Attitude toward the Church Scale

Abstract

The Attitude toward the Church Scale, developed by psychometric pioneer Louis Leon Thurstone and theologian Ernest John Chave in 1929, represents a foundational milestone in social science measurement. As the premier operationalization of the method of equal-appearing intervals, this instrument demonstrated for the first time that complex, subjective social attitudes could be quantitatively measured on a continuous, linear metric analogous to physical psychophysical scales. The instrument measures an individual’s net affective and cognitive disposition toward the church as an organized religious institution, spanning a continuum from intense antipathy and rejection to ardent devotion and institutional loyalty. The original scale exists in two parallel, psychometrically equivalent versions (Form A and Form B), each comprising 45 carefully calibrated statements selected from an initial pool of 130 candidate opinions. The response format departs from modern Likert summated ratings; instead, respondents endorse only those statements with which they substantively agree. Each item possesses an empirically determined scale value (ranging from approximately 0.2 for intensely favorable attitudes to 11.0 for intensely unfavorable attitudes), determined by the median sorting categorization of 300 independent human judges. Scoring is achieved by calculating the median or mean scale value of all statements endorsed by a respondent, with item ambiguity evaluated via the interquartile range ($Q$-value). Psychometric evaluations consistently demonstrate high parallel-form reliability ($r = .84$ to $.92$) and robust criterion-related validity, effectively discriminating between regular churchgoers, theological students, active religious professionals, and committed secularists or agnostics.

Keywords

Attitude toward the Church Scale, Thurstone scaling, method of equal-appearing intervals, Louis Leon Thurstone, Ernest John Chave, religious attitudes, psychometrics, attitude measurement, institutional religion, scale value, item ambiguity

Authors

The scale was collaboratively formulated by Louis Leon Thurstone (1887–1955) and Ernest John Chave (1886–1961) at the University of Chicago.

• Louis Leon Thurstone, Ph.D.: Professor of Psychology at the University of Chicago, founder of the Psychometric Society, and developer of the Law of Comparative Judgment, factor analysis techniques (including multiple factor analysis), and fundamental psychological scaling methodologies.
• Ernest John Chave, Ph.D.: Associate Professor of Religious Education at the University of Chicago Divinity School, whose scholarly focus centered on the psychological, moral, and character development of youth, religious education curriculum evaluation, and the empirical measurement of religious behavior.

Purpose

The primary objective of the Attitude toward the Church Scale was both theoretical and practical: to resolve the prevailing early-20th-century skepticism regarding whether human attitudes, values, and sentiments could be measured scientifically. Prior to Thurstone’s breakthroughs, subjective phenomena such as religious sentiment were widely considered too multifaceted, fluid, and qualitative to yield to mathematical modeling. Thurstone and Chave sought to demonstrate that an attitude toward a defined social institution—specifically, the Christian Church—could be projected onto a linear continuum representing the degree of favorable or unfavorable affect associated with that institution.

In clinical, pastoral, and educational research, the instrument serves multiple applied functions:

• Longitudinal Tracking of Religious Sentiment: Tracking shifts in religious commitment across educational milestones, such as matriculation through secondary and tertiary higher education, evaluating secularization hypotheses, and assessing whether scientific training attenuates institutional religious endorsement.
• Institutional Evaluation: Assisting religious denominations, parachurch organizations, and community planners in assessing the impact of sermons, liturgical reforms, evangelistic campaigns, or youth programs on congregational attitudes.
• Sociological and Political Research: Facilitating empirical inquiries into how institutional church attitudes intersect with broader political attitudes, civic participation, socio-economic status, authoritarianism, and intergroup prejudice.
• Counseling and Pastoral Psychology: Providing clinicians and pastoral counselors with a standardized metric to evaluate spiritual crisis, religious disillusionment, or faith reconciliation in clients navigating faith transitions.

Psychological Construct

The psychological construct assessed by the instrument is attitude toward the church, defined by Thurstone as the degree of positive or negative affect associated with the church as an organized social and spiritual institution. Unlike measures of general theological orthodoxy, mystical experience, or private spirituality, this construct explicitly targets the organized ecclesiastical body, its social functions, its doctrinal assertions, and its institutional integrity.

While designed by Thurstone and Chave as a unidimensional affective-evaluative continuum, the construct encompasses several interconnected sub-facets reflected across the statement pool:

• Social Utility and Cultural Progress: Evaluates the degree to which the church is perceived as a force for societal betterment, social welfare, moral leadership, and civic righteousness versus an outdated, parasitic, or reactionary impediment to human progress. Statements range from viewing the church as “the greatest influence for good government and right living” to condemning it as an “enemy of science and truth.”
• Spiritual Inspiration and Moral Guidance: Assesses personal reliance on church services, rituals, sermons, and community for ethical direction, emotional comfort, communion with the transcendent, and motivation for daily living. Favorable items highlight church attendance as restful, inspiring, and essential for optimal moral living, whereas critical items view church teachings as superficial platitudes.
• Institutional Cynicism and Hypocrisy: Captures evaluations of institutional integrity, financial management, denominational infighting, and clergy sincerity. Negative manifestations include perceptions of the church as bound by monetary interests, marked by sectarian quarrels, dominated by narrow-minded adherents, and harboring pervasive hypocrisy.
• Doctrinal and Scientific Rationality: Probes the compatibility between church dogma and modern scientific, educational, and intellectual paradigms. Unfavorable statements characterize ecclesiastical doctrines as “worn-out dogmas and medieval superstitions,” asserting that educated individuals must abandon or maintain mental reservations regarding church tenets.

Theoretical Framework

The scale is anchored in Thurstone’s psychophysical scaling theory, specifically the Law of Comparative Judgment and its practical operationalization, the Method of Equal-Appearing Intervals. Drawing from the classical psychophysics of Ernst Heinrich Weber and Gustav Theodor Fechner, Thurstone hypothesized that psychological stimuli (such as written statements of opinion) could be mapped onto an internal psychological continuum in a manner directly analogous to physical stimuli (such as weights, auditory frequencies, or visual brightness).

The methodology rests on four foundational assumptions:

1. Linearity of the Evaluative Continuum: Any specific attitude toward a social object can be conceptualized as falling along a single, continuous psychological baseline ranging from extreme positivity (favorableness) to extreme negativity (unfavorableness).
2. Independence of Scale Values from Judges’ Own Attitudes: Thurstone postulated that the scale value assigned to a statement reflects its intrinsic objective meaning along the psychological continuum and should remain invariant regardless of the personal attitudes held by the judges sorting the items. Although this assumption was later qualified by social judgment theorists (such as Muzafer Sherif and Carl Hovland, who demonstrated assimilation-contrast effects), Thurstone and Chave verified empirically that judges with divergent personal church commitments produced highly correlated scale values for the statements.
3. Equal-Appearing Distance: The subjective distance between adjacent interval categories (e.g., category 1 to category 2, versus category 6 to category 7 across an 11-step sorting sorting grid labeled A through K) is perceived by judges as psychologically uniform.
4. Measurement by Endorsement: A respondent’s position on the continuum corresponds to the central tendency (median) of the psychological scale values of the discrete items that the respondent endorses as reflecting their personal perspective.

Validity

Thurstone and Chave (1929) established extensive empirical evidence for the validity of the Attitude toward the Church Scale through multiple psychometric paradigms:

• Content Validity: Content validity was systematically engineered during initial instrument design. An exhaustive item pool of several hundred candid statements expressing opinions about the church was culated from literature, religious publications, university essays, and personal interviews. After eliminating obvious redundancies, colloquial ambiguities, and compound propositions, 130 candidate statements were retained for sorting, ensuring comprehensive representation of affective, cognitive, and social attitudes toward the church.
• Criterion-Related and Known-Groups Validity: The authors administered the calibrated scale to distinct demographic cohorts whose actual behavioral commitments to the church were independently known. The empirical distributions perfectly reflected predicted group dynamics:
  • Active divinity school students and regular churchgoers scored almost exclusively at the favorable end of the continuum (median scores ranging between 1.0 and 3.0).
  • General university undergraduate samples displayed a broad, approximately normal distribution centered around neutral-to-moderately favorable scores (median scores around 4.5 to 5.5).
  • Members of secularist societies, self-identified agnostics, and non-attenders clustered significantly at the unfavorable end of the scale (median scores between 7.5 and 10.5).
• Behavioral Correlation (Predictive/Concurrent Validity): Thurstone and Chave documented strong correlations between respondents’ median scale scores and their self-reported frequency of church attendance ($r > .60$). Individuals who attended church weekly demonstrated significantly lower (more favorable) scale values than individuals who rarely or never attended.
• Scale Invariance across Varying Judge Panels: High inter-judge correlation coefficients ($r > .98$) were obtained when comparing the median scale values derived from groups of judges who were themselves sympathetic to the church against groups of judges who were critical, confirming construct stability independent of judge biases.

Reliability

Because Thurstone scales utilize the endorsement of heterogeneous items across an equal-appearing interval continuum rather than summated homogeneous item pools (as in Likert scaling), classical internal consistency metrics such as Cronbach’s alpha are less theoretically aligned than parallel-form and split-half reliability coefficients:

• Alternate-Form Reliability: The primary reliability benchmark established by Thurstone and Chave (1929) is the correlation between the two parallel 45-item forms (Form A and Form B). When administered simultaneously or in immediate sequence to university student cohorts ($N > 200$), the alternate-form correlation coefficient reached $r = .848$. When corrected for attenuation via the Spearman-Brown prophecy formula to estimate the reliability of the combined 90-item assessment, reliability exceeded $.92$.
• Item Scale Ambiguity Index ($Q$): For every statement, the interquartile range ($Q = Q_3 – Q_1$) was calculated from the 300 judges’ sort distributions. Items exhibiting high $Q$-values (indicating wide judge disagreement regarding where the item belonged on the 11-point scale) were deemed ambiguous and eliminated from the final 45-item forms, ensuring that only items with high sorting precision were retained.
• Test-Retest Stability: Subsequent replications and pedagogical administrations (e.g., Robinson & Shaver, 1969; Zazeckis, 1981) have confirmed high temporal stability over brief intervals (two to four weeks), yielding test-retest coefficients consistently ranging between $r = .80$ and $.88$.

Factor Analysis

Although the scale was constructed under the conceptual assumption of a single unidimensional continuum, subsequent psychometric investigations utilizing exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) on Thurstone items have demonstrated the presence of underlying multidimensional sub-factors:

• Dominant General Factor: When items are coded dichotomously (endorsed vs. not endorsed) or analyzed via tetrachoric correlation matrices, a dominant general factor accounts for over 45% to 55% of the common variance, confirming the practical viability of Thurstone’s original single-score representation of overall favorable/unfavorable affect.
• Multidimensional Components: Secondary factor solutions typically reveal three to four distinct, correlated sub-dimensions:
  1. Personal Spiritual Utility: Items reflecting individual inspiration, peace, moral strength, and regular worship (e.g., Items 2, 4, 11, 45).
  2. Socio-Cultural Value: Items evaluating the church’s macro-level influence on national morality, civic justice, politics, and social welfare (e.g., Items 3, 14, 33, 44).
  3. Intellectual and Scientific Obsoleteness: Items critiquing church dogma, alleged hostility to scientific advancement, and retention of medieval superstition (e.g., Items 21, 34, 37, 41).
  4. Institutional Skepticism and Moral Critique: Items focusing on financial waste, factional bickering, hypocrisy, and social parasitism (e.g., Items 10, 17, 19, 26, 40).
• Model Fit in Modern Re-evaluations: Modern structural equation modeling of modified versions of the scale indicates that while a unidimensional model provides acceptable broad descriptive utility, a higher-order model incorporating a global attitude factor governing distinct personal, civic, and doctrinal sub-dimensions yields superior fit indices (e.g., $\text{CFI} > .93$, $\text{TLI} > .91$, $\text{RMSEA} < .06$).

Instrument / Measurement Tool

• Instrument Type: Equal-Appearing Interval Attitude Scale (Thurstone scale).
• Administration Format: Self-administered paper-and-pencil questionnaire (or computerized equivalent); individual or group administration.
• Completion Time: Approximately 10 to 15 minutes per 45-item form.
• Item Structure: 45 calibrated opinion statements per standardized form (Form A and Form B), selected from a comprehensive standardized pool of 130 statements.
• Response Format: Binary endorsement checklist. The respondent is instructed to read through the statements and place a checkmark ($checkmark$) next to every statement with which they fully agree, leaving blank those statements with which they disagree or toward which they feel indifferent.
• Scoring and Quantification Rules:
  • Each item possesses an assigned scale value ($S$) between approximately 0.0 and 11.0, where values near 0.0 represent extreme positive favorableness toward the church, values near 6.0 represent neutral/indifferent attitudes, and values near 11.0 represent extreme antagonism or hostility toward the church.
  • Respondent Individual Score: Calculated as the median (or arithmetic mean) of the scale values of all the statements checked/endorsed by the respondent. For example, if a participant endorses five statements with scale values of 1.2, 1.5, 1.7, 2.3, and 2.6, their calculated median score is 1.7, indicating a strongly favorable attitude.
  • If an individual endorses widely discordant statements (e.g., endorsing an item at 1.0 and an item at 10.5), the spread of endorsed items can be quantified using the range or interquartile spread, which serves as an index of internal attitude conflict, ambivalence, or careless responding.

Permissions & Fee and Test Year

• Publication Year: 1929 (Original monograph: The Measurement of Attitude: A Psychophysical Method and Some Experiments with a Scale for Measuring Attitude toward the Church, University of Chicago Press).
• Copyright Status & Permissions: The original 1929 work and its constituent scale items are in the public domain in the United States and most international jurisdictions due to the expiration of copyright protections.
• Usage Fees: Free of charge. Researchers, sociologists, educators, and clinicians may freely utilize, adapt, reproduce, and administer the items for non-commercial research, institutional evaluation, and scholarly inquiry without payment of royalties or mandatory licensing permissions.

References

• Robinson, J. P., & Shaver, P. R. (1969). Measures of political attitudes. Survey Research Center, Institute for Social Research, University of Michigan. https://catalog.hathitrust.org/Record/000282110
• Thurstone, L. L. (1928). Attitudes can be measured. American Journal of Sociology, 33(4), 529–554. https://doi.org/10.1086/214483
• Thurstone, L. L. (1929). Theory of attitude measurement. Psychological Review, 36(3), 222–241. https://doi.org/10.1037/h0070243
• Thurstone, L. L., & Chave, E. J. (1929). The measurement of attitude: A psychophysical method and some experiments with a scale for measuring attitude toward the church. University of Chicago Press. https://doi.org/10.1037/11574-000
• Zazeckis, T. M. (1981). The effect of variant attitudes on the standardization of Thurstone’s scale of attitude toward the church (Master’s thesis, University of Arizona). University of Arizona Libraries. http://hdl.handle.net/10150/557704

Items of the Scale