Experiences in Close Relationships Scale

1. Abstract

The Experiences in Close Relationships Scale (ECR) is a benchmark self-report instrument engineered to measure adult attachment orientations across two continuous, orthogonal dimensions: Attachment Anxiety and Attachment Avoidance. Developed by Kelly A. Brennan, Catherine L. Clark, and Phillip R. Shaver (1998), the scale was constructed following an exhaustive factor analysis of virtually all self-report adult romantic attachment items available in the empirical literature at the time. Composed of 36 items—18 measuring anxiety concerning rejection, abandonment, and emotional unresponsiveness, and 18 measuring avoidance of intimacy, self-disclosure, and interpersonal dependency—the ECR uses a 7-point Likert response scale ranging from 1 (“Strongly Disagree”) to 7 (“Strongly Agree”). Extensive psychometric evaluations demonstrate remarkable internal consistency, with Cronbach’s alpha coefficients typically exceeding .90 for both subscales across diverse populations, along with robust test-retest reliability over multi-month intervals. Confirmatory factor analyses across international samples continually confirm its two-factor structure. Furthermore, scores on the ECR demonstrate robust construct, convergent, discriminant, and predictive validity in relation to relationship satisfaction, dyadic communication, psychological distress, emotion regulation, and physiological stress responses. By superseding categorical taxonomies of attachment with a dimensional model, the ECR has fundamentally shaped contemporary relationship science, clinical assessment, and social psychology research.

2. Keywords

Experiences in Close Relationships, ECR, adult attachment, attachment anxiety, attachment avoidance, romantic relationships, psychometrics, relational insecurity, working models, Bartholomew four-category model

3. Authors

The Experiences in Close Relationships Scale was formulated and psychometrically validated by:

• Kelly A. Brennan, Ph.D. — Department of Psychology, State University of New York at Brockport (formerly at the University of California, Davis).
• Catherine L. Clark, Ph.D. — Department of Psychology, University of California, Davis.
• Phillip R. Shaver, Ph.D. — Distinguished Professor Emeritus of Psychology, University of California, Davis. Dr. Shaver is internationally recognized as one of the principal pioneers in adapting developmental attachment theory to adult romantic relationships.

Correspondence regarding the original development of the instrument was historically directed to the Department of Psychology, University of California, Davis, One Shields Avenue, Davis, CA 95616.

4. Purpose

The primary purpose of the Experiences in Close Relationships Scale (ECR) is to provide an empirically grounded, psychometrically sound, and dimensional assessment of adult romantic attachment styles. Prior to the publication of the ECR in 1998, the field of adult attachment was characterized by fragmented measurement traditions. Researchers frequently utilized discrete typologies, such as the three-category typology introduced by Hazan and Shaver (1987) or the four-category taxonomy introduced by Bartholomew and Horowitz (1991). While these early categorical measures offered conceptual breakthroughs, they suffered from significant psychometric limitations: categorical typologies artificially compress nuanced human behavioral patterns into rigid bins, forfeit statistical power, blur intra-individual variability, and exhibit lower reliability coefficients.

Brennan, Clark, and Shaver (1998) addressed this fragmentation by compiling a comprehensive corpus of 323 items drawn from across all known adult attachment questionnaires. Through rigorous exploratory and confirmatory factor analytic techniques on a large sample (N = 1,086), they demonstrated that the collective variance of romantic attachment measurement converges onto two overarching, continuous, and orthogonal dimensions: Attachment Anxiety (model of self) and Attachment Avoidance (model of other). The ECR was purposefully engineered to capture these two fundamental axes with maximum precision and minimal cross-loading items.

In clinical practice, the ECR provides therapists, marital counselors, and clinical psychologists with an objective diagnostic baseline to understand a patient’s internal working models, maladaptive relational dynamics, relational hyperactivating or deactivating strategies, and emotional vulnerabilities. For instance, high anxiety indicates hyper-vigilance toward cues of partner abandonment and chronic reassurance-seeking, whereas high avoidance indicates defensive emotional suppression, compulsive self-reliance, and discomfort with intimacy. In research contexts, the ECR is widely applied across developmental, social, personality, and health psychology to evaluate dyadic conflict, couple satisfaction, parental bonding, neurobiological stress responses, and psychotherapy process outcomes.

5. Psychological Construct

The psychological construct assessed by the ECR is adult romantic attachment, conceptualized within a two-dimensional continuous space. According to developmental and social-cognitive frameworks, early experiences with primary caregivers lead to the internal construction of cognitive-affective representations termed “internal working models.” These models dictate an individual’s expectations, emotional responses, and regulatory behaviors in adult romantic relationships.

Attachment Anxiety (Model of Self)

The Attachment Anxiety subscale measures the degree to which an individual fears rejection, emotional abandonment, and unreciprocated love. Individuals scoring high on this dimension maintain an insecure, negative model of the self. They chronically doubt their inherent lovability and worthiness, perceiving themselves as fragile or fundamentally deficient. Cognitively, this manifests as hyperactivating attachment strategies: relentless monitoring of relationship partners for signs of emotional withdrawal, catastrophizing minor relational conflicts, and persistent rumination. Emotionally, high anxiety generates intense jealousy, emotional volatility, clinging behaviors, and an overwhelming demand for proximity and constant reassurance. Sample items reflecting this dimension include: “I worry about being abandoned” and “I need a lot of reassurance that I am loved by my partner.”

Attachment Avoidance (Model of Other)

The Attachment Avoidance subscale captures the degree to which an individual feels uncomfortable with emotional intimacy, vulnerability, and mutual dependency. Individuals scoring high on this dimension maintain a negative model of others, viewing relational partners as inherently untrustworthy, unsupportive, or intrusive. Cognitively, this leads to deactivating attachment strategies: the systematic suppression of attachment-related thoughts, dismissal of relational distress, minimization of emotional needs, and emotional detachment. High avoidance results in extreme self-reliance, behavioral withdrawal when intimacy escalates, and an unwillingness to disclose authentic feelings or depend on a romantic partner during crises. Illustrative items include: “I prefer not to show a partner how I feel deep down” and “I find it difficult to allow myself to depend on romantic partners.”

The Four-Quadrant Attachment Space

When crossed orthogonally, the anxiety and avoidance dimensions yield four distinct attachment quadrants corresponding to the theoretical quadrants identified by Bartholomew and Horowitz (1991):

• Secure (Low Anxiety, Low Avoidance): Characterized by positive models of self and others. Secure individuals feel worthy of love, trust their partners, balance intimacy with autonomy, and resolve relational conflict constructively.
• Preoccupied / Anxious (High Anxiety, Low Avoidance): Characterized by a negative model of self and a positive model of others. These individuals intensely desire closeness and complete emotional merging but are plagued by persistent anxieties regarding abandonment.
• Dismissing-Avoidant (Low Anxiety, High Avoidance): Characterized by a positive model of self and a negative model of others. These individuals maintain defensive high self-esteem, reject emotional dependency, minimize the importance of close relationships, and assert absolute self-sufficiency.
• Fearful-Avoidant (High Anxiety, High Avoidance): Characterized by negative models of both self and others. Fearful individuals crave relational warmth and validation but simultaneously distrust others and experience profound fear of vulnerability, resulting in social and emotional paralysis.

6. Theoretical Framework

The ECR is grounded in John Bowlby‘s original attachment theory and Mary Ainsworth‘s observational studies of infant-caregiver dynamics. Bowlby postulated that human infants possess an evolutionarily evolved behavioral system designed to maintain proximity to primary caregivers (attachment figures) for safety and survival. When caregivers provide sensitive, reliable responsiveness, the child establishes a secure base from which to explore the environment and develops resilient internal working models of self and others.

In the late 1980s, Cindy Hazan and Phillip Shaver extended Bowlby’s framework into adult romantic life. They argued that the emotional bonds formed between adult romantic partners are functionally equivalent to infant-caregiver attachment bonds: romantic partners seek proximity, find emotional comfort in each other’s presence (safe haven), and utilize the partner as a base of support to engage with the broader world (secure base).

Subsequent work by Kim Bartholomew and Leonard M. Horowitz (1991) formalized the transition from three discrete styles into a 2 × 2 matrix based on the intersection of self-concept (anxiety/dependence) and other-concept (avoidance/intimacy). Brennan, Clark, and Shaver (1998) refined this theoretical foundation empirically. Rather than imposing an a priori structure, they applied factor analysis to an exhaustive collection of extant attachment scales. Their results revealed that individual differences in adult romantic attachment are best conceptualized as continuous positions along two core axes: Attachment Anxiety (regulating the vigilance and sensitivity of the attachment behavioral system) and Attachment Avoidance (regulating the behavioral proximity-seeking versus proximity-avoiding actions). The ECR thus operationalizes modern attachment theory as a continuous parameter space governed by dual regulatory mechanisms.

7. Validity

The construct, convergent, discriminant, and predictive validity of the Experiences in Close Relationships Scale have been demonstrated in hundreds of empirical investigations across international contexts.

Construct and Factorial Validity

During its initial construction, Brennan et al. (1998) verified that the 36 selected items exhibited high factor loadings on their intended dimensions and negligible cross-loadings. The two factors were found to be largely orthogonal, sharing minimal common variance (correlations typically range from r = .05 to r = .18 across normative community and university cohorts). Confirmatory factor analysis (CFA) models consistently demonstrate acceptable-to-superior goodness of fit across various non-clinical and clinical populations (CFI > .90, RMSEA < .06).

Convergent and Discriminant Validity

Convergent validity is supported by strong correlations with earlier attachment measures. The Anxiety dimension correlates robustly with Hazan and Shaver’s Anxious/Ambivalent category and Bartholomew’s Preoccupied scale (r > .60). The Avoidance dimension correlates heavily with Hazan and Shaver’s Avoidant style and Bartholomew’s Dismissing and Fearful scales (r > .65). In terms of the Five-Factor Model of personality, attachment anxiety demonstrates moderate positive correlations with Neuroticism (r ≈ .35–.45), while attachment avoidance correlates negatively with Agreeableness and Extraversion. Importantly, discriminant validity is demonstrated by weak correlations with social desirability scales (typically r < .15), proving that responses do not merely mirror socially conforming self-presentation.

Predictive and Ecological Validity

The predictive validity of the ECR is exceptionally robust. In longitudinal studies of couples, individuals with high Avoidance report lower relationship satisfaction, fewer supportive behaviors, reduced dyadic communication during conflict, and lower levels of commitment. Individuals with high Anxiety exhibit heightened affective distress during relationship transitions, amplified physiological stress responses (e.g., elevated salivary cortisol levels during marital conflict), and higher susceptibility to intrusive jealousy and chronic reassurance-seeking behaviors (Mikulincer & Shaver, 2016).

8. Reliability

The ECR possesses outstanding internal consistency and temporal stability. In the original derivation study by Brennan et al. (1998) involving 1,086 undergraduate students, internal consistency coefficients were remarkably high:

• Attachment Anxiety: Cronbach’s alpha (α) = .91
• Attachment Avoidance: Cronbach’s alpha (α) = .94

Subsequent psychometric replications across diverse multinational demographic cohorts, including clinical outpatients, marital dyads, and non-student adult community samples, report Cronbach’s alpha values typically ranging between .88 and .94 for Anxiety, and between .90 and .95 for Avoidance. Coefficient omega (ω) estimates similarly confirm high composite reliability.

Test-retest reliability assessments further substantiate the temporal stability of the instrument over varying durations. Sibley, Fischer, and Liu (2005) observed test-retest correlations of r = .70 for Avoidance and r = .68 for Anxiety across a 6-week interval. Over multi-year intervals, test-retest coefficients remain moderate to high (ranging from .50 to .70), mirroring theoretical expectations that adult attachment models are stable cognitive-affective architectures that nevertheless remain somewhat open to revision following major life transitions or corrective relational experiences.

9. Factor Analysis

The historical construction of the ECR represents one of the most comprehensive exploratory and confirmatory factor analytic undertakings in social psychometrics. Brennan et al. (1998) pooled 323 non-redundant items derived from 14 established romantic attachment scales—including measures developed by Hazan and Shaver, Bartholomew and Horowitz, Collins and Read, Feeney and Noller, and Simpson. These items were administered to 1,086 participants.

An exploratory factor analysis using principal components analysis followed by oblique and orthogonal (Varimax) rotations unequivocally revealed a two-factor structure accounting for the vast majority of the reliable variance. Items were systematically evaluated based on their primary loadings, conceptual clarity, and lack of cross-loadings. The researchers selected the 18 items with the highest unique loadings on the Anxiety factor and the 18 items with the highest unique loadings on the Avoidance factor.

Across validation studies using Confirmatory Factor Analysis (CFA):

• Standardized factor loadings for items on the Anxiety dimension typically range from .55 to .84.
• Standardized factor loadings for items on the Avoidance dimension typically range from .52 to .86.
• Model fit indices regularly meet standard empirical criteria for well-fitting psychometric models, typically yielding Comparative Fit Index (CFI) values between .91 and .95, Tucker-Lewis Index (TLI) values between .90 and .94, and Root Mean Square Error of Approximation (RMSEA) values between .045 and .062.

While minor sub-facets exist within each dimension (such as fear of abandonment versus need for reassurance within Anxiety, and avoidance of intimacy versus self-reliance within Avoidance), hierarchical and bi-factor models confirm that a two-dimensional solution represents the most parsimonious and psychometrically stable representation of the data.

10. Instrument / Measurement Tool

• Instrument Name: Experiences in Close Relationships Scale (ECR)
• Target Population: Adults (aged 18+) who have current or past romantic relationship experience.
• Scale Structure: 36 items divided into two subscales:
  • Attachment Avoidance: 18 items (Items 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31, 33, 35)
  • Attachment Anxiety: 18 items (Items 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36)
• Response Format: 7-point Likert scale:
  • 1 = Strongly Disagree
  • 2 = Disagree
  • 3 = Slightly Disagree
  • 4 = Neutral
  • 5 = Slightly Agree
  • 6 = Agree
  • 7 = Strongly Agree
• Reverse-Scored Items: Items 3, 15, 19, 22, 25, 27, 29, 31, 33, and 35. For these items, compute: Reversed Score = 8 − Original Score (e.g., 1 becomes 7, 2 becomes 6, etc.). Notice that item 22 is an Anxiety item, while items 3, 15, 19, 25, 27, 29, 31, 33, and 35 belong to the Avoidance subscale.
• Subscale Scoring:
  • Avoidance Score: Average the scores of the 18 Avoidance items (after reverse scoring items 3, 15, 19, 25, 27, 29, 31, 33, 35). Range: 1.00 to 7.00.
  • Anxiety Score: Average the scores of the 18 Anxiety items (after reverse scoring item 22). Range: 1.00 to 7.00.
• Interpretation: Higher scores reflect greater attachment insecurity along that respective axis. Individuals can be plotted on a 2D Cartesian coordinate plane or categorized into Bartholomew’s 4 quadrants based on sample median splits or midpoint splits (midpoint = 4.0).

11. Permissions & Fee and Test Year

The Experiences in Close Relationships Scale was formally published in 1998. In alignment with the ethical norms of scientific open access within personality and social psychology, the authors placed the ECR in the public domain for academic, educational, and clinical research purposes. No licensing fees, commercial royalties, or formal administrative permissions are required to use the instrument for non-commercial scientific research, academic dissertations, or clinical therapeutic assessment. Clinicians and researchers must appropriately cite the foundational publication (Brennan, Clark, & Shaver, 1998) in all resulting work and empirical reports.

12. References

• Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation. Lawrence Erlbaum Associates.
• Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61(2), 226–244. https://doi.org/10.1037/0022-3514.61.2.226
• Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). Basic Books.
• Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46–76). Guilford Press.
• Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis of self-report measures of adult attachment. Journal of Personality and Social Psychology, 78(2), 350–365. https://doi.org/10.1037/0022-3514.78.2.350
• Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52(3), 511–524. https://doi.org/10.1037/0022-3514.52.3.511
• Mikulincer, M., & Shaver, P. R. (2016). Attachment in adulthood: Structure, dynamics, and change (2nd ed.). Guilford Press.
• Sibley, C. G., Fischer, R., & Liu, J. H. (2005). Reliability and validity of the revised Experiences in Close Relationships (ECR-R) self-report measure of adult romantic attachment. Personality and Social Psychology Bulletin, 31(11), 1524–1536. https://doi.org/10.1177/0146167205276865

13. Items of the Scale