Attitude toward the Church Scale

Abstract

The Attitude toward the Church Scale, formulated by psychometrician Louis Leon Thurstone and religious educator Ernest J. Chave in 1929, stands as a foundational monument in the history of quantitative social science and psychometrics. Developed at the University of Chicago, it represents the primary empirical demonstration of the Method of Equal-Appearing Intervals, a psychophysical scaling paradigm adapted to measure non-physical, subjective psychological constructs. The instrument assesses an individual’s psychological disposition toward the church, conceptualized along a single, continuous evaluative spectrum extending from unreserved affection, loyalty, and reverence to severe antagonism, disparagement, and contempt. Comprising 45 carefully selected and calibrated statements, the instrument operates on a non-cumulative endorsement format wherein respondents check only those propositions with which they fully agree. The scale values of endorsed items—originally derived from the median sorting judgments of over 300 independent raters across an 11-interval psychological continuum—are aggregated to calculate a respondent’s personal attitude score via either the mean or median scale value. Across initial validation cohorts involving university undergraduates, graduate students, theological seminarians, and public forum participants, the scale demonstrated exceptional split-half reliability, yielding Spearman-Brown corrected coefficients between .92 and .94. Criterion-related and known-groups validity were robustly corroborated by marked distinctions between divinity students, frequent church attendees, and secular civic participants, as well as significant correlations with direct self-ratings and behavioral indices of church support. This article presents an exhaustive academic review of the scale’s historical architecture, theoretical foundations in psychophysics, psychometric characteristics, scoring methodologies, and enduring significance in attitude measurement.

Keywords

Attitude toward the Church Scale, Louis Leon Thurstone, Ernest J. Chave, Method of Equal-Appearing Intervals, psychometrics, attitude measurement, sociology of religion, psychology of religion, scale calibration, Thurstone scaling

Authors

The scale was collaboratively developed by two seminal figures at the University of Chicago:

• Louis Leon Thurstone, Ph.D. (1887–1955): Professor of Psychology at the University of Chicago and later the University of North Carolina at Chapel Hill. A revolutionary figure in psychometrics, Thurstone formulated the Law of Comparative Judgment, pioneered modern factor analysis (including the theory of Multiple Primary Mental Abilities), founded the Psychometric Society, and established the journal Psychometrika. His work demonstrated that social and subjective phenomena could be quantified with mathematical precision comparable to physical stimuli.
• Ernest John Chave, Ph.D. (1886–1960): Associate Professor of Religious Education in the Divinity School at the University of Chicago. Chave brought substantive domain expertise in religious sociology, character education, and moral development, orchestrating the item generation, theological nuances, and contextual validation necessary to translate Thurstone’s psychophysical techniques into religious and educational research.

Purpose

The primary purpose of the Attitude toward the Church Scale was both methodological and substantive. In the late 1920s, experimental psychology was dominated by sensory psychophysics and behaviorism; subjective mental phenomena, particularly social and religious attitudes, were widely regarded as ephemeral, multidimensional, and mathematically intractable. Thurstone set out to refute this skepticism by demonstrating that human affection, sentiment, and ideological affinity could be measured on an objective, linear, and quantitative metric scale. The Attitude toward the Church Scale was intentionally selected as the pioneer empirical proving ground for this paradigm.

Substantively, the instrument was engineered to quantify the full continuum of favorable and unfavorable attitudes held toward “the church.” The designers recognized that the church occupies a dual sociological role: it operates simultaneously as a generalized, transcendent institution (the historical, theological body claiming divine sanction and global moral stewardship) and as a localized, empirical nexus of collective worship, social fellowship, and denominational leadership. The scale was constructed to measure this comprehensive sentiment without injecting moralizing judgments or normative presuppositions regarding whether an individual ought to hold positive or negative orientations.

In applied and empirical contexts, the scale served several crucial functions:

• Educational Evaluation: Providing religious educators and collegiate administrators with a scientific instrument to trace shifts in theological affinity and institutional loyalty across academic matriculation (e.g., from freshman to senior years).
• Sociological and Demographic Stratification: Permitting systematic comparisons across diverse populations, such as sectarian affiliations (Catholics, Protestants, Jews), theological trainees (seminarians), and secular or socio-politically active assemblies (such as the Chicago Forum).
• Personality and Social Psychology Research: Facilitating correlational studies linking institutional religious sentiment with social conservatism, authoritarianism, ethnic prejudice, and broad ideological conservatism.
• Attitude-Behavior Consistency: Serving as a benchmark metric in foundational social psychological investigations examining the alignment (or discordance) between verbalized institutional sentiments and concrete behaviors, such as church attendance, financial giving, and civic action (e.g., Wicker, 1971).

Psychological Construct

The construct assessed by the instrument is an individual’s net evaluative disposition toward the Christian church as both an organized historical institution and an active social community. Within Thurstone’s classical psychometric framework, an attitude is defined as the degree of positive or negative affect associated with a psychological object. An attitude is viewed as an internal, latent inclination that mediates an individual’s verbal declarations, emotional reactions, and social behaviors. The scale operationalizes this disposition along a single, continuous, bipolar dimension ranging from extreme, uncritical veneration to intense, active hostility.

Although mathematically treated as a unidimensional continuum, the 45 items capture several distinct psychological facets of institutional sentiment:

• Affective Reverence vs. Moral Repulsion: Statements tap deep emotional attachments and spiritual uplift versus visceral revulsion, contempt, or alienation. At the favorable pole, items reflect an experience of the church as a sacred, indispensable refuge (“In the church I find my best companions and express my best self”); at the antagonistic pole, items express moral disgust, depicting the institution as an instrument of social degradation (“I regard the church as a parasite on society”; “I have nothing but contempt for the church”).
• Cognitive and Epistemic Appraisal: The scale incorporates evaluations regarding the church’s intellectual integrity, alignment with scientific progress, and philosophical viability. Unfavorable items frequently articulate the stance that modern intellectual advancement and institutional religion are mutually incompatible (“I regard the church as a monument to human ignorance”; “I believe that the church is losing ground as education advances”; “I believe the church is nonscientific, depending for its influence on fear of God and hell”). Conversely, moderately favorable items assert that the church adapts constructively to modernity (“I feel the church is trying to adjust itself to a scientific world and deserves support”).
• Social Utility and Cultural Progress: Another primary facet involves the church’s broader impact on societal ethics, human brotherhood, and political progress. Positive items frame the church as an indispensable moral vanguard (“I feel the church is the greatest agency for the uplift of the world”; “I feel that the church promotes a fine brotherly relationship between people and nations”). Negative statements frame the church as an archaic, corrupt, or regressive social barrier (“I regard the church as hopelessly allied with reactionary forces”; “I believe the church is bound hand and foot with monied interests and does not practice its ideals”).
• Pragmatic, Ambivalent, and Indifferent Orientations: Thurstone and Chave recognized that the middle of an attitude continuum is not merely occupied by people with moderate views, but also by those exhibiting indifference, utilitarian qualification, or spiritual ambivalence. The neutral-to-moderate regions of the scale include statements that acknowledge structural faults while valuing core ideals (“I believe in the ideals of my church, but I am tired of its denominationalism”), maintain passive non-involvement (“My attitude toward the church is best described as indifference”), or convey total cognitive detachment (“I know too little about any church to express an opinion”).

Theoretical Framework

The Attitude toward the Church Scale is rooted in psychophysics, specifically the adaptation of classical psychophysical scaling principles formulated by Ernst Heinrich Weber and Gustav Fechner to the psychological measurement of subjective social stimuli. Thurstone postulated that psychological stimuli (such as beliefs, social propositions, and ideological doctrines) evoke internal discriminal processes along a subjective psychological continuum, just as physical stimuli (such as light intensity, sound frequency, or lifted weights) evoke sensory discriminations along physical dimensions.

To quantify these latent distributions, Thurstone formulated the Method of Equal-Appearing Intervals. The construction methodology rested upon several rigorous theoretical assumptions:

1. Linear Metric Calibration: An attitude continuum can be partitioned into subjective intervals that appear equal in subjective magnitude to an unbiased observer. In the development of this scale, an initial pool of 130 candidate statements was compiled from diverse literary, philosophical, and personal sources. These statements were presented to a sorting panel of 300 judges, who were tasked with sorting each statement into one of 11 piles (labeled A to K), where pile A represented the most favorable orientation toward the church, pile F represented neutral or indifferent sentiment, and pile K represented the most unfavorable orientation.
2. Independence of Judgment: A foundational postulate of Thurstone’s method was that the scale value assigned to an item (its perceived degree of favorableness or unfavorableness) should be independent of the personal attitudes held by the judges themselves. Although subsequent psychometricians (notably Sherif and Hovland in their Social Judgment Theory) demonstrated that judges’ own reference points can systematically displace item perceptions, Thurstone and Chave maintained that median sorting values across a heterogeneous group of judges neutralize individual bias.
3. Scale Value Determination: For each candidate item, a cumulative frequency distribution of judge placements was constructed. The scale value (S) was formally defined as the median (the 50th percentile) of this distribution, representing the point along the 11-interval continuum above and below which exactly half of the judges categorized the statement.
4. Ambiguity and Irrelevance Indexing: The dispersion of judges’ ratings for any given item served as a direct empirical metric of that item’s semantic ambiguity. Thurstone utilized the interquartile range ($Q = Q_3 – Q_1$) as an index of ambiguity. Statements that yielded broad, scattered distributions (large $Q$ values) were diagnosed as ambiguous or subject to discordant interpretations and were eliminated. Items that elicited high consensus (small $Q$ values) were retained. Furthermore, items characterized by high internal irrelevance—those that failed to discriminate between individuals of varying attitude levels—were discarded through criterion-of-consistency checks.

From the original 130 statements, 45 items were retained. The authors systematically selected four statements to represent each of the 11 calibrated class-intervals across the continuum, plus one additional statement that fell into an extreme, twelfth psychological zone, ensuring uniform spacing and linear coverage throughout the psychological construct.

Validity

Thurstone and Chave established multiple layers of empirical and conceptual validity for the Attitude toward the Church Scale, providing an early model for psychometric instrument validation.

Known-Groups / Criterion-Related Validity

The scale was administered across divergent demographic and institutional cohorts whose expected attitudes were theoretically distinct. The observed empirical distributions conformed precisely to theoretical expectations:

• Theological Students: A cohort of 103 divinity students at the University of Chicago yielded the most homogeneous and overwhelmingly favorable scores, clustering sharply at the favorable end of the continuum with minimal dispersion.
• University Students: Samples of undergraduate classes (freshmen through seniors) and graduate students (totaling over 1,000 respondents) produced broad distributions centered around moderate-to-favorable scores, exhibiting slight bimodal tendencies that captured both traditional collegiate loyalties and emerging secular critiques.
• Civic Forum Participants: Respondents attending the Chicago Forum—a public-interest lecture platform convened by the Illinois League of Women Voters—exhibited substantially less favorable average scores than collegiate samples, accompanied by high internal variance that mirrored the politically and ideologically diverse nature of the urban audience.

Behavioral and Membership Correlates

Criterion validity was further substantiated through concrete self-reported behaviors. Respondents who reported active, frequent church attendance obtained significantly higher (more favorable) scale scores than non-attenders or irregular attenders. Similarly, individuals holding formal church membership scored substantially higher than non-members. In subsequent decades, landmark studies such as Wicker (1971) incorporated the scale into investigations of the attitude-behavior relationship, demonstrating that while overall institutional sentiment correlated with attendance, specific situational factors moderated actual physical participation.

Convergent Validity

Thurstone and Chave evaluated convergent validity by comparing participants’ scale-derived scores with their self-ratings on an independent, single-item graphical rating scale ranging from “strongly favorable to the church” to “strongly against the church.” The correlation between the comprehensive 45-item scale and this direct self-rating was moderately high and statistically robust. Interestingly, the authors observed that respondents—especially those who held negative underlying dispositions—tended to rate themselves slightly more favorably on the single direct rating than their endorsement of specific item statements revealed. Thurstone and Chave interpreted this discrepancy as the manifestation of normative social pressure, demonstrating that the multi-item objective scale was superior in circumventing social desirability bias.

Reliability

The internal consistency and measurement precision of the Attitude toward the Church Scale were evaluated using split-half reliability methods and parallel-form correlations:

• Split-Half Calibration: To construct parallel halves, the 45 statements were arranged in ascending order according to their empirically determined scale values. The researchers then split successive pairs of statements, alternating the assignment of the higher-scale item between Form A and Form B. The single remaining extreme item at the terminal boundary was assigned to both halves.
• Empirical Reliability Coefficients: When administered to an initial validation sample of 100 university students, the raw product-moment correlation between the two split halves was $r = .89$. When corrected for full-scale length using the Spearman-Brown formula, the estimated reliability coefficient reached $r_{xx} = .94$.
• Replication Sample: In a subsequent independent cohort of 200 university freshmen, the raw correlation between halves was $r = .85$, yielding a Spearman-Brown corrected reliability coefficient of $r_{xx} = .92$.

These values demonstrate that the scale possesses exceptional psychometric stability and internal consistency. Because the instrument uses an equal-appearing interval architecture rather than a summative Likert response format, traditional modern indices such as Cronbach’s alpha are technically less suited than split-half equivalence or test-retest indices; however, both the original empirical studies and later reassessments (e.g., Shaw & Wright, 1967; Robinson et al., 1999) confirm that the instrument reliably differentiates individuals along the attitudinal spectrum with minimal measurement error.

Factor Analysis

At the time of the scale’s construction in 1929, modern computational techniques for Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) had not yet been codified. Thurstone himself would soon develop Multiple Factor Analysis during the 1930s. Consequently, unidimensionality in the Attitude toward the Church Scale was originally established not through latent factor modeling, but through rigorous psychophysical sorting criteria, interquartile ambiguity filtering ($Q$-values), and the criterion of internal consistency.

In subsequent decades, as factor-analytic paradigms matured, psychometricians subjected Thurstone’s scales to empirical structural evaluation:

• Dominant General Factor: Factor-analytic reappraisals of the 45-item correlation matrix have consistently identified a massive primary general factor accounting for the bulk of common variance, corresponding to generalized institutional affect (favorableness vs. unfavorableness toward ecclesiastical authority and participation).
• Secondary Variance Dimensions: When multi-factor solutions are extracted, secondary factors emerge that mirror the conceptual sub-domains originally balanced by Thurstone and Chave: (a) Intellectual/Epistemic Relevance (items contrasting dogma with science and education), (b) Socio-Political Progressivism (items addressing wealth, reactionism, and institutional hypocrisy), and (c) Personal Fellowship/Spiritual Uplift (items regarding moral habituation, personal solace, and community life).
• Modern Item Response Theory (IRT) Alignments: Contemporary psychometric analyses view the equal-appearing interval mechanism as an early operationalization of non-monotonic or unimodal response functions, akin to Rasch models for unfoldment (e.g., Coombsian unfolding or Andrich’s hyperbolic cosine models). In this framework, respondents do not accumulate points monotonically as in a Likert scale; rather, they endorse statements whose calibrated location on the latent continuum matches their own latent trait position ($ heta$), confirming the scale’s fundamental unidimensional architectural integrity.

Instrument / Measurement Tool

The operational characteristics of the Attitude toward the Church Scale are summarized below:

• Test Type: Standardized self-report attitude scale based on the Psychophysical Method of Equal-Appearing Intervals.
• Format: Paper-and-pencil or digital administration; individual or group format; self-administered.
• Administration Time: Approximately 10 to 15 minutes; untimed.
• Item Count: 45 statements representing the entire evaluative continuum from extreme favorableness to extreme antagonism.
• Response Format: Non-cumulative endorsement. Respondents are instructed to place a checkmark next to every statement with which they fully agree and leave blank those with which they disagree or feel indifferent toward. Modifying statements is prohibited.
• Scoring Protocols: Scoring can be executed via either of two mathematically convergent methods:
  • Mean/Median Scale-Value Method: The examiner locates the pre-calibrated scale value (the median ranking determined by the 300 original judges) for each statement endorsed by the respondent. The individual’s final score is calculated as the arithmetic mean (or median) of these endorsed scale values.
  • Rank-Order Averaging Method: The 45 statements are ranked from 1 to 45 according to their reported scale values. The respondent’s score is computed as the mean rank order of the items endorsed.
• Transformed Scale and Score Interpretation: In the published standardized scoring manual, scale values were transformed by multiplying the original median rankings by 2.5 and rounding to the nearest whole integer. The resulting composite scores are interpreted using six primary evaluative categories:
  • Category 1 (0.0 – 2.9): Strongly favorable to the church.
  • Category 2 (3.0 – 4.9): Favorable to the church.
  • Category 3 (5.0 – 5.9): Mildly favorable to the church.
  • Category 4 (6.0 – 6.9): Mildly antagonistic toward the church.
  • Category 5 (7.0 – 8.9): Antagonistic toward the church.
  • Category 6 (9.0 – 11.0+): Strongly antagonistic toward the church.

Permissions & Fee and Test Year

The Attitude toward the Church Scale was originally developed and published in 1929 by the University of Chicago Press under the monograph title The Measurement of Attitude: A Psychophysical Method and Some Experiments with a Scale for Measuring Attitude toward the Church, authored by L. L. Thurstone and E. J. Chave. Supplementary forms and pedagogical applications were later cataloged in Chave’s 1939 volume, Measure Religion: Fifty-Two Experimental Forms.

Due to its publication date (prior to 1930) and the expiration of historical statutory copyrights, the instrument is in the public domain. Researchers, clinicians, and historians may reproduce, adapt, and administer the scale for academic, pedagogical, and scientific investigations without payment of licensing fees or formal publisher permissions, provided standard academic attribution is accorded to the original authors.

References

• Chave, E. J. (1939). Measure religion: Fifty-two experimental forms. University of Chicago Bookstore.
• Heber, S. (1996). Dimensionen der Religiosität: Skalen, Messmodelle und Ergebnisse einer empirisch orientierten Religionspsychologie. Universitätsverlag Freiburg.
• Jones, V. (1970). Attitudes of college students and their changes: A 37-year study. Genetic Psychology Monographs, 81, 3–80.
• Meredith, G. M. (1968). Personality correlates to religious belief systems. Psychological Reports, 23(3), 1039–1042. https://doi.org/10.2466/pr0.1968.23.3.1039
• Robinson, J. P., Shaver, P. R., & Wrightsman, L. S. (Eds.). (1999). Measures of political attitudes. Academic Press.
• Shaw, M. E., & Wright, J. M. (1967). Scales for the measurement of attitudes (pp. 97–99). McGraw-Hill.
• Siegman, A. W. (1962). A cross-cultural investigation of the relationship between religiosity, ethnic prejudice and authoritarianism. Psychological Reports, 11(2), 419–424. https://doi.org/10.2466/pr0.1962.11.2.419
• Struening, E., & Spilka, B. (1952). A study of certain social and religious attitudes of university faculty members. Psychological Newsletter, 43, 1–17.
• Thurstone, L. L. (Ed.). (1931). The measurement of social attitudes. University of Chicago Press.
• 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.
• Wicker, A. W. (1971). An examination of the “other variables” explanation of attitude-behavior inconsistency. Journal of Personality and Social Psychology, 19(1), 18–30. https://doi.org/10.1037/h0031090

Items of the Scale