Loving and Controlling God Scales

1. Abstract

The Loving and Controlling God Scales are foundational psychometric instruments developed by Peter L. Benson and Bernard Spilka in 1973 to assess multidimensional representations of the divine. Operationalized through the lens of semantic differential methodology, the instrument evaluates two primary phenomenological dimensions of religious cognition: a Loving God image (characterized by acceptance, benevolent salvation, affective warmth, and forgiveness) and a Controlling God image (characterized by restrictiveness, divine demand, rigid regulation, and disciplinary oversight). The complete measurement tool consists of 10 bipolar adjective pairs scored on a 7-point continuum ranging from 0 to 6, with five items allocated to the Loving God subscale and five items dedicated to the Controlling God subscale. Summed scale scores for each dimension range from 0 to 30, with established reverse-scoring protocols to preserve dimensional integrity. Initial psychometric evaluations across adolescent and early adulthood cohorts demonstrated moderate internal consistency, yielding Cronbach’s alpha / homogeneity coefficients of .72 for the Loving God dimension and .60 for the Controlling God dimension among denominations such as Lutherans and Roman Catholics. Extensive construct, convergent, and discriminant validity studies over five decades have validated the utility of these scales across clinical psychology, developmental psychology, and the sociology of religion. Research indicates that the Loving God scale positively predicts global self-esteem, secure attachment patterns, mature object relations development, and adaptive coping in the face of acute bereavement. Conversely, the Controlling God scale correlates with authoritarian belief structures, restrictive parental discipline perceptions, and clinical distress in dissociative populations. This paper provides an exhaustive review of the psychometric architecture, theoretical foundations, factor analytic stability, scoring paradigms, and empirical applications of the Loving and Controlling God Scales.

2. Keywords

Loving and Controlling God Scales, God image, psychology of religion, semantic differential, divine representation, Peter L. Benson, Bernard Spilka, object relations, attachment theory, religious coping

3. Authors

The Loving and Controlling God Scales were developed and introduced to the scientific literature by:

• Peter L. Benson, Ph.D. — Late Senior Fellow and former President of Search Institute, Minneapolis, Minnesota, United States. A pioneering developmental psychologist renowned for his groundbreaking theoretical and empirical formulations regarding adolescent developmental assets, religious development, and positive youth empowerment.
• Bernard Spilka, Ph.D. — Professor Emeritus of Psychology at the University of Denver, Denver, Colorado, United States. A foundational figure in the empirical psychology of religion, co-author of definitive academic treatises including The Psychology of Religion: An Empirical Approach, and a prolific contributor to the measurement of personal faith, religious attributions, and mystical experience.

Correspondence regarding the foundational validation studies was historically directed through the Department of Psychology at the University of Denver, Denver, CO 80208, and the archival literature published in the Journal for the Scientific Study of Religion.

4. Purpose

The overarching purpose of the Loving and Controlling God Scales is to provide a brief, psychometrically sound, self-administered measurement framework capable of capturing individuals’ emotional and relational representations of God, distinct from abstract intellectual doctrines or theological dogmas. Historically, empirical research within the social scientific study of religion suffered from an overreliance on crude, single-item demographic markers, such as institutional religious affiliation or frequency of attendance at worship services. Benson and Spilka recognized that formal denominational affiliation fails to account for the idiosyncratic, internal phenomenological experience of divine authority, benevolent intimacy, and moral regulation that governs human cognitive and affective functioning.

In clinical psychology and psychotherapy, the scales serve as a potent diagnostic and exploratory tool for assessing clients’ internal relational templates. An individual’s internal representation of God often functions as a psychological mirror reflecting formative interpersonal relationships, internalized parental representations, and emotional security. In clients presenting with moral scrupulosity, guilt-laden neurotic depression, or severe dissociative phenomena (such as dissociative identity disorder), the scales illuminate whether the client perceives a punitive, hyper-controlling transcendent overseer or a supportive, unconditional source of existential solace. Clinical practitioners utilize this profile to identify maladaptive spiritual distress, spiritual struggles, or fundamental developmental arrests within internal object relations.

Within empirical research, the scales provide a standard operationalization for testing major theoretical paradigms in developmental psychology, personality psychology, and social psychology. Specifically, researchers utilize the Loving and Controlling God Scales to:

• Investigate the transmission of parental socialization styles (e.g., authoritative versus authoritarian parenting) into divine mental models across childhood and adolescence.
• Examine the compensatory versus correspondence hypotheses formulated within religious attachment theory, assessing whether insecure interpersonal attachment yields an idealized loving deity or an alienating, controlling deity.
• Determine the psychological mechanisms governing coping strategies during acute stressors, including severe medical illness, terminal bereavement, interpersonal trauma, and institutional incarceration.
• Map the developmental trajectory of ego functioning and object relations maturity from early adolescence to late adulthood.

5. Psychological Construct

The operational framework of the scales delineates between two distinct, non-mutually exclusive dimensions of the divine representation: the Loving God Image and the Controlling God Image. Psychologists of religion differentiate between a God concept (the intellectual, theological, and doctrinal understanding of God learned through cultural instruction) and a God image (the experiential, emotional, dynamic, and relational mental representation of God). Benson and Spilka’s instrument is specifically calibrated to evaluate the subjective God image.

The Loving God Dimension

The Loving God construct captures an affective schemas dominated by benevolence, warmth, emotional accessibility, and soteriological mercy. A high score on this dimension reflects an internalized divine object that provides unconditional positive regard, emotional solace, and profound personal validation. Within this construct, the divine is experienced as actively engaged in the individual’s psychological well-being, ready to forgive moral infractions and alleviate existential alienation. The five adjective dyads characterizing this factor are:

• Rejecting vs. Accepting: Assesses the degree to which God is experienced as welcoming and validating versus dismissive and alienating.
• Loving vs. Hating: Taps the foundational affective tone of the relationship, ranging from profound care to perceived divine malice or hatred.
• Damning vs. Saving: Evaluates the soteriological and moral stance of the deity, contrasting ultimate condemnation with protective redemption.
• Unforgiving vs. Forgiving: Measures the accessibility of divine grace and relational reconciliation following personal failure or transgressions.
• Approving vs. Disapproving: Evaluates the degree of generalized positive appraisal and psychological affirmation attributed to the transcendent.

The Controlling God Dimension

The Controlling God construct operationalizes the perceived sovereign governance, restrictiveness, authoritarian oversight, and behavioral regulation exercised by the divine. Rather than measuring a benign protector, this dimension taps the extent to which the individual feels monitored, constrained, and held to exacting ethical or behavioral standards. A high score denotes an image of God that dominates individual autonomy, demands rigorous compliance, and limits human volition. The five adjective dyads structuring this dimension include:

• Demanding vs. Not Demanding: Measures the perceived magnitude of behavioral expectations, ritual obligations, and moral imperatives imposed by God.
• Freeing vs. Restricting: Evaluates whether divine intervention is felt as liberating psychological agency or confining behavioral and cognitive choices.
• Controlling vs. Uncontrolling: Directly queries the perceived omnipresence of divine micro-management versus perceived human autonomy in the cosmos.
• Strict vs. Lenient: Reflects the unyielding rigidity of divine justice versus flexibility, tolerance, and lenience regarding rules.
• Permissive vs. Rigid: Assesses the structural elasticity of divine governance, contrasting structural flexibility with dogmatic, unbending behavioral prescriptions.

6. Theoretical Framework

The theoretical foundations of the Loving and Controlling God Scales integrate principles from symbolic interactionism, social learning theory, psychodynamic object relations theory, and semantic differential theory.

Semantic Differential and Cognitive Attribution

Benson and Spilka constructed their tool upon the psycholinguistic foundations established by Charles E. Osgood, George Suci, and Percy Tannenbaum (1957) in their measurement of meaning. Osgood’s research identified three universal dimensions of affective meaning: Evaluation (good/bad), Potency (strong/weak), and Activity (active/passive). The Loving God subscale aligns fundamentally with the Evaluation dimension, capturing emotional valence, warmth, and existential safety. Conversely, the Controlling God subscale captures elements of Potency and structural constraint, quantifying power dynamics, sovereign dominance, and perceived behavioral constriction.

Object Relations Theory and Attachment Dynamics

The psychodynamic underpinnings of the scale are rooted in object relations theories formulated by W. R. D. Fairbairn, Donald Winnicott, and Ana-María Rizzuto. Rizzuto’s seminal 1979 treatise, The Birth of the Living God, posited that the God image is an internal transitional object synthesized from early parental introjects, specifically constructed out of the child’s primary relational experiences with attachment figures. A child reared by supportive, responsive, and affirming caretakers internalizes a benevolent object representation that is subsequently projected onto the divine, manifesting as a high Loving God orientation. Conversely, harsh, authoritarian, intrusive, or punitive parenting yields internalized representations characterized by fear, restriction, and hyper-vigilance, manifesting as an elevated Controlling God orientation.

In complementary fashion, contemporary attachment theorists (e.g., Lee A. Kirkpatrick and Phillip R. Shaver) conceptualized God as a primary attachment figure capable of serving as a safe haven and secure base. Within this paradigm, the Loving God dimension corresponds directly to secure attachment representations, while the Controlling God dimension captures elements of anxious-ambivalent or fearful-avoidant attachment configurations wherein the attachment figure is perceived as unpredictably demanding, hyper-critical, or suffocating of personal autonomy.

7. Validity

Substantial empirical validity has been established for the Loving and Controlling God Scales across half a century of quantitative and clinical investigations.

Construct and Convergent Validity

In their initial validation study, Benson and Spilka (1973) administered the instrument alongside measures of self-esteem (using the Coopersmith Self-Esteem Inventory) and generalized expectancy for internal versus external control (using Julian Rotter’s Locus of Control Scale). As theoretically hypothesized, self-esteem demonstrated a statistically significant positive relationship with the Loving God scale; individuals holding high self-regard conceptualized God as accepting, forgiving, and loving. In contrast, self-esteem was unrelated or negatively linked to an overly punitive representation.

Further convergent validity was established in developmental and psychoanalytic research by Brokaw and Edwards (1994). Investigating the relationship between divine representations and the maturation of internal object relations among Christian adults, they employed the Ego Function Assessment Questionnaire-Revised (EFAQ-R). Their findings revealed that the Loving God scale was significantly positively correlated with advanced object relations development, characterized by structural integration, empathy, and affective stability. In sharp contrast, the Controlling God scale was significantly negatively correlated with mature object relations development, confirming that a hyper-controlling divine image corresponds to primitive developmental levels dominated by splitting, projective identification, and fear of retaliatory abandonment.

Discriminant and Predictive Validity

Discriminant validity was evidenced in Benson and Spilka’s (1973) findings regarding locus of control: general internal-external locus of control was empirically distinct from and unrelated to the Controlling God scale, demonstrating that personal attributional style regarding mundane events does not conflate with perceived divine cosmological control. Subsequent research by Gabbard, Howard, and Tageson (1986) evaluated religious locus of control, demonstrating that the scales discriminate between active collaborative control and passive deferring control.

In health psychology and crisis coping, Park and Cohen (1993) administered the scales to college students experiencing the sudden death of a close peer. The Loving God scale predicted positive religious coping, personal growth, and cognitive resolution of grief, whereas the Controlling God scale was linked to perceived divine abandonment, unresolved negative affect, and elevated depressive symptomatology. Similarly, Bowman et al. (1987) observed elevated, unbalanced Controlling God profiles among patients suffering from multiple personalities (dissociative identity disorder), reflecting internalized persecutory altars and punitive childhood developmental environments.

8. Reliability

The Loving and Controlling God Scales have exhibited acceptable to good psychometric reliability across diverse empirical cohorts, particularly when considering the brevity of each 5-item scale.

Internal Consistency

In the original investigation by Benson and Spilka (1973), scale homogeneity was examined within a sample of 50 Lutheran participants using inter-item correlation matrices and standardized split-half/homogeneity estimation methods:

• Loving God Scale: Yielded an initial homogeneity coefficient of α = .72, reflecting coherent conceptual unity among the five affective descriptors.
• Controlling God Scale: Yielded an initial homogeneity coefficient of α = .60, a lower but psychometrically defensible index given the complex multidimensionality of autonomy versus constraint.

Subsequent investigations across larger, more demographically varied samples have observed elevated internal consistency coefficients. In studies of religious college students and adult community samples (e.g., Brokaw & Edwards, 1994; Hertel & Donahue, 1995), Cronbach’s alpha coefficients for the Loving God subscale routinely range between .78 and .86. The Controlling God subscale typically exhibits internal consistency coefficients ranging between .68 and .76. The slightly lower alpha for the Controlling God dimension is widely recognized in psychometric literature as an artifact of its bipolar pairs addressing both active sovereign interference (e.g., controlling, demanding) and boundary rigidity (e.g., strict, rigid).

Temporal Stability

Test-retest reliability assessments conducted across intervals ranging from 4 to 8 weeks in non-crisis adult populations have reported stability coefficients between r = .74 and r = .83 for the Loving God scale, and between r = .67 and r = .79 for the Controlling God scale. These metrics indicate that while the God image remains relatively stable as an internal representational trait, it retains sufficient sensitivity to register acute psychotherapeutic, spiritual, or existential interventions.

9. Factor Analysis

The structural dimensionality of the 10 adjective items comprising the Loving and Controlling God Scales has been rigorously scrutinized using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA).

Exploratory Factor Structure

Initial principal components analysis with varimax orthogonal rotation conducted on the original normative data yielded a distinct two-factor solution explaining the majority of total scale variance:

• Factor 1: Benevolent / Loving God: Accounted for approximately 38% to 45% of the total variance across diverse samples. The five items comprising this dimension loaded robustly onto this component, with factor loadings consistently exceeding .60 (e.g., Loving-Hating loading ≥ .75; Rejecting-Accepting loading ≥ .71; Forgiving-Unforgiving loading ≥ .68). Secondary cross-loadings onto the second factor remained minimal (typically ≤ .20).
• Factor 2: Authoritarian / Controlling God: Accounted for approximately 14% to 20% of the total variance. The five items assigned to this construct exhibited strong, clean loadings onto this axis (e.g., Controlling-Uncontrolling loading ≥ .70; Strict-Lenient loading ≥ .65; Demanding-Not Demanding loading ≥ .58).

Confirmatory Factor Modeling

Subsequent structural evaluations in contemporary psychometrics utilizing structural equation modeling (SEM) have tested whether the two dimensions represent independent factors, correlated factors, or polar ends of a single continuum. The empirical data decisively favor an oblique, two-factor correlated model over a single-factor unipolar model. Fit indices for the correlated two-factor model routinely meet acceptable psychometric thresholds:

• Comparative Fit Index (CFI) ≥ .92 to .95
• Tucker-Lewis Index (TLI) ≥ .90 to .93
• Root Mean Square Error of Approximation (RMSEA) ≤ .055 to .072
• Standardized Root Mean Square Residual (SRMR) ≤ .048 to .060

The correlation between the two latent factors is typically slightly negative or orthogonal depending on denominational background (ranging from r = -.12 to r = -.35), demonstrating that perceiving God as loving does not mechanically preclude perceiving God as controlling, an insight of critical importance in conservative theological environments where divine sovereignty and divine love are concurrently celebrated.

10. Instrument / Measurement Tool

The Loving and Controlling God Scales represent a self-administered, paper-and-pencil or computerized semantic differential assessment tool. Below are the structural specifications:

• Assessment Type: Self-report psychometric rating scale based on semantic differential bipolar adjective dyads.
• Target Population: Adolescents (ages 13+) and adults across clinical, collegiate, and community settings.
• Total Item Count: 10 items (5 items for the Loving God scale; 5 items for the Controlling God scale).
• Administration Time: Approximately 3 to 5 minutes.
• Response Continuum: 7-point semantic differential scale anchored from 0 to 6 between opposing adjective anchors.
• Subscales:
  • Loving God Subscale: Items 1, 2, 4, 6, 8
  • Controlling God Subscale: Items 3, 5, 7, 9, 10
• Scoring and Directionality Rules:
  • Unreversed Loving Items (Items 1, 2, 6): Scored directly (0 = lowest loving valence, 6 = highest loving valence). Adjective poles are positioned such that the high score represents divine love (e.g., Damning [0] to Saving [6]; Rejecting [0] to Accepting [6]; Unforgiving [0] to Forgiving [6]).
  • Reverse-Scored Loving Items (Items 4, 8): Must be inverted prior to subscale summation: Scored Value = 6 − Raw Value (Loving [0] to Hating [6] becomes Hating [0] to Loving [6]; Approving [0] to Disapproving [6] becomes Disapproving [0] to Approving [6]).
  • Unreversed Controlling Items (Items 5, 10): Scored directly: Freeing [0] to Restricting [6]; Permissive [0] to Rigid [6]. Higher values denote greater perceived restriction/rigidity.
  • Reverse-Scored Controlling Items (Items 3, 7, 9): Must be inverted prior to subscale summation: Scored Value = 6 − Raw Value (Demanding [0] to Not Demanding [6]; Controlling [0] to Uncontrolling [6]; Strict [0] to Lenient [6]).
• Total Score Range: 0 to 30 for the Loving God subscale; 0 to 30 for the Controlling God subscale. Higher composite scores signify greater intensity of that particular divine representation.

11. Permissions & Fee and Test Year

Publication Year: 1973.

Copyright and Accessibility: The Loving and Controlling God Scales were original research instruments published in an academic journal article by Peter L. Benson and Bernard Spilka in the Journal for the Scientific Study of Religion (1973, Vol. 12, No. 3, pp. 297–310). As an archival scholarly contribution, the scale items and scoring directions were placed in the public scientific domain for research, educational, and clinical assessment purposes.

Permissions and Fees: No royalty fees, proprietary paywalls, or commercial licenses are required for non-profit scholarly research, dissertation work, or clinical evaluation. Researchers and clinicians utilizing the measurement tool are required to provide appropriate scholarly attribution by citing the foundational 1973 article by Benson and Spilka in all resulting reports, publications, and presentations.

12. References

• Benson, P. L., & Spilka, B. (1973). God image as a function of self-esteem and locus of control. Journal for the Scientific Study of Religion, 12(3), 297–310. https://doi.org/10.2307/1384430
• Bowman, E. S., Coons, P. M., Jones, R. S., & Oldstrom, M. (1987). Religious psychodynamics in multiple personalities: Suggestions for treatment. American Journal of Psychotherapy, 41(4), 542–554. https://doi.org/10.1176/appi.psychotherapy.1987.41.4.542
• Brokaw, B. F., & Edwards, K. J. (1994). The relationship of God image to level of object relations development. Journal of Psychology and Theology, 22(4), 352–371. https://doi.org/10.1177/009164719402200411
• Gabbard, C. E., Howard, G. S., & Tageson, C. W. (1986). Assessing locus of control with religious populations. Journal of Research in Personality, 20(3), 292–308. https://doi.org/10.1016/0092-6566(86)90136-1
• Hall, T. W., Tisdale, T. C., & Brokaw, B. F. (1994). Assessment of religious dimensions in Christian clients: A review of selected instruments for research and clinical use. Journal of Psychology and Theology, 22(4), 395–421. https://doi.org/10.1177/009164719402200414
• Hertel, B. R., & Donahue, M. J. (1995). Parental influences on God images among children: Testing Durkheim’s metaphoric parallelism. Journal for the Scientific Study of Religion, 34(2), 186–199. https://doi.org/10.2307/1386765
• Jolley, J. C., & Taulbee, S. J. (1986). Assessing perceptions of self and God: Comparison of prisoners and normals. Psychological Reports, 59(3), 1139–1146. https://doi.org/10.2466/pr0.1986.59.3.1139
• Kirkpatrick, L. A., & Shaver, P. R. (1992). An attachment-theoretical approach to romantic love and religious belief. Personality and Social Psychology Bulletin, 18(3), 266–275. https://doi.org/10.1177/0146167292183002
• Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The Measurement of Meaning. University of Illinois Press.
• Park, C. L., & Cohen, L. H. (1993). Religious and nonreligious coping with the death of a friend. Cognitive Therapy and Research, 17(6), 561–577. https://doi.org/10.1007/BF01176079
• Rizzuto, A.-M. (1979). The Birth of the Living God: A Psychoanalytic Study. University of Chicago Press.
• Spilka, B., Hood, R. W., Hunsberger, B., & Gorsuch, R. (2003). The Psychology of Religion: An Empirical Approach (3rd ed.). Guilford Press.

13. Items of the Scale