Attachment to God Inventory

Abstract

The Attachment to God Inventory (AGI), developed by psychometricians and psychologists of religion Richard Beck and Angie McDonald in 2004, is a premier self-report psychometric instrument designed to assess individual differences in relational attachment dynamics directed toward the divine. Grounded firmly in contemporary attachment theory as formulated by John Bowlby and operationalized in adult relational frameworks by Cindy Hazan, Phillip Shaver, and Kelly Brennan, the AGI adapts the two-dimensional framework of adult romantic attachment—specifically anxiety and avoidance—to religious and spiritual experience. The instrument comprises 28 items balanced across two orthogonal subscales: Anxiety about God’s Love (14 items) and Avoidance of Intimacy with God (14 items). Participants respond via an authentic 7-point Likert scale ranging from 1 (Disagree strongly) to 7 (Agree strongly), with six items on the Avoidance subscale designated for reverse-scoring.

Psychometrically, the AGI demonstrates exceptional factor stability, robust internal consistency, and strong construct, convergent, discriminant, and predictive validity across diverse Christian traditions and broader monotheistic spiritual cohorts. Initial and confirmatory validation studies routinely document Cronbach’s alpha coefficients ranging between .80 and .88 for both the Anxiety and Avoidance dimensions. Confirmatory factor analytic investigations consistently support the hypothesized two-factor oblique and orthogonal structures over unidimensional or tripartite configurations. The scale has become an indispensable measurement tool within clinical psychology, pastoral counseling, spiritual direction, and the empirical psychology of religion, clarifying how internal working models of self and others map onto personal theological representations, psychological distress, existential well-being, and coping mechanisms in the face of suffering.

Keywords

Attachment to God Inventory, AGI, attachment theory, adult attachment, psychology of religion, internal working models, spiritual anxiety, spiritual avoidance, God image, psychometrics

Authors

The Attachment to God Inventory was originated and psychometrically validated by Richard Beck, Ph.D., and Angie McDonald, Ph.D.

• Richard Beck, Ph.D. is Professor and former Chair of the Department of Psychology at Abilene Christian University (ACU) in Abilene, Texas. Dr. Beck is an experimental and social psychologist whose prolific research trajectory spans the intersection of Christian theology, attachment theory, terror management theory, and the psychology of religious belief. Email contact: [email protected].
• Angie McDonald, Ph.D. is a licensed psychologist and academic researcher specializing in developmental psychology, attachment dynamics, marital and family systems, and the empirical measurement of religious functioning. She has served as an academic faculty member at institutions including Abilene Christian University and Palm Beach Atlantic University.

Purpose

The primary purpose of the Attachment to God Inventory is to systematically conceptualize, quantify, and investigate individual differences in how religious believers emotionally and cognitively relate to God. Throughout the early history of the psychology of religion, instruments measuring religious attitudes frequently conflated doctrinal assent, institutional attendance, extrinsic ritual observance, and generalized intrinsic commitment with affective and relational intimacy. Recognizing that an individual may demonstrate intense religious orthodoxy while harboring profound underlying relational ambivalence, mistrust, or fear of abandonment by God, Beck and McDonald (2004) sought to establish a structurally sound, theoretically coherent psychometric measure capable of differentiating the relational dimensions of faith.

From an applied clinical perspective, the AGI serves as a vital diagnostic and therapeutic instrument within spiritually integrated psychotherapy and pastoral care. Spiritual struggles—often manifesting as spiritual dryness, feelings of divine rejection, anger toward God, or compulsive scrupulosity—frequently stem from insecure attachment orientations. By administering the AGI, clinicians can ascertain whether a client’s emotional crisis is characterized predominantly by hyperactivating strategies (e.g., intense preoccupation with divine rejection, chronic fear that one is insufficient for God, obsessive reassurance-seeking) or deactivating strategies (e.g., suppression of emotional reliance on God, self-reliant aloofness, disengagement from prayer during crises). Such differentiation directly informs treatment planning, facilitating the safe therapeutic exploration of the client’s “God image” versus their cognitive “God concept,” thereby fostering affective healing and secure attachment reparative experiences.

In empirical research, the AGI addresses crucial theoretical questions regarding correspondence versus compensation models of religious attachment. Specifically, scholars utilize the scale to evaluate whether an individual’s adult interpersonal attachment style (e.g., romantic attachment measured by the Experiences in Close Relationships [ECR] scale) directly mirrors their attachment to the divine (correspondence hypothesis), or whether individuals with insecure human attachments utilize a benevolent, omnipotent God as a compensatory attachment figure to achieve psychological safety (compensation hypothesis). Furthermore, researchers employ the AGI to predict physiological stress responses, existential well-being, neuroticism, defensive theological fundamentalism, coping efficacy during bereavement, and symptoms of depression and generalized anxiety.

Psychological Construct

The Attachment to God Inventory operationalizes the construct of divine attachment along two continuous, conceptually orthogonal axes: Anxiety about God’s Love and Avoidance of Intimacy with God. These two dimensions reflect the universal internal working models of the self (worthiness of receiving love and protection) and others (accessibility, dependability, and benevolence of the attachment figure), mapped specifically onto the person-deity dynamic.

1. Anxiety about God’s Love (Model of Self)

The Anxiety about God’s Love subscale captures the degree to which an individual experiences chronic apprehension, fear of abandonment, hypervigilance regarding divine favor, and profound insecurity concerning their worthiness before God. This dimension corresponds directly to a negative internal working model of the self. Individuals scoring high on this dimension operate under a pervasive belief that they are fundamentally flawed, inadequate, or incapable of maintaining divine acceptance.

Behaviorally and emotionally, high attachment anxiety toward God manifests as hyperactivating attachment strategies. Sufferers exhibit intense cognitive preoccupation with sin, persistent worry that God will withdraw His presence, emotional lability following minor transgressions, and an insatiable desire for immediate emotional reassurance of divine affection. For instance, endorsement of items such as “I often worry about whether God is pleased with me” and “I crave reassurance that God loves me” typifies this orientation. When God feels distant or prayers appear unanswered, highly anxious individuals do not detach; rather, they escalate emotional protest, experiencing acute distress, jealousy toward other believers perceived as closer to God, and profound existential panic.

2. Avoidance of Intimacy with God (Model of Divine Other)

The Avoidance of Intimacy with God subscale assesses the extent to which an individual systematically minimizes emotional vulnerability, resists relational dependence, and maintains defensive self-reliance in their dynamic with God. This dimension corresponds to a negative internal working model of the attachment figure, wherein God is implicitly or explicitly construed as distant, unresponsive, untrustworthy, or unconcerned with personal emotional needs.

In response to this mental model, individuals employ deactivating attachment strategies designed to suppress attachment needs and avoid the psychological pain of anticipated divine neglect or intrusion. High-scoring individuals demonstrate discomfort with emotional forms of worship or private prayer, maintain a formal, transactional, or purely cognitive theological posture, and actively resist surrendering control to God in moments of personal crisis. This dynamic is illustrated by items such as “I prefer not to lean on God when I am making decisions” and “I am uncomfortable being emotional in my communication with God.” In contrast, low scores reflect comfortable intimacy, high relational trust, ease of vulnerability, and authentic emotional expression in prayer.

Quadratic Attachment Typologies

When crossed orthogonally, these two continuous dimensions generate four distinct attachment quadrants analogized from interpersonal attachment literature:

• Secure Attachment (Low Anxiety, Low Avoidance): Individuals view themselves as inherently loved and valued by God, and they perceive God as reliably accessible, compassionate, and dependable. They utilize God as a “safe haven” during acute distress and a “secure base” for exploratory autonomous living.
• Preoccupied / Anxious Attachment (High Anxiety, Low Avoidance): Individuals deeply desire divine closeness and emotional communion, yet they are paralyzed by chronic dread of divine displeasure, feelings of unworthiness, and fear of abandonment.
• Dismissing Avoidant Attachment (Low Anxiety, High Avoidance): Individuals project defensive independence, dismissing the importance of emotional intimacy with God. They focus on self-sufficiency, intellectualized religious dogma, or moral self-governance, feeling little need for personal prayer or divine comforting.
• Fearful Avoidant Attachment (High Anxiety, High Avoidance): Individuals harbor an intense internal conflict; they deeply crave divine reassurance and communion, yet they distrust God and fear emotional closeness, resulting in ambivalence, profound spiritual alienation, and unresolved grief.

Theoretical Framework

The conceptual infrastructure of the AGI is rooted within classical attachment theory, originally formulated by British psychoanalyst John Bowlby (1969, 1973, 1980) and empirically expanded in child-mother dyadic interactions by Mary Ainsworth (1978). Bowlby posited that human beings possess an innate evolutionary behavioral system designed to maintain proximity to protective, more capable caregivers (attachment figures) to mitigate environmental threats and regulate internal homeostasis.

The Extension to Adult Relational Dynamics

In the late 1980s, Hazan and Shaver (1987) extended Bowlby’s infant-caregiver paradigms into adult romantic functioning, arguing that emotional bonds between romantic partners represent an activation of the same underlying attachment system. Subsequently, Brennan, Clark, and Shaver (1998) synthesized dozens of disparate attachment measures via rigorous factor analysis, proving that all self-report attachment assessments converge on two overarching, fundamental dimensions: Attachment Anxiety (hyperactivation) and Attachment Avoidance (deactivation). These two dimensions map onto Bowlby’s theoretical postulates regarding internal working models of self and others.

Kirkpatrick’s Attachment Paradigm in the Psychology of Religion

The theoretical bridge connecting adult attachment theory directly to monotheistic religious faith was comprehensively articulated by Lee A. Kirkpatrick (1992, 1999, 2005). Kirkpatrick asserted that God functions not merely as a metaphorical target of religious reverence, but meets all four evolutionary criteria of an authentic Bowlby-defined attachment figure:

1. Proximity Maintenance: Believers actively seek closeness to God through prayer, meditation, church attendance, and sacred rituals.
2. Safe Haven: In the presence of acute illness, trauma, loss, or existential dread, believers turn instinctually to God for solace, protection, and psychological soothing.
3. Secure Base: A felt sense of God’s abiding presence, unconditional acceptance, and omnipotent protection empowers the individual to step out into the world, take healthy risks, engage in altruism, and explore their environment.
4. Separation Anxiety / Protest: The perceived absence, silence, or desertion of God evokes profound psychological grief, anxiety, and behavioral protest (e.g., lamentations, persistent entreaty).

The Psychometric Innovation of Beck and McDonald (2004)

Prior to the development of the AGI, early attempts to quantify attachment to God—such as Kirkpatrick’s early adaptation of Hazan and Shaver’s categorical three-paragraph measure, and Rowatt and Kirkpatrick’s (2002) initial multi-item scales—exhibited psychometric limitations. Several early measures suffered from skewed distributions, ceiling effects (wherein most religious respondents categorized themselves as securely attached), and items that blended theological dogma with emotional dynamics. Beck and McDonald (2004) addressed these vulnerabilities by intentionally adapting the psychometrically elite structure of the Experiences in Close Relationships (ECR) inventory (Brennan et al., 1998) specifically to the God-believer relationship. They ensured that items were emotionally evocative, structurally orthogonal, balanced against acquiescence bias, and capable of capturing genuine psychological variation across devout and casual religious populations alike.

Validity

The Attachment to God Inventory has undergone extensive empirical validation across multiple independent studies, confirming its construct, convergent, discriminant, and predictive validity.

1. Construct and Structural Validity

In their seminal scale development studies involving over 1,500 undergraduate and community participants across diverse Christian denominations, Beck and McDonald (2004) established the structural validity of the two-factor model. Exploratory and confirmatory factor analyses supported two independent factors accounting for substantial variance without problematic cross-loadings. The correlation between the Anxiety and Avoidance subscales was minimal to low (typically r = .15 to .25), affirming that they capture distinct, non-redundant behavioral tendencies rather than a single bipolar secure-insecure continuum.

2. Convergent Validity

Convergent validity has been rigorously demonstrated through correlations with interpersonal attachment instruments, God image measures, and generalized psychological health scales:

• Interpersonal Attachment: Supporting the correspondence hypothesis, AGI Anxiety correlates significantly and positively with romantic attachment anxiety on the Experiences in Close Relationships (ECR) inventory (r values typically between .35 and .48, p < .001). Similarly, AGI Avoidance demonstrates robust positive correlations with romantic attachment avoidance on the ECR (r values typically between .30 and .45, p < .001).
• God Concept and Image Measures: AGI Avoidance exhibits strong negative correlations with measures of Loving God images (e.g., Lawrence’s God Image Inventory), spiritual intimacy, and daily devotional engagement (r = -.55 to -.70). AGI Anxiety correlates strongly with feelings of divine wrath, guilt, scrupulosity, and spiritual instability (r = .50 to .65).
• General Well-Being and Distress: AGI Anxiety consistently correlates with higher scores on the Beck Depression Inventory (BDI-II), the Spielberger State-Trait Anxiety Inventory (STAI), and the Neuroticism domain of the NEO Personality Inventory. Conversely, low scores on both AGI dimensions correlate robustly with existential well-being, life satisfaction, and subjective happiness.

3. Discriminant Validity

Discriminant validity is supported by the finding that the AGI does not merely measure social desirability, general religious orthodoxy, or religious fundamentalism. Research by Beck and McDonald (2004) demonstrated that correlations between AGI subscales and the Marlowe-Crowne Social Desirability Scale were negligible (r < .10, non-significant). Furthermore, individuals possessing identical levels of doctrinal orthodoxy (assessed via the Christian Orthodoxy Scale) demonstrated markedly divergent AGI profiles, proving that cognitive agreement with religious tenets is psychometrically dissociable from emotional security in attachment to God.

4. Predictive and Criterion Validity

The AGI demonstrates remarkable predictive validity across multiple life domains. In clinical populations, high AGI Anxiety prospectively predicts chronic grief, post-traumatic stress severity, and spiritual distress during acute medical illness (e.g., cancer survivorship studies). In cross-cultural and cross-confessional studies, AGI Avoidance reliably predicts religious deconversion, decline in private prayer, and disaffiliation from religious communities over longitudinal observation periods.

Reliability

The psychometric evaluation of the Attachment to God Inventory demonstrates exemplary reliability indices across varied demographic, cultural, and theological cohorts.

Internal Consistency

In the initial instrument validation studies conducted by Beck and McDonald (2004), internal consistency estimates (Cronbach’s alpha) for both 14-item subscales were consistently high:

• Anxiety about God’s Love: Baseline validation samples yielded alpha coefficients of α = .87 (Sample 1, N = 247), α = .88 (Sample 2, N = 217), and α = .86 (Sample 3, N = 312).
• Avoidance of Intimacy with God: Across the same validation cohorts, the Avoidance subscale yielded alpha coefficients of α = .86 (Sample 1), α = .85 (Sample 2), and α = .86 (Sample 3).

Subsequent independent studies in diverse populations—including Catholic seminarians, Evangelical undergraduate cohorts, mainline Protestant congregations, Eastern Orthodox believers, and international translated versions (e.g., Korean, Italian, Portuguese, and Spanish adaptations)—have consistently replicated internal consistency values, with Cronbach’s alphas routinely falling between .81 and .90, and McDonald’s omega (ω) coefficients matching or slightly exceeding these parameters (typically ω = .84 to .91).

Temporal Stability (Test-Retest Reliability)

Investigations examining test-retest reliability across multi-week intervals demonstrate that attachment to God represents an enduring, trait-like psychological disposition rather than a transient, state-dependent religious mood. Beck and McDonald reported test-retest stability coefficients across a 4-to-6-week test-retest window of r = .82 for the Anxiety subscale and r = .84 for the Avoidance subscale (p < .001). Longitudinal studies tracking participants across 12-month periods continue to show high stability (coefficients averaging between .68 and .76), while maintaining adequate sensitivity to register genuine therapeutic shifts resulting from prolonged relational psychotherapy or spiritual transformation.

Factor Analysis

The underlying latent dimensional structure of the Attachment to God Inventory has been scrutinized via both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) across numerous empirical studies.

Exploratory Factor Analysis (EFA)

During scale construction, Beck and McDonald (2004) subjected an initial 56-item pool to principal components and principal axis factor extraction followed by Promax (oblique) and Varimax (orthogonal) rotations. Scree test evaluations and eigenvalues clearly established a two-factor solution, accounting for approximately 42% to 46% of total item variance. The two latent dimensions emerged cleanly:

• Factor 1 (Anxiety): Dominated by items measuring intense fear of abandonment, perception of conditional acceptance, envy of others’ divine intimacy, and emotional instability in faith. All 14 retained items demonstrated primary factor pattern loadings exceeding .45 (ranging from .48 to .78) with negligible cross-loadings (< .20) on the second factor.
• Factor 2 (Avoidance): Dominated by items measuring emotional restraint in prayer, reluctance to depend upon God in crisis, and resistance to vulnerability. All 14 retained items demonstrated primary factor pattern loadings exceeding .42 (ranging from .46 to .75), with reverse-scored items exhibiting consistent negative loadings on un-reversed iterations.

Confirmatory Factor Analysis (CFA) and Model Fit

Subsequent confirmatory factor analytic studies have systematically contrasted the hypothesized two-factor model against alternative plausible nested models, such as a unidimensional “global spiritual security” model, an uncorrelated orthogonal model, and a tripartite model segregating dependent self-reliance from emotional avoidance. The two-factor oblique model consistently displays superior goodness-of-fit across standard psychometric benchmarks:

• Root Mean Square Error of Approximation (RMSEA): Values consistently range between .045 and .058 (with 90% confidence intervals well below the .08 threshold for acceptable fit).
• Comparative Fit Index (CFI): Values routinely exceed .92 to .95, indicating excellent fit relative to baseline independence models.
• Tucker-Lewis Index (TLI): Consistently recorded between .91 and .94.
• Standardized Root Mean Square Residual (SRMR): Observed values typically span .048 to .062.

Invariance testing has further confirmed metric and scalar invariance across biological sex, religious affiliation, and developmental age brackets, verifying that the AGI operates with equivalent psychometric precision across diverse sub-populations.

Instrument / Measurement Tool

• Test Type: Self-report psychometric inventory / psychological questionnaire
• Target Population: Adolescents and adults (ages 16+) with a personal belief in or relational dynamic with God (primarily utilized within monotheistic contexts, especially Christian, Jewish, and general theistic backgrounds)
• Administration Format: Paper-and-pencil questionnaire, computerized survey, or secure online clinical assessment
• Administration Time: Approximately 7 to 12 minutes
• Total Item Count: 28 items
• Response Scale: Authentic 7-point Likert scale:
  • 1 = Disagree strongly
  • 2 = Disagree
  • 3 = Disagree slightly
  • 4 = Neutral/Mixed
  • 5 = Agree slightly
  • 6 = Agree
  • 7 = Agree strongly
• Subscales:
  • Anxiety about God’s Love: 14 items (Items 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27)
  • Avoidance of Intimacy with God: 14 items (Items 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28)
• Scoring Rules:
  • Six items on the Avoidance subscale must be reverse-scored prior to computing subscale metrics: Items 4, 8, 18, 22, 26, and 28. For these items, convert values as follows: (1 → 7, 2 → 6, 3 → 5, 4 → 4, 5 → 3, 6 → 2, 7 → 1), or mathematically: Recoded Score = 8 – Original Score.
  • Subscale scores are derived by calculating the mean of the 14 items comprising each subscale (yielding an average score ranging from 1.00 to 7.00 for each dimension), or alternatively by calculating the total composite sum (ranging from 14 to 98 per subscale).
  • Higher scores on the Anxiety subscale denote greater attachment anxiety, hypersensitivity to rejection, and fear of divine abandonment.
  • Higher scores on the Avoidance subscale denote greater defensive detachment, discomfort with intimacy, and emotional self-reliance.
  • Attachment quadrants can be operationalized by dividing scores along sample medians or absolute scale midpoints (score of 4.00): Secure (Low Anxiety, Low Avoidance), Preoccupied (High Anxiety, Low Avoidance), Dismissing (Low Anxiety, High Avoidance), and Fearful (High Anxiety, High Avoidance).

Permissions & Fee and Test Year

The Attachment to God Inventory was originally published in 2004 in the Journal of Psychology and Theology by Dr. Richard Beck and Dr. Angie McDonald. The copyright for the instrument is held by Rosemead School of Psychology, Biola University, and the authors.

The authors have graciously placed the AGI into broad academic circulation for non-commercial, clinical, educational, and empirical research purposes. Researchers and licensed clinicians are permitted to utilize, administer, and reproduce the scale free of charge without formal royalty fees, provided that appropriate scholarly attribution and standard bibliographic citation are maintained in all resulting publications, theses, dissertations, and conference presentations. For commercial use, commercial digital platform distribution, or major funded clinical trials, written permission should be requested directly from the primary author, Dr. Richard Beck, via the Department of Psychology at Abilene Christian University.

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.
• Beck, R. (2006). God as a secure base: Attachment to God and theological exploration. Journal of Psychology and Theology, 34(2), 125–132. https://doi.org/10.1177/009164710603400202
• Beck, R., & McDonald, A. (2004). Attachment to God: The Attachment to God Inventory, tests of working model correspondence, and an exploration of faith group differences. Journal of Psychology and Theology, 32(2), 92–103. https://doi.org/10.1177/009164710403200202
• Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books.
• Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. Basic Books.
• Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression. 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.
• 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
• Kirkpatrick, L. A. (1992). An attachment-theory approach to the psychology of religion. The International Journal for the Psychology of Religion, 2(1), 3–28. https://doi.org/10.1207/s15327582ijpr0201_2
• Kirkpatrick, L. A. (1999). Attachment and religious representations and behavior. In J. Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and clinical applications (pp. 803–822). Guilford Press.
• Kirkpatrick, L. A. (2005). Attachment, evolution, and the psychology of religion. Guilford Press.
• Miner, M. (2009). The Attachment to God Inventory: A test of construct validity in an Australian context. Journal of Psychology and Theology, 37(4), 263–274. https://doi.org/10.1177/009164710903700404
• Rowatt, W. C., & Kirkpatrick, L. A. (2002). Two dimensions of attachment to God: Avoidance and anxiety. Journal for the Scientific Study of Religion, 41(4), 637–651. https://doi.org/10.1111/1468-5906.00143

Items of the Scale