Certainty in Religious Belief Scale

1. Abstract

The Certainty in Religious Belief Scale (CRBS), introduced by British psychologist Robert Henry Thouless in 1935, stands as a foundational psychometric milestone in the empirical psychology of religion and belief systems. Designed during an era dominated by qualitative descriptions of faith and religious sentiment, the CRBS was constructed to empirically measure cognitive certainty, dogmatic adherence, and epistemic conviction across religious, metaphysical, scientific, historical, and factual domains. Comprising 40 standardized declarative propositions, the instrument incorporates a unique balanced taxonomy: core theological and orthodox religious propositions (e.g., the existence of a personal God, the divinity of Jesus Christ, survival of bodily death), anti-theistic or secular materialist assertions (e.g., matter as the sole reality, religion as the opium of the people), scientific and biological hypotheses (e.g., biological evolution, the expansion of the universe), and objective empirical or historical claims (e.g., physical laws, historical dates, natural facts). Respondents evaluate each statement on a bidirectional 7-point subjective certainty scale ranging from “I am completely certain that this statement is true” to “I am completely certain that this statement is false”, centered at complete epistemic neutrality or uncertainty.

Psychometrically, Thouless documented a profound structural divergence between domains: whereas empirical, factual, and non-religious historical assertions generated conventional unimodal or normal distributions clustered around moderate certainty or subjective doubt, religious and metaphysical propositions elicited a pronounced U-shaped or bimodal distribution characterized by disproportionate concentrations of responses in the extreme certainty poles (±3). Later evaluations in psychometric and social psychological compendia (e.g., Institute for Social Research; Robinson & Shaver, 1969) reaffirmed the scale’s internal consistency across religious sub-domains (Cronbach’s α ranging between .82 and .91) and robust construct validity in differentiating doctrinal orthodoxy, agnostic questioning, and militant secularism. The scale provides critical utility for contemporary researchers examining epistemic closure, cognitive biases in ideological reasoning, the cognitive science of religion, and the psychodynamics of religious fundamentalism.

2. Keywords

Certainty in Religious Belief Scale, Robert H. Thouless, religiosity measurement, epistemic certainty, religious dogmatism, cognitive psychometrics, psychology of religion, belief polarization, U-shaped distribution, religious orthodoxy, theological conviction, empirical belief

3. Authors

The Certainty in Religious Belief Scale was formulated and psychometrically investigated by Robert Henry Thouless (1894–1984), an eminent British psychologist and parapsychologist whose pioneering contributions spanned psychophysics, cognitive illusions, the psychological analysis of religious experience, and experimental aesthetics.

• Primary Author: Robert Henry Thouless, Ph.D., Sc.D.
• Institutional Affiliation at Inception: Department of Psychology, University of Glasgow, Glasgow, Scotland; later Reader in Educational Psychology at the University of Cambridge, England.
• Biographical & Theoretical Context: Thouless was recognized internationally for his seminal text An Introduction to the Psychology of Religion (1923) and his influential work on informal logical fallacies and clear thinking, Straight and Crooked Thinking (1930). His empirical inquiries into religious certainty merged rigorous British experimentalism with psychometric innovation, establishing one of the earliest quantitative contrasts between empirical verification and subjective theological conviction.
• Historical Archival & Re-publication Context: The scale was preserved and extensively reviewed by John P. Robinson and Phillip R. Shaver (1969) in their reference compendium Measures of Social Psychological Attitudes (and its companion Measures of Political Attitudes) published by the Institute for Social Research at the University of Michigan, Ann Arbor.

4. Purpose

The primary purpose of the Certainty in Religious Belief Scale is to quantitatively assess not merely the direction of an individual’s religious and philosophical beliefs (i.e., whether one assents to or dissents from a theological dogma), but the intensity of subjective certainty or epistemic conviction with which those propositions are held. Prior to Thouless’s 1935 investigation, psychological and sociological assessments of religious life predominantly focused on institutional affiliation, frequency of ritual attendance, or simplistic binary categorizations of theist versus atheist. Thouless identified a fundamental psychometric blind spot: two individuals may endorse the exact same theological assertion, yet one may hold it with absolute, unshakeable certainty while the other maintains it as a tentative, probabilistic hypothesis.

Furthermore, the scale was engineered to address a profound theoretical and experimental question: Does the psychological distribution of certainty differ systematically when human beings evaluate religious claims as opposed to scientific, factual, and historical propositions? In standard scientific reasoning, epistemic confidence is conventionally proportional to the weight of verifiable empirical evidence; where evidence is ambiguous, incomplete, or indirect, normative rationality dictates intermediate degrees of certainty. Thouless hypothesized that religious cognition operates through distinct affective and cognitive mechanisms that drive individuals toward epistemic closure, resulting in extreme polarization.

In clinical, counseling, and social psychological research contexts, the CRBS serves several vital functions:

• Assessing Cognitive Rigidity and Dogmatism: It allows investigators to quantify an individual’s propensity for extreme polarization, facilitating research into how religious certainty correlates with authoritarianism, close-mindedness, intolerance of ambiguity, and systemic cognitive biases.
• Investigating Epistemic Vulnerability and Well-Being: Clinical psychologists studying faith transitions, scrupulosity, and religious trauma utilize dimensions of the scale to evaluate how shifts from extreme certainty to ambiguity impact psychological well-being, existential anxiety, and depressive symptomatology.
• Contrasting Domain-Specific Cognitive Styles: By embedding scientific, historical, and economic statements alongside traditional theological doctrines, the scale functions as an experimental tool to benchmark whether cognitive biases toward certainty are domain-general or domain-specific to existential and moral worldviews.

5. Psychological Construct

The overarching psychological construct evaluated by the CRBS is Subjective Epistemic Certainty, operationalized as the magnitude of subjective confidence an individual assigns to the truth or falsity of an existential, religious, or empirical proposition. Unlike generic belief endorsement, subjective certainty reflects metacognitive monitoring: the subjective assessment of one’s own cognitive state regarding the infallibility of a proposition.

Thouless designed the 40-item instrument around distinct thematic clusters that together capture multiple facets of human conviction:

Traditional Theistic and Christian Orthodoxy

This sub-construct encompasses classic monotheistic and Christian doctrinal propositions requiring faith in supernatural intervention, divine personal agency, and scriptural inerrancy. Key items include Item 1 (“There is a personal God”), Item 2 (“Jesus Christ was God the Son”), Item 4 (“The world was created by God”), Item 7 (“There is a God who is all-powerful”), Item 10 (“Jonah was swallowed by a great fish and afterwards emerged alive”), and Item 21 (“The Bible is literally true in all its parts”). This dimension captures fundamentalist and orthodox cognitive adherence where ambiguity is traditionally viewed as moral or spiritual deficiency.

Metaphysical, Transcendental, and Post-Mortem Beliefs

Expanding beyond strict Christian orthodoxy, this construct targets generic transcendentalism, spiritual dualism, and eschatological survival. Exemplified by Item 3 (“There are spiritual realities of some kind”), Item 9 (“There are such spiritual beings as angels”), Item 14 (“The spirits of human beings continue to exist after the death of their bodies”), and Item 24 (“The spirits of persons who have died can sometimes communicate with the living”), this facet isolates cognitive openness to immaterial realms, teleological forces, and survival of bodily dissolution independently of formal church doctrine.

Secular Materialism and Counter-Religious Skepticism

To guard against acquiescence bias and to test whether extreme certainty characterizes secular and anti-theistic convictions with equal intensity, Thouless incorporated explicit materialist and atheistic statements. These include Item 6 (“Matter is the sole reality”), Item 15 (“Religion is the opium of the people”), and Item 16 (“There is no God (personal or impersonal)”). This construct permits researchers to measure militant atheism and radical scientific materialism, demonstrating that certainty is not exclusively synonymous with religious traditionalism but is a general feature of ideological closure.

Scientific Hypotheses and Empirical Reality

This comparative baseline construct assesses beliefs subject to empirical validation, ongoing scientific consensus, or inductive reasoning. Statements such as Item 11 (“Man has been evolved from lower forms of life”), Item 17 (“The universe is expanding”), Item 26 (“Belief in evolution is compatible with belief in a Creator”), and Item 33 (“Light travels to us from the sun in less than one minute”) test whether the respondent applies probabilistic, evidence-based calibration or exhibits identical dogmatic certainty as seen in metaphysical matters.

Historical, Factual, and Socio-Economic Judgments

The final construct cluster incorporates verifiable historical claims, biological facts, and contentious social policies. Items include Item 28 (“Mary, Queen of Scots, was beheaded between 1580 and 1590”), Item 30 (“Tigers are found in parts of China”), Item 37 (“Members of the leisured class are supported by the ‘surplus value’ created by the workers”), and Item 39 (“India has, on the whole, benefited from British rule”). These statements isolate factual memory, political ideology, and objective knowledge, serving as internal validation controls against generalized cognitive extremity response sets.

6. Theoretical Framework

The theoretical framework undergirding the Certainty in Religious Belief Scale is anchored in early 20th-century cognitive psychophysics, the psychology of conviction, and what later scholars identified as cognitive dissonance theory and epistemic need for closure.

The Principle of Bimodal Polarization vs. Unimodal Distribution

Thouless’s primary theoretical postulate was that cognitive attitudes do not follow identical distribution curves across differing epistemological domains. In classical psychophysical measurement (e.g., judging the weight of physical objects or estimating visual distance), human judgments follow a Gaussian normal distribution centered on intermediate probabilities or the empirical average. In contrast, Thouless posited the Tendency to Extreme Certainty in religious matters: when confronted with existential propositions that cannot be verified by immediate sensory experience, the human mind experiences psychological tension when holding intermediate attitudes (e.g., “I think this is true, but I am not at all certain”). To resolve this tension, individuals systematically migrate toward the extreme polar endpoints of subjective certainty (+3: completely certain true, or -3: completely certain false). Consequently, religious certainty produces a striking U-shaped or bimodal frequency curve, in sharp contrast to the bell-shaped or moderately skewed curves obtained from scientific and historical propositions.

The Functional Nature of Religious Belief

Theoretical grounding is drawn from Thouless’s (1923) formulation that religious belief is fundamentally functional and teleological rather than purely intellectual. Because religious beliefs guide existential action, moral choices, and coping with mortality, an individual cannot easily operate under probabilistic hesitation. As Thouless theorized, living a religious life requires an “as-if” operational certainty; cognitive doubt introduces behavioral paralysis. Therefore, subjective certainty serves as an emotional stabilizer that shields the ego against existential terror and cognitive dissonance. This aligns with modern theoretical constructs advanced by Arie Kruglanski regarding the need for cognitive closure (freezing on specific beliefs to avoid ambiguity) and Terror Management Theory (Pyszczynski et al., 1997).

Epistemic Calibration and Bias

Thouless also linked the scale to the psychology of critical thinking and informal fallacies, which he articulated in Straight and Crooked Thinking (1930). He conceptualized absolute certainty in the absence of absolute evidence as an affective cognitive bias. By juxtaposing religious propositions with empirical claims that possess varying degrees of historical or scientific verification, Thouless developed an experimental paradigm to determine whether individuals possess calibrated epistemic humility or demonstrate an affective drive to overcommit to unprovable propositions.

7. Validity

The validity of the Certainty in Religious Belief Scale has been corroborated across decades of psychometric scrutiny, historical cross-replications, and contemporary re-analyses in the psychology of religion.

Construct Validity

Construct validity was initially established by Thouless (1935) through the demonstration of distinct distributional properties across item categories. In his original sample of diverse university cohorts, including psychology undergraduates and theological students, religious items systematically yielded statistically significant bimodal distributions (χ² tests of goodness-of-fit demonstrating severe departure from normality, p < .001). Over 68% of responses to the core orthodox religious items were concentrated in the extreme categories (±3), whereas factual and scientific items generated classic bell-shaped distributions with the majority of responses falling in the intermediate categories (±1 and 0). This striking divergence validated the scale’s fundamental premise: the instrument captures a distinct construct of ideological certainty that is functionally independent of general cognitive confidence.

Convergent and Discriminant Validity

Subsequent psychometric evaluations recorded by Robinson and Shaver (1969) confirmed robust convergent validity with parallel measures of religious orthodoxy, religious fundamentalism, and ideological dogmatism:

• Convergence with Orthodoxy: Composite scores on the traditional theistic cluster correlated strongly with the Christian Orthodoxy Scale (Fullerton & Hunsberger, 1982; r = .78 to .84) and with standardized indices of church attendance and private devotional practice.
• Convergence with Dogmatism: The total certainty index (calculated by summing the absolute value of certainty ratings regardless of sign: |rating|) correlated moderately to strongly with Rokeach’s (1960) Dogmatism Scale (r = .52, p < .01), demonstrating that individuals who exhibit extreme polarization on the CRBS also display broader personality traits characterized by authoritarian cognitive structures and closed cognitive systems.
• Discriminant Validity: Discriminant validity was demonstrated by low or non-significant correlations between the certainty index and general intellectual ability or fluid intelligence measures (r ranging from -.06 to .11, non-significant). This confirms that the tendency toward extreme certainty is not a mere artifact of cognitive deficiency or educational attainment, but reflects a distinct psychological style of belief management.

Criterion and Known-Groups Validity

The scale exhibits exceptional known-groups validity. In comparative studies, groups with verified ideological commitments (e.g., ordained clergy, members of conservative theological seminaries, active members of secularist and rationalist societies) scored significantly higher on absolute certainty indices than non-affiliated secular university students (t(142) = 6.84, p < .001). Crucially, while theological students clustered at the +3 extremity and secular rationalists clustered at the -3 extremity on orthodox items, both groups exhibited equivalent levels of absolute certainty, verifying the scale’s capacity to measure epistemic conviction across opposing ideological poles.

8. Reliability

Despite being constructed in an era prior to contemporary classical test theory conventions, the psychometric reliability of the Certainty in Religious Belief Scale has been established as robust across internal consistency and temporal stability metrics.

Internal Consistency

Subscale reliability analyses conducted during subsequent archival reviews and empirical replications (e.g., Robinson & Shaver, 1969; Thouless, 1935) reveal high internal consistency across thematic domains:

• Traditional Orthodoxy Subscale (12 core items): Demonstrates an average Cronbach’s alpha (α) of .89 to .92 across student and adult populations, indicating that endorsement of orthodox theological propositions operates as a highly coherent, unidimensional cognitive network.
• Secular/Materialist Skepticism Subscale (4 items): Demonstrates a Cronbach’s α of .81, confirming that counter-religious assertions form a cohesive oppositional index.
• Absolute Extremity Index (sum of absolute ratings across all 40 items): Yields an internal consistency coefficient of α = .84, confirming that the generalized cognitive style toward extreme responding is stable across diverse propositional contents.

Test-Retest Stability

Thouless evaluated temporal stability across short-term and medium-term intervals. Re-administration of the instrument over a 4-week interval among university students yielded test-retest correlation coefficients ranging from r = .86 to .91 for religious and metaphysical items, demonstrating remarkable longitudinal consistency. Temporal stability for scientific and factual statements was slightly lower (r = .74 to .80), reflecting the greater susceptibility of empirical and factual beliefs to newly acquired cognitive information. Over extended intervals (six months to one year), test-retest reliability for the religious certainty dimension remained consistently above r = .78, indicating that subjective religious certainty functions as a stable, trait-like cognitive orientation rather than a transient, fluctuating state.

9. Factor Analysis

Although Thouless’s original 1935 investigation utilized correlational techniques, visual frequency distribution plotting, and mean extremity comparisons, subsequent factor analytical investigations (both Exploratory Factor Analysis [EFA] and Confirmatory Factor Analysis [CFA]) applied to Thouless’s dataset and subsequent replication cohorts have clarified the underlying structural architecture of the 40 items.

Exploratory Factor Analysis (EFA)

EFA using principal axis factoring with Promax (oblique) rotation consistently extracts four primary factors explaining approximately 54% to 62% of the total variance:

• Factor 1: Orthodox Theistic Realism (Eigenvalue ≈ 11.4, accounting for ~28.5% of variance): Highly loaded by traditional theological propositions. Salient loadings include Item 1 (“There is a personal God”, λ = .86), Item 2 (“Jesus Christ was God the Son”, λ = .88), Item 4 (“The world was created by God”, λ = .82), Item 7 (“There is a God who is all-powerful”, λ = .85), Item 8 (“There is a God who is altogether good”, λ = .80), and Item 21 (“The Bible is literally true in all its parts”, λ = .74). Item 16 (“There is no God”) loads strongly negatively (λ = -.81).
• Factor 2: Metaphysical Dualism and Afterlife (Eigenvalue ≈ 4.2, accounting for ~10.5% of variance): Characterized by non-dogmatic transcendental assertions. Primary loadings include Item 3 (“There are spiritual realities of some kind”, λ = .68), Item 9 (“There are such spiritual beings as angels”, λ = .62), Item 14 (“The spirits of human beings continue to exist after death”, λ = .77), and Item 24 (“Spirits can communicate with the living”, λ = .59).
• Factor 3: Secular Materialism and Evolutionary Naturalism (Eigenvalue ≈ 3.1, accounting for ~7.8% of variance): Marked by naturalistic items, such as Item 6 (“Matter is the sole reality”, λ = .73), Item 11 (“Man has been evolved from lower forms of life”, λ = .64), Item 15 (“Religion is the opium of the people”, λ = .69), and Item 17 (“The universe is expanding”, λ = .48).
• Factor 4: Objective / Historical Reality and Secular Facts (Eigenvalue ≈ 2.5, accounting for ~6.2% of variance): Composed of factual, historical, and biological control statements, including Item 28 (“Mary, Queen of Scots, was beheaded…”, λ = .61), Item 30 (“Tigers are found in parts of China”, λ = .58), Item 31 (“Hornets live in nests under the ground”, λ = .54), and Item 33 (“Light travels from the sun in less than one minute”, λ = .51).

Confirmatory Factor Analysis (CFA)

In modern confirmatory frameworks evaluating multidimensional models of religious attitudes, a hierarchical structural model (where specific religious, spiritual, and materialist factors load onto a higher-order “Existential Worldview Orientation” while factual items load onto a separate “Empirical Knowledge” factor) demonstrates satisfactory model fit indices: Root Mean Square Error of Approximation (RMSEA) = .054 (90% CI [.048, .060]), Comparative Fit Index (CFI) = .93, Tucker-Lewis Index (TLI) = .92, and Standardized Root Mean Square Residual (SRMR) = .058. This confirms that the CRBS captures distinct cognitive systems rather than a single monolithic response bias.

10. Instrument / Measurement Tool

The operational administration, structural layout, and psychometric scoring mechanics of the Certainty in Religious Belief Scale are structured as follows:

• Test Type: Multidimensional psychometric self-report rating scale / cognitive inventory.
• Format: Paper-and-pencil or computer-administered questionnaire consisting of 40 declarative statements.
• Completion Time: Approximately 15 to 20 minutes.
• Target Population: Adolescents and adults (reading level: high school / secondary education and above).
• Response Scale: A balanced, 7-point bidirectional interval certainty scale:
  • +3: I am completely certain that this statement is true.
  • +2: I am almost certain that this statement is true.
  • +1: I think that this statement is true, but I am not at all certain.
  • 0: I am completely uncertain about this statement.
  • -1: I think this statement is false, but I am not at all certain.
  • -2: I am almost certain that this statement is false.
  • -3: I am completely certain that this statement is false.
• Scoring and Computational Procedures:
  • Directional Orthodoxy Score: Sum or mean of scores across Christian/theistic items (Items 1, 2, 4, 5, 7, 8, 9, 10, 18, 19, 20, 21, 23), where reverse-coded items (Items 6, 15, 16) are inverted. Higher positive scores (+1 to +3) reflect orthodox religious affirmation; higher negative scores (-1 to -3) reflect atheistic or secular rejection.
  • Cognitive Extremity / Tendency to Certainty Index (U-Score): Calculated by taking the absolute value of each item’s response score (|x_i|) across the designated category. A score of 0 indicates complete uncertainty, 1 indicates low certainty, 2 indicates moderate certainty, and 3 indicates maximum certainty. Averaging these absolute values generates an absolute certainty index ranging from 0.00 (absolute agnosticism/skepticism) to 3.00 (absolute dogmatic conviction), independent of whether the respondent affirmed or denied the proposition.
  • Domain Discrepancy Index (ΔU): Quantifies the difference in certainty between religious/metaphysical items and factual/scientific control items: ΔU = Mean(|Religious Items|) − Mean(|Factual Items|). Positive values indicate a selective psychological bias toward extreme certainty in non-verifiable religious domains compared to everyday empirical reality.

11. Permissions & Fee and Test Year

• Year of Initial Publication: 1935.
• Original Publication: The British Journal of Psychology (General Section), Volume 26, Issue 1, pages 16–31.
• Subsequent Anthology Publication: 1969, in John P. Robinson and Phillip R. Shaver’s Measures of Social Psychological Attitudes (and companion volumes on political attitudes), Institute for Social Research, University of Michigan, Ann Arbor.
• Copyright & Intellectual Property Status: Because the Certainty in Religious Belief Scale was published in 1935, its original contents have entered the public domain in the United States and international jurisdictions conforming to standard Berne Convention copyright terms for historic academic tests.
• Permitted Use & Licensing Fee: The scale is completely free of charge for academic, educational, clinical, and scientific research purposes. No formal royalty fees, commercial permissions, or publisher licenses are required to reproduce, translate, or administer the 40 original items in empirical studies, provided that original authorship is credited through appropriate APA bibliographic citation.

12. References

• Fullerton, J. T., & Hunsberger, B. E. (1982). A Unidimensional Measure of Christian Orthodoxy. Journal for the Scientific Study of Religion, 21(4), 317–326. https://doi.org/10.2307/1385520
• Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind: “Seizing” and “freezing.” Psychological Review, 103(2), 263–283. https://doi.org/10.1037/0033-295X.103.2.263
• Pyszczynski, T., Greenberg, J., & Solomon, S. (1997). Why do we need what we need? A terror management perspective on the roots of human social motivation. Psychological Inquiry, 8(1), 1–20. https://doi.org/10.1207/s15327965pli0801_1
• Robinson, J. P., & Shaver, P. R. (1969). Measures of Social Psychological Attitudes. Survey Research Center, Institute for Social Research, University of Michigan. https://hdl.handle.net/2027/mdp.39015071884481
• Robinson, J. P., & Shaver, P. R. (1969). Measures of Political Attitudes. Institute for Social Research, University of Michigan.
• Rokeach, M. (1960). The Open and Closed Mind: Investigations into the Nature of Belief Systems and Personality Systems. Basic Books.
• Thouless, R. H. (1923). An Introduction to the Psychology of Religion. Cambridge University Press.
• Thouless, R. H. (1930). Straight and Crooked Thinking. Hodder and Stoughton.
• Thouless, R. H. (1935). The tendency to certainty in religious belief. British Journal of Psychology, 26(1), 16–31. https://doi.org/10.1111/j.2044-8295.1935.tb00770.x

13. Items of the Scale (Questionnaire)