Trust Scale – Trust within close interpersonal relationships

Abstract

The Trust Scale, developed by John K. Rempel, John G. Holmes, and Mark P. Zanna (1985), is an essential psychometric instrument designed to assess trust within close, adult interpersonal relationships. Grounded in interdependence theory and attributional models of relational dynamics, the instrument operationalizes trust not as an undifferentiated emotional state or generalized personality trait, but as an evolving, multi-tiered dyadic construct. The scale captures three developmentally sequenced dimensions: Predictability, which evaluates the consistency and stability of a partner’s past behaviors based on empirical interaction history; Dependability, which gauges dispositional attributions regarding the partner’s personal integrity, reliability, and benevolent motives in the face of vulnerability; and Faith, representing an emotional conviction that transcends available evidence, reflecting confidence in the partner’s ongoing support, care, and responsiveness into an inherently uncertain future.

Comprising 18 self-report items administered via a 7-point Likert-type scale ranging from -3 (“Strongly Disagree”) through 0 (“Neutral”) to +3 (“Strongly Agree”), the Trust Scale yields both discrete subscale indices and a composite global trust score. Extensive psychometric evaluations demonstrate sound psychometric integrity, with internal consistency coefficients (Cronbach’s alpha) typically ranging from .70 to .85 across subscales and exceeding .85 for the overall measure. Construct, convergent, and discriminant validities are substantiated through robust correlations with dyadic commitment, intrinsic motivation, relationship satisfaction, secure attachment, and partner responsiveness, alongside negative associations with relational ambivalence and chronic vigilance. As one of the most widely cited measures in social, personality, and clinical psychology, the Trust Scale remains foundational for investigating relational stability, risk regulation, betrayal trauma, and couple therapy outcomes.

Keywords

Trust Scale, interpersonal trust, close relationships, Predictability, Dependability, Faith, dyadic interaction, psychometrics, Rempel Holmes and Zanna, interdependence theory, relationship satisfaction, romantic attachment

Authors

The scale was developed and psychometrically validated by:

• John K. Rempel, Ph.D. — Department of Psychology, St. Jerome’s University / University of Waterloo, Waterloo, Ontario, Canada.
• John G. Holmes, Ph.D. — Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada.
• Mark P. Zanna, Ph.D. — Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada.

Subsequent psychometric dissemination, archival hosting, and cross-validation initiatives have also involved Ronald Okada, Professor Emeritus of Psychology, Department of Psychology, Faculty of Health, York University, Toronto, Ontario, Canada (Email: [email protected]).

Purpose

The central purpose of the Trust Scale is to provide an empirically validated, theoretically sophisticated measurement tool for capturing subjective perceptions of trustworthiness and emotional safety in ongoing, committed romantic partnerships and close interpersonal attachments. Prior to its construction in 1985, empirical research into trust was dominated by generalized social trust instruments—such as Julian Rotter’s Interpersonal Trust Scale—which conceptualized trust as a stable, generalized expectancy that others’ verbal statements or written commitments could be believed. While valuable for broad sociological inquiries and stranger interactions, these generalized instruments proved inadequate for understanding the nuances of intimate relationships, where individuals must routinely negotiate mutual interdependence, physical and emotional vulnerability, and asymmetrical risk.

Rempel, Holmes, and Zanna sought to remedy this empirical gap by formulating a domain-specific instrument that addresses the structural evolution of trust over the lifespan of a partnership. Specifically, the scale was engineered to differentiate between the behavioral verifiability of a partner’s actions and the psychological inferences a person draws regarding that partner’s fundamental motivations and emotional commitments. This distinction is critical in both basic psychological research and applied clinical interventions:

• Clinical Assessment and Couple Therapy: The Trust Scale allows clinicians to identify the specific cognitive-affective tier at which dyadic distress occurs. For instance, couples recovering from infidelity or severe relational ruptures may demonstrate intact ratings of partner predictability (e.g., routines and logistic reliability remain known) yet experience total collapse in the Faith dimension. Therapists can use subscale profiles to tailor interventions toward behavioral contracts (rebuilding predictability), characterological consistency (rebuilding dependability), or affective risk-taking (rebuilding faith).
• Basic Relationship Science: Researchers utilize the scale to evaluate how trust regulates interpersonal risk, dyadic communication, conflict resolution styles, and distress-maintaining versus relationship-enhancing attributions. It serves as a cornerstone instrument for testing models of attachment security, vulnerability negotiation, and longitudinal relationship longevity.
• Health and Stress Psychophysiology: In behavioral medicine, the Trust Scale is employed to investigate how intimate trust moderates neuroendocrine responses to social stress, cortisol reactivity, and cardiovascular recovery during conflict discussions.

Psychological Construct

Interpersonal trust within the Rempel et al. framework is conceptualized as an individual’s generalized expectation that their partner will act in ways that support, protect, and enhance their well-being, especially when potential costs or vulnerabilities are high. Rather than treating trust as a static, unitary phenomenon, the authors conceptualized it as a three-dimensional psychological construct that develops progressively across distinct evolutionary strata of relationship maturation.

1. Predictability

Predictability represents the most concrete, empirically based component of trust. It is grounded in the perceptual regularity, consistency, and stability of a partner’s overt behavioral patterns accumulated over time. In early relationship phases, partners observe each other across varied contexts to ascertain whether behaviors conform to identifiable routines or social norms. When predictability is high, an individual experiences situational certainty: they can reliably anticipate how their partner will act from one day to the next. However, predictability alone provides an incomplete foundation for deep intimacy, as it relies heavily on external constraints, established habits, or social scripts rather than the partner’s internalized benevolence. A partner may be entirely predictable out of rigid habit or fear of sanctions without harboring authentic goodwill.

2. Dependability

Dependability shifts the locus of trust from observable behaviors to internal, dispositional qualities. It reflects the individual’s belief that their partner possesses personal integrity, honesty, benevolence, and psychological stability that warrant confidence, particularly in contexts involving interpersonal vulnerability, sacrifice, or risk. While predictability asks, “Do I know what my partner will do?”, dependability asks, “Can my partner be counted on when I need them, even if doing so is costly to them?”. Dependability requires attributional processing: the partner’s trustworthy actions must be attributed to intrinsic prosocial motivation rather than external pressures. High dependability signifies that the partner has proven honest, kept commitments, and demonstrated reliability under conditions of emotional exposure.

3. Faith

Faith represents the apex of relational trust—an emotional conviction and psychological leap that extends beyond empirical evidence or past behavioral data. Given that life continually presents unprecedented dilemmas, developmental transitions, and unpredictable crises, individuals cannot rely solely on historical precedents to guarantee their partner’s future benevolence. Faith is defined as feelings of confidence that the partner will remain supportive, caring, and responsive in the face of an uncertain, unwritten future. It embodies unconditional positive expectations regarding the partner’s devotion, allowing individuals to expose deep emotional vulnerabilities, reveal hidden shame, and face unknown challenges with emotional security. In mature relationships, Faith functions as the central cognitive-emotional buffer against existential anxiety and prospective betrayal.

Theoretical Framework

The construction of the Trust Scale is deeply embedded within two major theoretical paradigms: Interdependence Theory (developed by Harold H. Kelley and John W. Thibaut) and Attribution Theory (originating with Fritz Heider and expanded by Kelley).

Interdependence and the Transformation of Motivation

According to interdependence theory, close relationships are characterized by behavioral and outcome interdependence. In situations of mutual dependence, partners frequently confront situations involving non-correspondent outcomes—scenarios where one partner’s immediate self-interest conflicts with the partner’s or dyad’s well-being. Trust is intrinsically tied to how partners navigate these diagnostic situations. When a partner voluntarily departs from immediate self-interest to enact behaviors that facilitate the other’s welfare, this “transformation of motivation” provides compelling diagnostic evidence of relational commitment. Rempel and colleagues posited that trust cannot emerge in the absence of perceived vulnerability; true trust requires an individual to place vital outcomes in the hands of the partner under conditions where exploitation is possible.

Attributional Progression of Relational Knowledge

The scale integrates attribution theory to explain how experiential data is converted into psychological security over developmental time. In the foundational stage of a relationship, trust is anchored in behavioral consistency (Predictability). Over time, as partners navigate cross-situational challenges involving interpersonal risk, individuals engage in dispositional attribution: they discount situational pressures and attribute benevolent actions to the partner’s enduring moral character and love (Dependability). Finally, through repeated experiences of safety and affirmation, these dispositional attributions crystallize into a generalized emotional schema (Faith) that no longer requires continual evidentiary verification. Faith operates as a proactive cognitive heuristic, coloring ambiguous future scenarios with expectations of safety.

The Risk Regulation System

In modern relational psychometrics, the Rempel et al. construct aligns directly with the Risk Regulation Model developed by Sandra Murray, John Holmes, and Nancy Collins. The model posits that when individuals encounter relational threats, an internal executive system balances two competing goals: the promotion goal of seeking closeness and the protection goal of avoiding rejection. The Trust Scale directly indexes the resting state of this regulatory system. High trust—particularly high Faith—biases the system toward connection, enabling vulnerability and pro-relationship accommodation. Low trust forces the system into hypervigilance, emotional withdrawal, and self-protective defensiveness.

Validity

The empirical validation of the Trust Scale has been substantiated through rigorous psychometric testing across diverse community and student samples, cross-sectional surveys, and longitudinal observational designs.

Construct and Factorial Validity

In the original validation study by Rempel, Holmes, and Zanna (1985), involving couples in dating, cohabiting, and married relationships, construct validity was established through confirmatory factor analysis and structural equation modeling. The tripartite operationalization (Predictability, Dependability, and Faith) demonstrated superior conceptual and statistical fit compared to unidimensional models. Furthermore, developmental hypotheses were confirmed: the correlation between Faith and relationship duration, as well as its predictive weight on overall relationship happiness, was significantly higher in long-term married couples than in early-stage dating couples, supporting the theoretical progression from behavioral predictability to affective faith.

Convergent Validity

Convergent validity is robustly documented through significant positive correlations with gold-standard relationship metrics:

• Love and Affection: High positive associations with Zick Rubin’s Loving Scale ($r \approx .65$ to $.78$) and Hatfield and Sprecher’s Passionate Love Scale.
• Dyadic Commitment and Satisfaction: Strong correlations with the Dyadic Adjustment Scale (DAS) ($r = .55$ to $.70$) and the Investment Model Scale developed by Caryl Rusbult ($r = .60$ to $.75$).
• Attachment Security: Moderate to strong positive correlations with secure adult attachment styles assessed via the Experiences in Close Relationships (ECR) inventory, and strong inverse correlations with attachment anxiety and attachment avoidance.

Discriminant and Predictive Validity

Discriminant validity was established by confirming that the Trust Scale shares only modest variance with generalized interpersonal trust scales ($r < .30$), demonstrating t\hat intimate dyadic trust is distinct from stranger-oriented trust expectancies. Furthermore, the scale exhibits negligible correlations with generalized social desirability response sets (e.g., the Marlowe-Crowne Social Desirability Scale,$r < .15$).

Predictive validity has been demonstrated in longitudinal studies evaluating relationship longevity and conflict behavior. In laboratory conflict paradigms, partners with higher baseline Trust Scale scores—particularly in Dependability and Faith—exhibit significantly fewer negative affective expressions, less physiological arousal, and lower rates of demand-withdraw cycles. Longitudinally, baseline trust scores reliably predict relationship stability, buffering couples against external socioeconomic stressors and mitigating the destabilizing effects of normative relationship transitions, such as the birth of a first child.

Reliability

The Trust Scale demonstrates high psychometric reliability across internal consistency and temporal stability paradigms across multiple cross-cultural replications.

Internal Consistency

In the foundational validation samples (Rempel et al., 1985), internal consistency coefficients (Cronbach’s alpha) were documented as follows:

• Faith Subscale: Consistently yields the highest internal consistency, with $\alpha$ coefficients ranging from .80 to .88 across diverse populations. This reflects the high conceptual cohesion among items tapping future-oriented emotional certainty.
• Dependability Subscale: Demonstrates satisfactory to strong internal consistency, with $\alpha$ coefficients typically falling between .72 and .82.
• Predictability Subscale: Shows adequate reliability, with $\alpha$ coefficients typically reported between .70 and .76. The slightly lower alpha values for this subscale reflect the factual, varied nature of behavioral consistency domains across differing lifestyle contexts.
• Total Scale Score: The composite 18-item index achieves high overall internal reliability, typically yielding $\alpha$ values between .85 and .91.

Test-Retest Reliability

Temporal stability evaluations demonstrate that the Trust Scale is sensitive to true relational transformations while remaining stable in the absence of critical dyadic events. Test-retest reliability over a 4-to-6-week interval in stable couples ranges from $r = .75$ to $.84$ for the total score. Longitudinal studies tracking couples across one to two years confirm moderate-to-high stability ($r = .55$ to $.68$), with fluctuations systematically explained by diagnostic relationship events, such as discovered deceit or major therapeutic breakthroughs.

Factor Analysis

The dimensional architecture of the Trust Scale was originally determined through Exploratory Factor Analysis (EFA) and later corroborated using Confirmatory Factor Analysis (CFA).

Exploratory Factor Analyses

Rempel, Holmes, and Zanna (1985) subjected initial item pools to principal components analysis (PCA) followed by both orthogonal (Varimax) and oblique (Oblimin) rotations. The analyses yielded a clear three-factor solution accounting for substantial total variance:

• Factor 1: Faith — Items tapping feelings of emotional confidence, future responsiveness, and willingness to expose deepest vulnerabilities loaded heavily onto this factor (primary loadings typically ranging from .58 to .81). Cross-loadings on other factors were negligible.
• Factor 2: Dependability — Items measuring dispositional reliability, honesty, commitment keeping, and avoidance of intentional harm loaded cleanly onto this dimension (loadings ranging from .52 to .76).
• Factor 3: Predictability — Items evaluating behavioral stability, consistency of action, and behavioral knowability formed the third discrete component (loadings ranging from .48 to .73).

Confirmatory Factor Analyses and Model Fit

Subsequent psychometric investigations evaluating the latent structure of the scale have compared three competing specifications: (a) a unidimensional model, (b) an orthogonal three-factor model, and (c) an oblique (correlated) three-factor model. The correlated three-factor model consistently produces the best fit to the empirical data:

• Comparative Fit Index (CFI): Values regularly exceed .92 to .95, indicating acceptable to excellent fit.
• Root Mean Square Error of Approximation (RMSEA): Values consistently range between .045 and .068, within standard thresholds for good model fit.
• Standardized Root Mean Square Residual (SRMR): Indices typically remain below .06.

Moderate latent correlations emerge between Dependability and Faith ($r \approx .55$ to $.65$), and between Dependability and Predictability ($r \approx .40$ to $.52$), while the correlation between Predictability and Faith is weaker ($r \approx .25$ to $.35$). This empirical pattern strongly validates the theoretical taxonomy: dependability acts as the conceptual bridge linking early behavioral consistency (predictability) to deep, future-oriented emotional security (faith).

Instrument / Measurement Tool

• Test Name: Trust Scale (Trust within close interpersonal relationships)
• Authors: John K. Rempel, John G. Holmes, and Mark P. Zanna (1985)
• Test Type: Psychometric Self-Report Rating Scale
• Target Population: Adults and adolescents involved in close, ongoing interpersonal or romantic partnerships
• Administration Format: Paper-and-pencil questionnaire or computerized/online self-report
• Completion Time: Approximately 5 to 8 minutes
• Item Count: 18 items
• Response Format: 7-point Likert-type scale scored as:
  • -3: Strongly Disagree
  • -2: Moderately Disagree
  • -1: Mildly Disagree
  •  0: Neutral
  • +1: Mildly Agree
  • +2: Moderately Agree
  • +3: Strongly Agree

  (Note: For computerized or linear scoring purposes, items may alternatively be coded on a standard 1 to 7 continuum where 1 = Strongly Disagree and 7 = Strongly Agree).

• Subscale Breakdown:
  • Predictability (P): Items 3, 6, 9, 11, 17
  • Dependability (D): Items 1, 5, 7, 13, 15
  • Faith (F): Items 2, 4, 8, 10, 12, 14, 16, 18
• Reverse-Scoring Instructions:

  Items 4, 7, 9, 11, 15, 16, and 17 are negatively valenced. Before calculating subscale or global scores, these items must be reverse-coded (e.g., on a -3 to +3 scale, multiply by -1; on a 1 to 7 scale, compute: $\text{New Score} = 8 – \text{Original Score}$).

• Scoring Procedure:

  Calculate mean or summed scores for each subscale after reversing keyed items. The global Trust score is computed by summing the scores of all three subscales or averaging all 18 items. Higher values reflect greater levels of interpersonal trust.

Permissions & Fee and Test Year

The Trust Scale was initially published in 1985 by the American Psychological Association (APA) in the Journal of Personality and Social Psychology. The instrument is considered open for educational, academic, clinical, and non-commercial research purposes without royalty fees, provided proper formal academic citation is granted to the original authors. Researchers seeking to incorporate the instrument into commercial software, fee-for-service diagnostic platforms, or published books should consult APA permissions policies or contact the primary developers (e.g., Dr. John K. Rempel or Dr. Ronald Okada, York University, at [email protected]).

References

• Holmes, J. G., & Rempel, J. K. (1989). Trust in close relationships. In C. Hendrick (Ed.), Review of personality and social psychology: Close relationships (Vol. 10, pp. 187–220). Sage Publications.
• Kelley, H. H., & Thibaut, J. W. (1978). Interpersonal relations: A theory of interdependence. John Wiley & Sons. https://psycnet.apa.org/record/1979-05241-000
• Murray, S. L., Holmes, J. G., & Collins, N. L. (2006). Optimizing assurance: The risk regulation system in relationships. Psychological Bulletin, 132(5), 641–666. https://doi.org/10.1037/0033-2909.132.5.641
• Rempel, J. K., Holmes, J. G., & Zanna, M. P. (1985). Trust in close relationships. Journal of Personality and Social Psychology, 49(1), 95–112. https://doi.org/10.1037/0022-3514.49.1.95
• Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35(4), 651–665. https://doi.org/10.1111/j.1467-6494.1967.tb01454.x
• Simpson, J. A. (2007). Psychological foundations of trust. Current Directions in Psychological Science, 16(5), 264–268. https://doi.org/10.1111/j.1467-8721.2007.00517.x
• 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

13. Items of the Scale (Questionnaire)