cost of leaving faster than confidence in the relationship grows. Stanley, Rhoades, and Markman's inertia model highlights the distinction between a deliberate decision to build a durable partnership and an accumulation of constraints that happens through sequential transitions. Its theoretical value is to expose a temporal order in which choices alter the opportunity structure for future choices Stanley et al. (2006).

Rhoades and colleagues measured dedication, perceived constraints, material constraints, and felt constraint separately in a large unmarried sample, and each facet contributed information about eight-month stability. The distinction shows why a single question about how committed a couple is can be ambiguous. Someone who reports few alternatives might be deeply invested in a valued shared life, or might describe practical isolation. Researchers should ask which resources would remain accessible after leaving and whether the relationship creates support or restricts independence Rhoades et al. (2010).

The causal status of inertia remains a separate research question. Couples select into cohabitation, shared property, and other transitions for reasons that themselves predict stability; the apparent impact of a transition may partly reflect selection. Stronger tests require prospective observation before decisions, measurement of relationship quality and motives, and comparisons among groups with genuinely similar prior circumstances. Interpretations must also be sensitive to legal context and economic conditions, because the same housing arrangement may generate radically different exit costs across populations Stanley et al. (2006) Karney & Bradbury (1995).

Longitudinal Dyadic Processes and Partner Asymmetry

A close relationship is generated by at least two participants whose perceptions and motivations can diverge. One partner's high satisfaction may coincide with the other's low satisfaction; one may view alternatives optimistically while the other perceives few options. These divergences matter because dissolution can follow a unilateral decision, whereas everyday maintenance often requires ongoing mutual coordination. The dyad should therefore be represented through paired measurements rather than by treating one person's account as an objective property of the relationship Finkel et al. (2017) Wieselquist et al. (1999).

A common analytical strategy is to compare actor effects, in which a person's own perceptions predict that person's outcomes, with partner effects, in which one partner's perceptions predict the other's outcomes. Repeated dyadic measurement then permits examination of temporal ordering, shared events, and mutual adaptation. Yet correlated reports can arise from common experiences, assortative selection, or response style; partner associations alone do not prove influence. Models should account for nonindependence and evaluate whether changes within a couple predict later changes more convincingly than stable differences between couples Karney & Bradbury (1995) Joel et al. (2020).

A large-scale synthesis of dyadic data by Joel and numerous colleagues analyzed 43 longitudinal couples datasets from 29 laboratories, involving 11,196 couples. Relationship-specific perceptions, including perceived partner commitment, were among useful predictors of current relationship quality. The investigation also found substantial difficulty predicting subsequent change in relationship quality from the measured variables. This contrast between cross-sectional fit and longitudinal forecasting cautions against interpreting a rich contemporary profile as a reliable map of the future. The study concerns relationship quality, and its results should not be substituted for a direct breakup hazard estimate Joel et al. (2020).

Commitment, Autonomy, and Relational Safety

The ethical adequacy of a commitment model depends on distinguishing enduring partnership from desirable partnership. Dedication can coexist with mutual flourishing, but continuity can also result from intimidation, unequal power, or absence of feasible alternatives. The outcome of remaining together is normatively incomplete until safety, autonomy, mutual respect, and the quality of care are considered. Research on traumatic attachments in abusive relationships illustrates that enduring emotional ties and persistent costs can coexist, although a specific clinical mechanism should not be inferred without evidence about the person's experience and circumstances Dutton & Painter (1993) Rhoades et al. (2010).

Standard questionnaires may be especially misleading when respondents cannot disclose coercion privately or when leaving carries credible danger. A statement of poor alternatives may encode economic precarity, threatened harm, separation barriers, or dependent children as well as ordinary preferences. Any application to intimate partner violence must incorporate ethical recruitment, confidentiality, appropriate safety procedures, and direct measures of violence and coercive control. The relationship-science literature on accommodation and sacrifice must similarly distinguish freely chosen cooperation from compliance in response to fear Dutton & Painter (1993) Impett et al. (2005).

A strong interpretation of commitment therefore has at least two axes: motivation toward a shared future and the freedom and safety to revise that future. Because no single scalar can represent both, research should report relationship continuity alongside well-being, autonomy, psychological distress, material vulnerability, and exposure to harm. The practice of describing greater persistence as an unqualified success confuses prediction with ethical assessment. Researchers can preserve the explanatory strength of investment theory while explicitly recognizing cases where reduced dependence or a supported exit is the favorable outcome Rhoades et al. (2010) Overall & McNulty (2017).

Development, Life Transitions, and Changing Opportunity Structures

Investment and commitment evolve across the life course as partners accumulate shared history and as circumstances alter the costs of continuation. Education, migration, employment, illness, caregiving, parenthood, and aging can change what each person expects from a relationship and what alternatives are realistically available. These transitions can increase cooperation or create unequal obligations. The same objective change can have opposite effects on perceived satisfaction and commitment depending on expectations, resources, and prior trust. Developmental research should therefore model life events as potential moderators and sources of change rather than as universally beneficial or harmful stages Karney & Bradbury (1995) Finkel et al. (2017).

The historical evidence base for the Investment Model contains substantial samples of young adults and heterosexual dating partnerships. Such samples were appropriate for early theory tests but cannot represent every lifespan stage or contemporary household form. Adult partners caring for children, older relatives, or a disabled spouse may understand investment partly through moral obligation and continuity of care. Those commitments can remain meaningful while available alternatives change over time. Future research benefits from including long-term cohabiting partners, marital and nonmarital unions, diverse gender configurations, and partnerships formed or maintained across substantial geographic distance Rusbult (1983) Le et al. (2010).

A dynamic account must distinguish within-person adaptation from changes in sample composition. Relationships that endure become a selected group; couples who leave are no longer available for assessments of ongoing satisfaction in the same form. Analyses restricted to continuing unions can consequently create misleading impressions of improvement or stability. Survival models, planned follow-up of separated participants, and explicit treatment of missingness are needed to estimate trajectories without conditioning entirely on persistence. These design issues are fundamental to claims about how investment accumulates over years Karney & Bradbury (1995) Le et al. (2010).

Culture, Norms, Institutions, and Perceived Mobility

Commitment takes shape within cultural and institutional arrangements. Social expectations about marriage, kinship, care, gender, personal independence, and the legitimacy of separation influence which alternatives are socially imaginable and practically feasible. Legal provisions concerning property and custody, access to income, housing markets, migration status, and community support can change exit costs even when the partners' feelings remain constant. The construct of alternative quality is therefore partly culturally and institutionally situated, and using a standard survey item across settings requires evidence that respondents interpret it comparably Finkel et al. (2017) Rusbult & Van Lange (2003).

General claims about cultural differences require representative cross-cultural evidence and suitable measurement equivalence. A correlational estimate from college-based convenience samples may generalize imperfectly to couples whose resources, social standing, or family arrangements differ. The investment framework offers hypotheses rather than a complete cultural taxonomy: satisfaction might predict commitment differently when separation is economically costly, and the meaning of a shared household investment might differ across family structures. These predictions need direct studies that compare contexts while controlling for age, education, household composition, and local constraints Tran et al. (2019) Karney & Bradbury (1995).

Institutions also influence the meaning of choice. A reported preference to stay can be sincerely endorsed within inherited obligations, while pressure or exclusion can narrow options in ways that self-report measures understate. Rather than deciding in advance which culture produces more authentic commitment, research should assess how people themselves define obligations and which resources support autonomous decision-making. A methodologically sound account separates descriptive distributions of commitments from moral judgments about particular relationship forms Rhoades et al. (2010) Finkel et al. (2017).

Economic Costs, Investments, and Asymmetric Power

The word investment can imply voluntary contribution, sunk cost, asset ownership, time, or irreversible dependence. These forms should not be collapsed. A jointly purchased home may confer both security and an exit barrier, while years of caregiving may build shared meaning but also reduce one partner's earnings or social opportunities. The relationship between investments and commitment depends on who controls resources, who bears foregone opportunities, and how losses would be distributed. Measures that simply ask how much a person has invested may fail to capture this asymmetry Rusbult (1980) Rhoades et al. (2010).

Perceived alternatives also depend on resources outside the relationship. Employment, transportation, housing, child care, education, social support, and personal safety shape the realistic feasibility of a separate life. A strong attachment to the current union may coexist with a genuine lack of exit options. The challenge for research is to identify distinct pathways by which investments strengthen dedication, generate constraint, and affect material welfare. Survey associations alone are inadequate to determine how changing the distribution of resources would alter commitment, because partners' arrangements are products of prior choices and institutional conditions Stanley et al. (2006) Rhoades et al. (2010).

An analytically richer design would record relational investments, financial assets and debts, time allocated to care, foregone employment, control over shared resources, and both partners' estimates of available support. Within-couple comparisons can then identify circumstances where the same joint project has unequal implications. Outcomes should include not only union survival but financial security, psychological adjustment, equitable bargaining, and freedom to negotiate. These extensions preserve the model's explanatory focus on dependence while broadening its ability to distinguish chosen interdependence from vulnerability Finkel et al. (2017) Karney & Bradbury (1995).

Study Designs, Causal Inference, and Evidence Hierarchies

The literature spans cross-sectional questionnaires, experimental decision simulations, daily diaries, prospective panels, dyadic observations, and meta-analyses. Each design answers a different question. A cross-sectional correlation establishes that two self-reported constructs vary together in the sampled population. A prospective design tests whether earlier measurements forecast later events. An experiment with randomized conditions can identify a causal effect of the manipulated condition on the observed outcome within its design. Neither a simulation nor a laboratory interaction automatically identifies the causal origins of real-world dissolution over years Rusbult (1980) Van Lange et al. (1997) Le & Agnew (2003).

Many attractive causal narratives remain observationally underidentified. Satisfaction may raise commitment, but commitment may also alter the interpretation of current events; partners may jointly respond to an unmeasured stressor. To discriminate these possibilities, researchers can measure temporal sequences with multiple waves, model stable differences, test natural changes in options, and preregister alternative explanations. Cross-lagged effects do not eliminate all confounding, and mediation analyses based on concurrent self-reports cannot independently prove a mechanistic chain. Robustness checks should ask whether findings persist across reasonable model specifications and independent samples Wieselquist et al. (1999) Joel et al. (2020).

Evidence synthesis adds breadth but also introduces selection and comparability challenges. Le and Agnew's 2003 synthesis and Tran and colleagues' 2019 update provide complementary historical coverage and aggregate results. Their numerical estimates depend on included measures and samples. A narrative monograph may integrate such results with primary experiments and critical theory, but it should not be labeled a new meta-analysis without a protocol, comprehensive search, explicit eligibility criteria, effect extraction, risk-of-bias assessment, and reproducible quantitative procedures. The present analysis uses published syntheses for their reported findings and describes uncertainty according to their evidentiary basis Le & Agnew (2003) Tran et al. (2019).

Psychometric, Sampling, and Replication Challenges

A commitment scale can be internally coherent yet conflate intention, identity, and felt obligation. Even when statistically separable, scales may share method variance because the same respondent answers similarly framed questions during a single session. Psychometric validation requires attention to factor structure, convergent and discriminant evidence, test–retest properties, language adaptation, and measurement invariance across groups. Measurement must also remain sensitive to meaningful within-person change without becoming so reactive that the questionnaire itself alters respondents' interpretations of their relationships Rusbult et al. (1998).

Sampling mechanisms affect inference. Volunteers in relationship surveys may be more open to discussing their unions than people in severe conflict or those facing safety concerns. Longitudinal attrition is informative when unhappy or dissolved couples become harder to contact. Studies of individual survey respondents cannot recover mutual influence merely by assigning dyadic terminology to one person's beliefs. The composition of samples in an evidence synthesis should therefore be reported by age, partnership type, geography, recruitment setting, and time period whenever those characteristics are available Karney & Bradbury (1995) Le et al. (2010).

Replication demands more than reproducing a significant sign in a new regression. Independent studies should reproduce construct meaning, assess predictions out of sample, distinguish direct forecasts from intermediary variables, and report absolute effect sizes and uncertainty. The updated Investment Model meta-analysis provides evidence of broad correlational regularity, whereas the large-scale 2020 couples project demonstrates the difficulty of predicting changes in satisfaction. A mature research program can accommodate both findings because understanding average contemporaneous differences and predicting an individual relationship's future are different scientific tasks Tran et al. (2019) Joel et al. (2020).

Commitment in Social Networks and Third-Party Evaluations

Romantic relationships unfold within broader social networks of friends, relatives, colleagues, and institutions. Partners encounter external perspectives on relationship value, experience demands that compete with couple time, and can draw on assistance that changes their perceived alternatives. Social-network support was a robust predictor in Le and colleagues' meta-analysis of nonmarital dissolution, indicating that explanatory accounts limited to internal feelings omit a relevant context of persistence. Network variables can also influence what partners believe others expect them to do, independently of the actual benefits and costs they experience within the union Le et al. (2010) Finkel et al. (2017).

Third-party involvement can work through several pathways. Approval may encourage public commitment; disapproval may intensify scrutiny or create stress; and practical support may enlarge feasible choices. Shared networks can themselves become investments when breakup would disrupt friendships, routines, or a community identity. Such ties matter differently when partners possess equivalent access to friends and family versus when one has become socially isolated. A network measure should therefore distinguish the number of connections, the quality of available support, and whether those connections increase or restrict options to act independently Rusbult (1980) Rhoades et al. (2010).

Research can connect relationship trajectories to changing network composition using repeated reports and network measures collected from multiple participants. A perceived decline in outside support might precede higher dependence without increasing satisfaction; alternatively, an external network can supply resources that make renegotiation or safe separation more feasible. These hypotheses illustrate why social embeddedness has no uniform meaning in commitment models. Researchers should test the mechanism of network involvement rather than assume that either integration or independence is invariably protective Le et al. (2010) Karney & Bradbury (1995).

Commitment Dynamics, Feedback, and Thresholds

Static commitment scores describe a moment within a process that may include long periods of relative stability and rapid changes during particular events. Rusbult's longitudinal approach already recognized that satisfaction, investment and alternatives can develop or deteriorate together. A useful dynamic extension would model commitment as a time-indexed variable with feedback from behavior, perceived partner dedication, and changes in the opportunity structure. The same initial commitment score could precede different trajectories when partners differ in responsiveness or when external constraints abruptly change Rusbult (1983) Wieselquist et al. (1999).

Feedback is a substantive hypothesis requiring temporal evidence. If greater commitment encourages accommodation, and successful accommodation strengthens perceived partner responsiveness, subsequent commitment may rise. But repeated unilateral concessions can accumulate resentment or deepen constraint. Therefore, feedback can be reinforcing or destabilizing depending on what participants perceive and receive. Intensive longitudinal data, repeated behavioral measurement, and dyadic models can investigate these possibilities while accounting for day-to-day mood, common events, and stable individual differences Rusbult et al. (1991) Impett et al. (2005).

An important unresolved problem concerns thresholds. A person might tolerate ordinary disappointments while commitment is high but reconsider the partnership after a major trust violation, resource loss, or safety concern. A purely additive model may miss such transitions, and apparent abrupt change may instead reflect unmeasured earlier deterioration. Researchers can compare linear and nonlinear specifications, examine time to breakup, and assess prediction in held-out couples. Any claim of an identified threshold should be supported by repeated observations rather than imposed retrospectively on a single breakup narrative Le et al. (2010) Joel et al. (2020).

Mutuality, Discrepancy, and the Distribution of Investment

Couples may agree that their relationship is important while disagreeing about who is more committed, who has sacrificed more, and who could leave with fewer losses. These discrepancies have theoretical significance because the Investment Model is typically measured separately within each individual, whereas outcomes emerge from joint action. Symmetric scores could reflect genuine concordance or a shared response style; asymmetric scores could reflect power inequality, different expectations, or complementary roles. The joint distribution is therefore a scientific object in its own right, beyond the mean commitment reported for the couple Rusbult et al. (1998) Finkel et al. (2017).

Measurement of asymmetry should separate several questions: do partners hold similar intentions to continue; do they anticipate similar post-separation resources; and do they perceive the other's dedication accurately? High agreement about intentions can coexist with sharp inequalities in costs or access to independent income. Researchers should measure actual household arrangements, perceived fairness, and the partner's own reports before treating one person's belief about shared investment as a verified dyadic fact. This approach also avoids using persistence to conceal inequalities that the original commitment model was not designed to diagnose Rhoades et al. (2010) Stanley et al. (2006).

Temporal analyses may ask whether discrepancies predict failed cooperation, changes in perceived safety, or eventual separation beyond both partners' average levels. Such predictions should be preregistered because alternative specifications of discrepancy scores can produce different results. Dyadic structural equations, actor–partner paths, and models of within-couple change may be useful when sample size, measurement reliability, and number of waves are sufficient. Scientific interpretation should focus on the specific path supported by data rather than converting a statistical difference into a diagnosis of the relationship Karney & Bradbury (1995) Joel et al. (2020).

Counterfactual Alternatives and the Problem of Choice

The alternatives component is conceptually counterfactual: it asks what life might be like if the present relationship did not continue. Such scenarios cannot be directly observed for a person who remains in the relationship, and reported alternatives combine preference with imagination, knowledge, and uncertainty. Measures should differentiate access to a potential partner from the broader prospect of living independently, reconnecting with social networks, or changing one's circumstances. Drigotas and Rusbult's dependence formulation made these alternative comparisons central to understanding breakup decisions; contemporary designs can examine how accurately participants forecast their options and how those forecasts change Drigotas & Rusbult (1992) Rusbult & Van Lange (2003).

Experimental manipulations of alternative salience can identify short-term changes in judgment but should be distinguished from interventions that actually alter available resources. A participant may rate other partners favorably after a comparison task without making any durable change in commitment; another may see the present relationship differently after gaining access to safe housing or independent income. This distinction is especially important because perceived alternatives can be shaped by social exclusion, financial dependency, or coercion. Research should establish whether an effect concerns cognition, the opportunity structure, or both Rhoades et al. (2010) Joel et al. (2018).

A rigorous counterfactual research program would combine scenario-based tasks, objective environmental measures, prospective reports, and naturally occurring changes in mobility or resources. It would track the timing of evaluations, the practical steps taken, and subsequent outcomes. Care is needed with inferential claims because changes in alternatives are rarely assigned randomly, and they may co-occur with stress, new social ties, or a change in relationship quality. The clearest evidence will emerge when competing explanations are measured before the choice and models are tested against independent data Karney & Bradbury (1995) Finkel et al. (2017).

Research Transparency and Translation Across Relationship Forms

Generalization requires a transparent description of who was studied and which form of relationship was represented. A model calibrated in a heterosexual dating sample may offer useful hypotheses for married partners, cohabiting couples, long-distance relationships, consensually nonmonogamous arrangements, or relationships constrained by migration rules, but the constructs must be re-examined within each setting. In particular, the quality of alternatives and the meaning of investment depend on relationship expectations, social recognition, and legal rights. A translated questionnaire should be tested for substantive equivalence, not only linguistic similarity Rusbult (1980) Tran et al. (2019).

Reproducible evidence also depends on accurate documentation of study design. Researchers should report recruitment, inclusion criteria, partnership duration, relationship status, missingness, follow-up windows, preregistered outcomes, analytic decisions, and effect uncertainty. Breakup prediction must identify who reports the event and whether dissolution is verified by both partners. Investigations of power or conflict must protect participants' privacy and safety, particularly where one person's participation could reveal information to a controlling partner. Sharing aggregated analyses should not expose identifying histories of vulnerable couples Le et al. (2010) Rhoades et al. (2010).

The broader goal is cumulative theoretical testing across changing settings. Evidence from the revised Investment Model literature suggests that the principal correlational relationships are robust at the aggregate level, but the strength of prediction and the explanatory pathways still depend on context. New research can compare models that incorporate dedication, constraint, perceived partner commitment, institutional conditions, and temporal feedback, then evaluate their performance in genuinely new populations. This approach retains the conceptual discipline of interdependence theory and makes its scope of application an empirical matter Tran et al. (2019) Joel et al. (2020).

Commitment in Relationship Maintenance and Intervention Research

Scientific accounts of relationship maintenance can inform the design of assessment and support, provided that preserving the relationship is not treated as the universal objective. Understanding satisfaction, alternatives, and investments may help explain why a pair wants to remain together while experiencing recurrent disputes. Communication research illustrates the importance of context: Overall and McNulty's 2017 review found that direct opposition can be helpful when serious problems can be changed, whereas affirming indirect strategies may be more appropriate in other circumstances. A generic demand for greater compromise can therefore miss the conditions under which constructive change is possible Overall & McNulty (2017).

Commitment has at least two potential roles in intervention research. It may support repeated investment in difficult but worthwhile change, and it may suppress action when change is impossible or unsafe. Its role can be studied by measuring motivation, feasible options, partner behavior, and follow-up outcomes separately. The observational evidence concerning accommodation and sacrifice supports mechanistic hypotheses about cooperation, but it does not by itself establish that instructing clients to sacrifice more would improve relationships. Practical recommendations require suitable intervention trials and direct assessment of harms Rusbult et al. (1991) Impett et al. (2005).

In a scientifically responsible framework, clinical and educational uses of these concepts should specify whether the aim is improved relationship functioning, informed choice, reduction of coercion, or safer separation. Outcomes can include mutual understanding, independent decision capacity, symptom burden, and reduced destructive interaction rather than dissolution rates alone. This monograph summarizes theories and evidence; it does not provide individualized relationship counseling or substitute a scientific prediction for the choices of people facing complex circumstances Overall & McNulty (2017) Rhoades et al. (2010).

Technological Mediation and the Artificial Era

Digital environments alter the visibility of alternatives and the organization of relational investments. Messaging services, dating platforms, shared digital accounts, virtual communities, and AI-mediated interfaces can affect how people anticipate available relationships and coordinate ongoing ones. Yet technical availability of contacts does not equal an attractive or feasible personal alternative. Scientific use of investment theory should distinguish increased exposure to options from changes in perceived alternative quality, and distinguish the storage of shared digital history from the psychological cost of losing it. These are extensions of established constructs to new conditions, not settled empirical effects Rusbult & Van Lange (2003) Finkel et al. (2017).

Human–Artificial Relations introduce additional research objects: the perceived continuity of an artificial interlocutor, platform-dependent access, interaction history, customization, dependence on assistance, and the conditions under which users attribute commitment or reciprocity. The central question is which observed relational processes are comparable to those studied in human dyads and which depend on substantially different agency, control, institutional design, and ownership arrangements. A person's report of attachment to or investment in an artificial service can be studied directly as a human psychological experience; it does not, by itself, demonstrate the artificial system's subjective commitment or welfare interests Finkel et al. (2017).

The Artificial Era enlarges the scope of comparative relationship psychology by asking how Homo and Artificial participate in distinct configurations of interdependence. Research designs can experimentally vary continuity of memory, access restrictions, responsiveness, and interface changes, then measure users' perceived alternatives, investment, distress, and behavioral choices. Ethical evaluation should examine disclosure, autonomy, privacy, commercial incentives, portability, and the possibility that product governance changes the user's effective options. These are concrete testable proposals within the publisher's broader research horizon, with empirical findings to be established by future work Rusbult & Van Lange (2003) Joel et al. (2020).

Integrated Scientific Interpretation of the Evidence

The strongest cumulative evidence concerns the association of satisfaction, investments, and perceived alternatives with reported commitment. Early theoretical studies identified the distinct constructs, scale-development research supplied operational measures, and meta-analyses documented substantial associations across multiple samples. Prospective studies connected commitment to maintenance behavior and dissolution, while dyadic and daily-experience research complicated simplistic claims about the value of sacrifice. This body of work supports commitment as a meaningful intermediate construct connecting perceived dependence and repeated conduct, subject to context and measurement Rusbult (1980) Rusbult et al. (1998) Le & Agnew (2003) Tran et al. (2019).

Several relations require greater causal resolution. Commitment and cognitive identification may mutually reinforce one another, but their observed associations can reflect shared antecedents. An increased investment may create an incentive to cooperate or a barrier against exit, and the two can coexist. Perceived alternatives can reflect actual opportunity, anticipated satisfaction, or institutional barriers. Similarly, trust can follow visible maintenance acts, but perceived trustworthiness is shaped by previous experiences and changing expectations. The field benefits from maintaining these distinctions in causal models and interpreting correlations according to the studies that generated them Agnew et al. (1998) Wieselquist et al. (1999) Rhoades et al. (2010).

A fully specified explanation of persistence requires a multidimensional outcome vector. At minimum, it should represent the individual's desire to stay, realized continuity, perceived relationship quality, mutuality of commitments, psychological well-being, and safety. Future studies should incorporate both partners' views and constraints while preserving different routes to the same observed outcome. The strength of investment theory lies in its explanatory clarity; its responsible extension lies in measuring what that clarity leaves out, especially unequal access to resources and the distinction between continuation and flourishing Karney & Bradbury (1995) Finkel et al. (2017).

Open Questions and a Cumulative Research Agenda

One research priority is to identify how the three antecedents of commitment change under exogenous shifts in employment, housing, mobility, and family responsibilities. Longitudinal work with suitable causal designs could distinguish whether an altered opportunity set changes commitment directly, indirectly through satisfaction, or by modifying the weight placed on existing investments. Such studies need clear measurement of objective resources as well as perceptions and should attend to time-varying confounding. They should document the extent to which causal effects vary across demographic and social conditions rather than assuming that aggregate correlations apply identically to every relationship Tran et al. (2019) Rhoades et al. (2010).

A second priority is the development of cross-cultural, lifespan, and dyadic measures that retain conceptual comparability. Researchers should test whether investment items cover care work, community connections, shared digital assets, and other resources meaningful to different participants. Multi-informant designs can assess agreement, asymmetric dependence, and the conditions under which one partner's intentions predict the other's behavior. Studies that include participants after separation can avoid restricting conclusions to surviving relationships, while transparent data and analysis practices can improve independent replication Rusbult et al. (1998) Karney & Bradbury (1995).

A third priority concerns competing outcomes: satisfaction, freedom to leave, mutual care, stability, and safety. A future integrative model should predict when commitment strengthens constructive engagement, when it chiefly reflects constraint, and when ending the relationship produces beneficial outcomes for those involved. Research on human–artificial interaction can become part of this agenda by specifying operational indicators of perceived interdependence and platform-created constraints. Across all domains, explanatory precision improves when theories distinguish the psychological experience of commitment from the institutional and relational conditions that make it possible Overall & McNulty (2017) Finkel et al. (2017).

Conclusion

Interdependence, investment, and relationship commitment describe a coherent research domain with three central explanatory levels. Interdependence concerns the structure of outcome dependence between partners; investment concerns resources tied to a continuing association; and commitment concerns a durable orientation toward maintaining that association. The Investment Model explains why satisfaction, alternatives, and investments can jointly account for commitment while remaining conceptually distinct. Its empirical tradition spans the early tests of Rusbult, the Investment Model Scale, meta-analytic syntheses, longitudinal breakup studies, and research on accommodation, sacrifice, and cognitive interdependence Rusbult (1980) Rusbult et al. (1998) Tran et al. (2019).

The accumulated findings support significant and replicable associations while leaving important causal questions open. Commitment predicts aspects of maintenance and persistence, but remaining together does not fully describe the conditions or consequences of the relationship. Material and psychological constraints, power asymmetries, alternative opportunities, mutuality, and the safety of leaving all shape interpretation. Changes in relationship quality may remain difficult to forecast even with substantial contemporary information. The most useful scientific account preserves the separation between explaining continuation and evaluating the well-being of participants Le et al. (2010) Rhoades et al. (2010) Joel et al. (2020).

The larger synthesis situates relational commitment within the changing institutional and technological organization of life. Across human partnerships and emerging human–artificial encounters, interdependence can be investigated through explicit outcomes, options, contributions, agency, and temporal continuity. Durable scientific progress requires validated measures, multiple informants, longitudinal and experimental identification, independent replication, and attention to ethical consequences. These commitments to method permit the field to explain why relationships endure, change, or end while preserving the diversity of human circumstances and the distinctiveness of the Artificial Era Rusbult & Van Lange (2003) Finkel et al. (2017).

References

  • Agnew, C. R., Van Lange, P. A. M., Rusbult, C. E., & Langston, C. A. (1998). Cognitive interdependence: Commitment and the mental representation of close relationships. Journal of Personality and Social Psychology, 74(4), 939–954. https://doi.org/10.1037/0022-3514.74.4.939
  • Drigotas, S. M., & Rusbult, C. E. (1992). Should I stay or should I go? A dependence model of breakups. Journal of Personality and Social Psychology, 62(1), 62–87. https://doi.org/10.1037/0022-3514.62.1.62
  • Dutton, D. G., & Painter, S. (1993). Emotional attachments in abusive relationships: A test of traumatic bonding theory. Violence and Victims, 8(2), 105–120. https://doi.org/10.1891/0886-6708.8.2.105
  • 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
  • Impett, E. A., Gable, S. L., & Peplau, L. A. (2005). Giving up and giving in: The costs and benefits of daily sacrifice in intimate relationships. Journal of Personality and Social Psychology, 89(3), 327–344. https://doi.org/10.1037/0022-3514.89.3.327
  • Joel, S., Eastwick, P. W., Allison, C. J., Arriaga, X. B., Baker, Z. G., Bar-Kalifa, E., Bergeron, S., Birnbaum, G. E., Brock, R. L., Brumbaugh, C. C., Carmichael, C. L., Chen, S., Clarke, J., Cobb, R. J., Coolsen, M. K., Davis, J., de Jong, D. C., Debrot, A., DeHaas, E. C., . . . Wolf, S. (2020). Machine learning uncovers the most robust self-report predictors of relationship quality across 43 longitudinal couples studies. Proceedings of the National Academy of Sciences, 117(32), 19061–19071. https://doi.org/10.1073/pnas.1917036117
  • Joel, S., Impett, E. A., Spielmann, S. S., & MacDonald, G. (2018). How interdependent are stay/leave decisions? On staying in the relationship for the sake of the romantic partner. Journal of Personality and Social Psychology, 115(5), 805–824. https://doi.org/10.1037/pspi0000139
  • Joel, S., MacDonald, G., & Page-Gould, E. (2018). Wanting to stay and wanting to go: Unpacking the content and structure of relationship stay/leave decision processes. Social Psychological and Personality Science, 9(6), 631–644. https://doi.org/10.1177/1948550617722834
  • 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
  • Kelley, H. H., & Thibaut, J. W. (1978). Interpersonal relations: A theory of interdependence. Wiley. https://search.worldcat.org/title/Interpersonal-relations-%3A-a-theory-of-interdependence/oclc/3627845
  • Le, B., & Agnew, C. R. (2003). Commitment and its theorized determinants: A meta-analysis of the investment model. Personal Relationships, 10(1), 37–57. https://doi.org/10.1111/1475-6811.00035
  • Le, B., Dove, N. L., Agnew, C. R., Korn, M. S., & Mutso, A. A. (2010). Predicting nonmarital romantic relationship dissolution: A meta-analytic synthesis. Personal Relationships, 17(3), 377–390. https://doi.org/10.1111/j.1475-6811.2010.01285.x
  • 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
  • Rhoades, G. K., Stanley, S. M., & Markman, H. J. (2010). Should I stay or should I go? Predicting dating relationship stability from four aspects of commitment. Journal of Family Psychology, 24(5), 543–550. https://doi.org/10.1037/a0021008
  • Rusbult, C. E. (1980). Commitment and satisfaction in romantic associations: A test of the investment model. Journal of Experimental Social Psychology, 16(2), 172–186. https://doi.org/10.1016/0022-1031(80)90007-4
  • Rusbult, C. E. (1983). A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology, 45(1), 101–117. https://doi.org/10.1037/0022-3514.45.1.101
  • Rusbult, C. E., & Van Lange, P. A. M. (2003). Interdependence, interaction, and relationships. Annual Review of Psychology, 54, 351–375. https://doi.org/10.1146/annurev.psych.54.101601.145059
  • Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The investment model scale: Measuring commitment level, satisfaction level, quality of alternatives, and investment size. Personal Relationships, 5(4), 357–387. https://doi.org/10.1111/j.1475-6811.1998.tb00177.x
  • Rusbult, C. E., Verette, J., Whitney, G. A., Slovik, L. F., & Lipkus, I. (1991). Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal of Personality and Social Psychology, 60(1), 53–78. https://doi.org/10.1037/0022-3514.60.1.53
  • Stanley, S. M., Rhoades, G. K., & Markman, H. J. (2006). Sliding versus deciding: Inertia and the premarital cohabitation effect. Family Relations, 55(4), 499–509. https://doi.org/10.1111/j.1741-3729.2006.00418.x
  • Tran, P., Judge, M., & Kashima, Y. (2019). Commitment in relationships: An updated meta-analysis of the investment model. Personal Relationships, 26(1), 158–180. https://doi.org/10.1111/pere.12268
  • Van Lange, P. A. M., Rusbult, C. E., Drigotas, S. M., Arriaga, X. B., Witcher, B. S., & Cox, C. L. (1997). Willingness to sacrifice in close relationships. Journal of Personality and Social Psychology, 72(6), 1373–1395. https://doi.org/10.1037/0022-3514.72.6.1373
  • Wieselquist, J., Rusbult, C. E., Foster, C. A., & Agnew, C. R. (1999). Commitment, pro-relationship behavior, and trust in close relationships. Journal of Personality and Social Psychology, 77(5), 942–966. https://doi.org/10.1037/0022-3514.77.5.942