The Measure of Attachment Qualities<\/strong> (MAQ) is a psychometric instrument developed by Charles S. Carver<\/a> in 1997 designed to assess individual differences in adult attachment tendencies through a continuous, multidimensional framework. Departing from earlier categorical typologies that classified individuals strictly into discrete attachment styles, the MAQ conceptualizes attachment through 14 self-report items mapped across four distinct yet interrelated continuous dimensions: Security<\/em>, Avoidance<\/em>, Ambivalence-Worry<\/em>, and Ambivalence-Merger<\/em>. Items are rated on a 4-point Likert-type scale ranging from 1 (indicating strong agreement) to 4 (indicating strong disagreement), or vice versa depending on scoring orientation. Carver derived the instrument by decomposing earlier prototype-based descriptions\u2014most notably those from Hazan and Shaver\u2019s Adult Attachment Questionnaire (AAQ), the Adult Attachment Scale (AAS) by Collins and Read, and the Attachment Style Measure (ASM) by Simpson\u2014into discrete behavioral and affective components. Psychometric evaluations across multiple undergraduate and community cohorts have demonstrated robust internal consistency (Cronbach\u2019s alphas typically ranging from .66 to .78 across subscales), replicated four-factor structural validity via exploratory and confirmatory factor analyses, and meaningful convergent and discriminant validity with generalized distress, coping strategies, interpersonal functioning, and personality inventories. By distinguishing the desire for extreme interpersonal enmeshment (merger) from chronic fears of abandonment and rejection (worry), the MAQ provides a nuanced theoretical and empirical contribution to attachment theory<\/a> and clinical relationship science.<\/p>\n Measure of Attachment Qualities, MAQ, adult attachment, attachment security, attachment avoidance, ambivalence-worry, ambivalence-merger, internal working models, interpersonal relationships, psychometrics, Charles S. Carver<\/p>\n The Measure of Attachment Qualities was authored by Charles S. Carver, Ph.D.<\/strong> (1947\u20132019), a distinguished professor of psychology at the University of Miami<\/a>, Coral Gables, Florida, United States. Carver was an internationally recognized scholar in the areas of personality psychology, self-regulation theory, behavioral medicine, and coping mechanisms. His prolific contributions include the development of widely adopted psychometric instruments such as the COPE Inventory, the Brief COPE, the Life Orientation Test (LOT\/LOT-R), and the Behavioral Inhibition System\/Behavioral Activation System (BIS\/BAS) scales. Inquiries regarding archival access and academic use of Carver\u2019s psychometric resources are maintained through the Department of Psychology at the University of Miami.<\/p>\n The central purpose of the Measure of Attachment Qualities is to evaluate adult romantic and close relational attachment orientations along continuous dimensions rather than forced categorical types. In the formative decades of adult attachment research, researchers heavily relied on single-item paragraph prototypes\u2014such as those popularized by Hazan and Shaver (1987)\u2014which required respondents to endorse one mutually exclusive description: secure, avoidant, or anxious\/ambivalent. Carver (1997) argued that this categorical methodology introduced substantial psychometric limitations. First, forced-choice categorization assumes within-category homogeneity, obscuring meaningful individual differences in the intensity or valence of attachment behaviors. Second, categorical classification prevents researchers from detecting individuals who endorse high or low levels of multiple orientations simultaneously, such as displaying elements of both avoidant detachment and anxious preoccupation. Third, paragraph-length items conflate multiple theoretical propositions (e.g., trust, comfort with intimacy, fear of abandonment, and interpersonal dependency) into a single composite stimulus, making it impossible to determine which specific feature of the prototype drives a participant’s endorsement.<\/p>\n To overcome these methodological shortcomings, the MAQ was designed to unpack and measure these underlying relational qualities independently. Carver\u2019s empirical objective was to create a brief, theoretically grounded, and psychometrically sound instrument that could capture the primary dimensions of relational security, defensive avoidance, and two distinct manifestations of anxious ambivalence: chronic fears of rejection and abandonment (ambivalence-worry) versus an intense, boundary-blurring impulse to dissolve one\u2019s identity within a partner (ambivalence-merger). By isolating these constructs, the MAQ enables researchers and clinicians to test sophisticated mediation and moderation models involving relational processes.<\/p>\n In research settings, the MAQ is widely employed to investigate how attachment qualities influence psychological stress, subjective well-being, cognitive appraisals, emotion regulation, and dyadic communication. It allows investigators to evaluate whether specific interpersonal vulnerabilities differentially predict maladaptive health behaviors, coping styles, or susceptibility to affective disorders. In clinical contexts, the MAQ serves as an efficient diagnostic screener to evaluate clients’ internal working models of self and others. By distinguishing between ambivalence-worry and ambivalence-merger, clinicians can tailor therapeutic interventions: individuals high in ambivalence-worry often require cognitive-behavioral restructuring of hyperactivating strategies, catastrophic fears of rejection, and low self-worth, whereas individuals elevated on ambivalence-merger typically require interventions targeting interpersonal boundaries, self-differentiation, and distress tolerance related to autonomous functioning.<\/p>\n The Measure of Attachment Qualities assesses four primary psychological dimensions that reflect an individual’s chronic cognitive-affective expectations, regulatory strategies, and behavioral patterns within close relationships. These dimensions originate from the dynamic interplay between the attachment behavioral system and an individual’s internal working models<\/a> of self and other.<\/p>\n The Security<\/em> dimension reflects felt security, comfortable interpersonal closeness, and confidence in the accessibility, responsiveness, and benevolence of relational partners. Individuals scoring high in security possess an internalized working model characterized by self-worth, positive expectations of others, and low relational distress. They experience intimacy as a source of emotional restoration and strength, feeling at ease when depending on partners or serving as a source of support. Rather than viewing vulnerability as dangerous, secure individuals use close relationships as a secure base from which to explore the environment and an emotional haven in times of adversity. In the MAQ, security is captured by items assessing the capacity to derive deep comfort, relaxation, and functional strength from close interpersonal bonds.<\/p>\n The Avoidance<\/em> dimension captures a dispositional discomfort with psychological closeness, vulnerability, and mutual dependency. Rooted in an internalized working model of others as untrustworthy, rejecting, or suffocating, individuals high in avoidance activate deactivating strategies. They maintain emotional equilibrium by defensively maximizing cognitive and physical distance from others, suppressing attachment needs, and cultivating compulsive self-reliance. When relational partners attempt to establish deeper intimacy, avoidant individuals experience heightened discomfort, nervousness, or irritation. On the MAQ, avoidance is reflected through items tapping defensive reservations about trusting others, feelings of unease when proximity increases, and preferences for preserving rigid emotional boundaries.<\/p>\n The Ambivalence-Worry<\/em> subscale assesses hyperactivating attachment strategies focused on chronic vigilance, anxiety over relational stability, and persistent fear of abandonment or loss of love. Driven by a negative working model of the self as defective, unlovable, or easily discarded, individuals high on this dimension continuously monitor their partners for subtle cues of emotional withdrawal or diminishing affection. They experience intrusive worries that their partner does not genuinely love them, does not care for them to an equal degree, or will abruptly terminate the relationship. This facet represents the hyper-anxious, distress-laden, and vigilant cognitive component of anxious attachment.<\/p>\n The Ambivalence-Merger<\/em> dimension reflects an intense, intrusive desire for interpersonal enmeshment, absolute proximity, and the fusion of identities. Unlike ambivalence-worry, which centers on fear of loss, ambivalence-merger represents the behavioral and affective drive to eliminate interpersonal separation. Individuals elevated on this dimension experience their own relational boundaries as permeable and seek to engulf or be engulfed by their attachment figure. Because this intense drive for psychological merger frequently overwhelms relationship partners, it often elicits withdrawal, which paradoxically reinforces the anxious individual’s underlying panic. On the MAQ, this construct is represented by items capturing the self-recognized desire to merge completely with others and the awareness that this consuming intensity can alarm or drive away prospective partners.<\/p>\n The Measure of Attachment Qualities is theoretically situated at the intersection of classical attachment theory<\/a>, originally formulated by John Bowlby<\/a> (1969, 1973, 1980) and operationalized in infants by Mary Ainsworth<\/a> (Ainsworth et al., 1978), and adult romantic attachment paradigms pioneered by Cindy Hazan<\/a> and Phillip Shaver (1987). Bowlby postulated that human infants possess an innate evolutionary motivational system designed to maintain proximity to a primary caregiver (the attachment figure) for protection and survival. Repeated interactions with attachment figures become cognitively encoded into durable mental representations\u2014termed “internal working models”\u2014of the self and the social environment.<\/p>\n Ainsworth\u2019s seminal strange situation laboratory experiments delineated three primary attachment patterns: secure, anxious-ambivalent (resistant), and anxious-avoidant. In 1987, Hazan and Shaver demonstrated that these infant-caregiver attachment dynamics naturally generalize to adult romantic partnerships. In adulthood, romantic partners replace primary caregivers as the principal source of safety and comfort. However, adult attachment dynamics operate within mutual, reciprocal relationships, incorporating reciprocal caregiving, reproductive sexuality, and abstract cognitive schemas.<\/p>\n Carver (1997) constructed the MAQ during a period of extensive theoretical refinement within personality and social psychology, as the field actively debated the dimensionality of adult attachment. Kim Bartholomew<\/a> and Leonard M. Horowitz (1991) introduced a four-category model based on the intersection of two orthogonal axes: the model of self (positive vs. negative) and the model of other (positive vs. negative), producing secure, preoccupied, dismissing-avoidant, and fearful-avoidant attachment styles. Simultaneously, Collins and Read (1990) and Simpson (1990) decomposed Hazan and Shaver’s prototypes into multiple continuous dimensions (e.g., Close, Depend, Anxiety).<\/p>\n Carver\u2019s unique theoretical insight was that the classical “anxious-ambivalent” prototype conflated two functionally distinct relational processes: abandonment anxiety<\/em> (worry) and boundary dissolution<\/em> (merger). While both stem from an insecure, hyperactivated attachment system, they manifest via divergent cognitive and behavioral profiles. Fear of abandonment focuses on the anticipated loss of an external source of security, whereas merger reflects an attempt to compensate for fragile self-boundaries by dissolving the self into the other. Furthermore, Carver’s conceptualization aligned with psychometric evidence suggesting that security and avoidance are not simply diametric opposites on a single linear continuum, but separate constructs that can exhibit varying degrees of orthogonal variance depending on context and measurement design.<\/p>\n The psychometric validity of the Measure of Attachment Qualities has been extensively supported across multiple empirical studies conducted by Carver (1997) and independent researchers in personality, health psychology, and clinical science.<\/p>\n In Carver\u2019s (1997) initial four validation studies, scores on the MAQ subscales were systematically compared against traditional attachment classification measures, including Hazan and Shaver\u2019s (1987) Adult Attachment Questionnaire (AAQ), Bartholomew and Horowitz\u2019s (1991) Relationship Questionnaire (RQ), and Collins and Read\u2019s (1990) Adult Attachment Scale (AAS). Consistent with theoretical predictions:<\/p>\n Discriminant validity was established by comparing the MAQ dimensions with broad personality traits, generalized negative affectivity, and neuroticism. While Ambivalence-Worry shares modest variance with trait neuroticism (as measured by the NEO-PI), Ambivalence-Merger and Avoidance demonstrate unique variance unpredicted by generalized distress or negative emotionality. Furthermore, factor analyses combining the MAQ items with general distress scales have consistently confirmed that attachment security, avoidance, and merger load on latent constructs distinct from general anxiety, depressive affect, or self-esteem.<\/p>\n The MAQ exhibits strong predictive validity regarding interpersonal behavior, stress appraisals, and coping responses. Carver (1997) demonstrated that the MAQ subscales meaningfully predict how individuals cope with major stressors:<\/p>\n The internal consistency and temporal stability of the Measure of Attachment Qualities have been confirmed across undergraduate samples, clinical cohorts, and diverse adult populations.<\/p>\n In Carver\u2019s (1997) initial four studies involving over 1,000 participants, the Cronbach\u2019s alpha reliability coefficients for the four subscales consistently met acceptable to good psychometric standards, particularly given the concise nature of the subscales (each containing between 3 and 5 items):<\/p>\n Given the brevity of the scale, these alpha coefficients indicate strong item homogeneity and coherent shared variance within each targeted latent dimension without redundant item phrasing.<\/p>\n Test-retest assessments conducted across intervals ranging from four weeks to six months have demonstrated that MAQ scores exhibit moderate to high temporal stability, consistent with the theoretical premise that adult attachment internal working models operate as enduring cognitive schemas. Test-retest correlation coefficients across an eight-week test-retest window typically range between r<\/em> = .68 and .82<\/strong> across all four subscales. While scores remain stable under normative conditions, research confirms that MAQ scores remain sensitive to major relational transitions, such as relationship dissolution, marriage, or intensive psychotherapy, confirming the scale’s sensitivity to meaningful clinical change<\/span>.<\/p>\n The structural integrity of the Measure of Attachment Qualities was established through extensive exploratory and confirmatory factor analyses reported by Carver (1997).<\/p>\n In Study 1, Carver subjected the 14 candidate items to an exploratory factor analysis using principal-axis factoring with an oblique (promax or oblimin) rotation<\/em>, reflecting the theoretical expectation that attachment dimensions are correlated rather than strictly orthogonal. The analysis yielded a distinct four-factor solution with eigenvalues exceeding 1.0, which accounted for approximately 61% of the total variance<\/strong>. The empirical item loadings conformed cleanly to the hypothesized structure:<\/p>\n To examine broader conceptual configurations, Carver conducted a second-order factor analysis on the correlations among the four primary factors. This higher-order analysis extracted two essentially orthogonal factors (second-order correlation r<\/em> = .03)<\/span>:<\/p>\n This structural finding aligns directly with the two-dimensional model of attachment (Attachment Anxiety and Attachment Avoidance) formalized by Brennan, Clark, and Shaver (1998), while demonstrating the empirical value of preserving lower-order facets.<\/p>\n Carver also performed cluster analyses on the MAQ subscale profiles, evaluating both three-cluster and four-cluster models against Bartholomew\u2019s typology. A three-cluster solution (Secure, Avoidant, and Ambivalent) exhibited the highest statistical stability and clinical clarity. Attempts to differentiate dismissing from fearful avoidance were less clear-cut within the MAQ’s item pool, demonstrating that the instrument’s primary strength lies in delineating the three traditional Ainsworth-derived spaces with enhanced diagnostic differentiation of anxious ambivalence.<\/p>\n The Measure of Attachment Qualities was formally published in 1997<\/strong> in the Journal of Social and Personal Relationships<\/em>. Dr. Charles S. Carver made the instrument freely accessible for educational, non-commercial research, and clinical use without requiring licensing fees or formal administrative permission, provided that appropriate scholarly attribution is maintained. Copyright remains with the original publisher (SAGE Publications) and the estate of Charles S. Carver. Researchers and clinicians intending to integrate the MAQ into commercial diagnostic software, fee-for-service enterprise systems, or commercial publications should obtain formal permission through SAGE Publications or the University of Miami.<\/p>\n Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the strange situation<\/em>. Lawrence Erlbaum Associates.<\/p>\n Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology<\/em>, 61(2), 226\u2013244. https:\/\/doi.org\/10.1037\/0022-3514.61.2.226<\/a><\/p>\n Bowlby, J. (1969\/1982). Attachment and loss: Vol. 1. Attachment<\/em> (2nd ed.). Basic Books.<\/p>\n Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger<\/em>. Basic Books.<\/p>\n Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression<\/em>. Basic Books.<\/p>\n 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<\/em> (pp. 46\u201376). Guilford Press.<\/p>\n Carver, C. S. (1997). Adult attachment and cookie-cutter styles: A measure of attachment qualities with validation from self-report and social science research. Journal of Social and Personal Relationships<\/em>, 14(2), 245\u2013263. https:\/\/doi.org\/10.1177\/0265407597142006<\/a><\/p>\n Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology<\/em>, 58(4), 644\u2013663. https:\/\/doi.org\/10.1037\/0022-3514.58.4.644<\/a><\/p>\n Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology<\/em>, 52(3), 511\u2013524. https:\/\/doi.org\/10.1037\/0022-3514.52.3.511<\/a><\/p>\nKeywords<\/h2>\n
Authors<\/h2>\n
Purpose<\/h2>\n
Psychological Construct<\/h2>\n
1. Security<\/h3>\n
2. Avoidance<\/h3>\n
3. Ambivalence-Worry<\/h3>\n
4. Ambivalence-Merger<\/h3>\n
Theoretical Framework<\/h2>\n
Validity<\/h2>\n
Construct and Convergent Validity<\/h3>\n
\n
Discriminant Validity<\/h3>\n
Predictive and Criterion Validity<\/h3>\n
\n
Reliability<\/h2>\n
Internal Consistency<\/h3>\n
\n
Temporal Stability (Test-Retest Reliability)<\/h3>\n
Factor Analysis<\/h2>\n
Exploratory Factor Analysis (EFA)<\/h3>\n
\n
Second-Order Factor Analysis and Higher-Order Dimensions<\/h3>\n
\n
Cluster Analysis<\/h3>\n
Instrument \/ Measurement Tool<\/h2>\n
\n
\n
\n
Permissions & Fee and Test Year<\/h2>\n
References<\/h2>\n