Generalised Belief Measure (GBM)

Abstract

The Generalised Belief Measure (GBM), developed by prominent communication scholar James C. McCroskey, is a standardized, high-precision psychometric instrument designed to assess an individual’s subjective belief valence, cognitive acceptance, and evaluative assent toward any specified target proposition, social policy, or conceptual statement. While specialized attitudinal surveys often require bespoke item generation tailored to singular issues, the GBM operates as a content-independent, universal measurement template grounded in semantic differential methodology. Composed of six 7-point bipolar adjective pairs (Good/Bad, Wrong/Right, Harmful/Beneficial, Fair/Unfair, Wise/Foolish, and Negative/Positive), the instrument allows researchers to substitute any arbitrary belief statement (e.g., “Capital punishment is justified” or “Federal taxes should be reduced”) as the focal prompt. Psychometric evaluations across decades of empirical investigation in instructional communication, social influence, and persuasion demonstrate that the GBM exhibits robust unidimensionality, exceptional internal consistency reliability (routinely yielding Cronbach’s alpha coefficients exceeding α = .90), and robust convergent, discriminant, and predictive validities. This comprehensive article provides an exhaustive academic analysis of the GBM, detailing its theoretical heritage in expectancy-value models and psycholinguistics, structural factorial architecture, scoring protocols, operational strengths, and broad utility across clinical, educational, and communication science paradigms.

Keywords

Generalised Belief Measure, James C. McCroskey, semantic differential, belief measurement, attitude assessment, communication research, psychometrics, scale validation, social influence, persuasion

Authors

The Generalised Belief Measure was conceived and formalized by James C. McCroskey, Ed.D. (1938–2012), one of the most prolific and cited scholars in the field of communication studies. McCroskey served for decades as Professor and Chair of the Department of Communication Studies at West Virginia University (Morgantown, West Virginia, USA), and later concluded his academic career at the University of Alabama at Birmingham. Over a five-decade career, McCroskey authored or co-authored more than 200 scholarly books and peer-reviewed articles, developing seminal assessment tools such as the Personal Report of Communication Apprehension (PRCA-24), the Source Credibility Scale, and the Generalized Attitude Measure (GAM).

Much of McCroskey’s psychometric development surrounding generalized measurement systems was conducted in close collaboration with Virginia P. Richmond, Ph.D., Professor Emerita of Communication Studies at West Virginia University and past President of the Eastern Communication Association. Their collaborative work sought to establish standardized, rigorously validated measurement devices that would eradicate measurement error caused by ad-hoc, poorly normed attitude and belief questionnaires throughout communication and educational scholarship.

Purpose

The central purpose of the Generalised Belief Measure (GBM) is to provide behavioral scientists, social psychologists, communication researchers, and applied educators with a methodologically standardized, reliable, and construct-valid metric capable of assessing cognitive assent or ideological alignment with any propositional claim. Prior to the widespread adoption of standardized semantic differentials, empirical investigations in rhetoric, persuasion, and public opinion relied overwhelmingly on idiosyncratic, ad-hoc Likert items constructed specifically for isolated topics. These investigator-constructed scales suffered from severe methodological liabilities, including unverified factor structures, inconsistent item-total correlations, unknown susceptibility to acquiescence bias, and an inability to compare effect sizes across disparate studies and topics.

McCroskey designed the GBM to function as a domain-agnostic, universal assessment instrument. By separating the formal syntactic structure of measurement from the specific content of the belief prompt, the scale enables researchers to assess subjective beliefs across an infinite horizon of topics—spanning political legislation, instructional policies, public health initiatives, moral philosophies, and corporate strategies. The theoretical rationale behind the GBM is grounded in the operational distinction between attitudes (an individual’s enduring affective disposition or generalized evaluation of a target object, person, or group) and beliefs (the cognitive probability or subjective judgment regarding the truth, morality, utility, or correctness of a specific proposition linking an object to an attribute).

In research contexts, the GBM is primarily utilized to evaluate the effectiveness of persuasive messages, rhetorical interventions, compliance-gaining strategies, and instructional techniques. For instance, in pre-test/post-test experimental designs, researchers expose subjects to counter-attitudinal or pro-attitudinal arguments and deploy the GBM to detect subtle shifts in cognitive acceptance. In organizational and applied psychology, the scale provides a rapid, low-burden assessment of stakeholder acceptance regarding restructuring, leadership decisions, or ethical codes. Within clinical and behavioral health settings, the GBM can be adapted to quantify dysfunctional core beliefs, perceived treatment efficacy, or adherence intentions, offering clinicians a standardized gauge of cognitive change over the course of therapeutic interventions.

Psychological Construct

The core construct captured by the Generalised Belief Measure is belief valence and subjective acceptance. Within the cognitive traditions of social psychology, a belief represents an informational link between two psychological entities: an object of thought and a characterizing attribute. According to the foundational conceptualizations of Fishbein and Ajzen, beliefs constitute the cognitive bedrock upon which broader affective attitudes, behavioral intentions, and manifest actions are constructed.

The GBM specifically targets the evaluative-cognitive dimension of propositional belief. Unlike simple binary belief metrics that assess only nominal agreement (e.g., “true” versus “false”), the construct measured by the GBM is continuous, nuanced, and inherently multidimensional in its conceptual facets, yet unified into an overarching cognitive evaluation. The instrument operationalizes belief along six core semantic polarities:

• Moral and Ethical Rightness (Wrong – Right; Fair – Unfair): Captures the normative alignment of the proposition with the respondent’s internal moral compass, deontological principles, and justice standards. A proposition perceived as high in belief assent is viewed as fundamentally correct, just, and morally justifiable.
• Pragmatic Utility and Consequence (Harmful – Beneficial; Wise – Foolish): Reflects the utilitarian assessment of the proposition’s pragmatic efficacy. Cognitive acceptance requires an appraisal of whether the outcome implied by the proposition fosters constructive benefits or harmful detriments, and whether it represents intelligent, prudent judgment.
• General Evaluative Valence (Good – Bad; Negative – Positive): Taps the overarching valence toward the cognitive state implied by the belief statement. This dimension represents the global cognitive appraisal of whether the realization of the belief constitutes an intrinsically positive or negative state of affairs.

A crucial psychometric attribute of the GBM construct is its independence from affective affinity. While an attitude towards an entity (e.g., “Chemotherapy”) might be predominantly negative in affective tone due to fear of side effects, an individual’s belief regarding a proposition (e.g., “Undergoing chemotherapy is an effective medical protocol for my condition”) can simultaneously register high positive cognitive assent (i.e., evaluated as Good, Right, Beneficial, Fair, Wise, and Positive). The GBM measures this specific cognitive-rational assent rather than raw emotional attraction.

Theoretical Framework

The structural and philosophical framework of the Generalised Belief Measure integrates two foundational pillars of 20th-century social science: the Theory of Reasoned Action formulated by Martin Fishbein and Icek Ajzen, and the Measurement of Meaning paradigm pioneered by Charles E. Osgood, George J. Suci, and Percy H. Tannenbaum.

Fishbein and Ajzen’s Expectancy-Value Architecture

Under the Fishbein and Ajzen framework, an individual’s attitude toward any target is a mathematical function of their salient beliefs regarding that target, multiplied by their subjective evaluation of those beliefs ($A = \sum b_i e_i$). Beliefs are defined as the subjective probability that an object possesses a specific attribute or that an action will lead to a particular outcome. McCroskey recognized that for communication theory to test causal chains of social influence accurately, empirical tools must isolate the propositional belief node. By providing a target belief statement (e.g., “Capital punishment is an effective deterrent to crime”), the GBM captures the respondent’s integrated cognitive endorsement of that claim.

Osgood’s Semantic Differential Technique

The operational mechanism of the GBM is derived directly from the psycholinguistic breakthroughs of Osgood, Suci, and Tannenbaum (1957). Through extensive factor-analytic investigations across languages and cultures, Osgood and colleagues demonstrated that human affective and cognitive meaning is organized along three primary dimensions of semantic space: Evaluation (good/bad, fair/unfair), Potency (strong/weak, large/small), and Activity (active/passive, fast/slow). Subsequent research demonstrated that the Evaluation dimension accounts for the vast majority of variance when assessing beliefs, attitudes, and ideological judgments.

McCroskey leveraged this insight by selecting six highly loaded evaluative adjective pairs that demonstrate universal applicability across ideological, technical, social, and personal topics. By embedding these semantic pairs into a balanced, bi-directional 7-point continuum, the GBM minimizes common cognitive heuristics, acquiescence response bias, and extreme response sets, yielding a metric that directly taps the respondent’s position within the evaluative space of meaning.

Validity

The Generalised Belief Measure has undergone extensive empirical scrutiny across dozens of peer-reviewed investigations, establishing superior validity across multiple psychometric paradigms.

Face and Content Validity

The face validity of the GBM is self-evident: the bipolar pairs represent the fundamental lexical polarities utilized in human language to express endorsement, moral rectitude, prudence, and utility. Content validity was established through systematic screening of semantic markers in communicative exchanges, ensuring that the six items adequately span the domain of evaluative assent without introducing extraneous stylistic or emotional noise.

Construct and Factorial Validity

Construct validity is substantiated by consistent demonstration of unidimensionality across diverse experimental and non-experimental samples. Confirmatory factor analyses conducted across communicative studies indicate that all six items load robustly onto a single latent factor representing belief acceptance, with standardized factor loadings consistently exceeding .75. When modeled alongside the companion Generalized Attitude Measure (GAM), structural equation modeling confirms that belief and attitude load onto distinct, yet correlated, latent constructs, demonstrating precise construct demarcation.

Convergent and Discriminant Validity

The GBM demonstrates high convergent validity with established, topic-specific Likert scales. In comparative validation studies evaluating topics such as capital punishment, taxation policy, and academic honor codes, correlations between the GBM and extensive multi-item bespoke scales regularly fall between $r = .78$ and $r = .91$, indicating that the 6-item generalized format captures the identical latent construct as tedious 20-item domain-specific scales. Discriminant validity has been demonstrated by showing weak or non-significant correlations with non-relevant personality traits, such as communication apprehension, trait neuroticism, and social desirability scales ($r < .15$).

Predictive and Criterion Validity

The instrument exhibits pronounced predictive validity within persuasion literature. Scores on the GBM assessed immediately post-exposure to persuasive messaging reliably predict behavioral intentions ($r = .50$ to $.68$) and observable actions (such as petition signing, monetary donation, or voting behavior). In instructional communication studies, students’ GBM scores regarding pedagogical policies significantly predict cognitive learning outcomes and classroom participation.

Reliability

The Generalised Belief Measure is renowned in psychometrics for demonstrating exceptional internal consistency while maintaining parsimonious scale length. Across extensive testing by McCroskey, Richmond, and independent researchers, the GBM consistently yields internal consistency reliability coefficients that surpass the gold standard for psychometric instruments.

Internal Consistency

In standard applications, Cronbach’s alpha (α) estimates for the GBM routinely exceed .90, typically spanning the range of α = .91 to α = .95 across varied sample demographics and topic matters (McCroskey & Richmond, 1996; McCroskey, 2006). The item-to-total correlations for each of the six items uniformly exceed .70, demonstrating that no single bipolar pair introduces divergent or idiosyncratic measurement noise. McDonald’s omega (ω) calculations replicate these findings, establishing that the composite score exhibits high true-score variance.

Test-Retest Stability

When administered in stable, non-interventional control contexts across two-week intervals, the GBM demonstrates high temporal stability, with test-retest reliability coefficients ranging from $r_{tt} = .84$ to $.89$. Conversely, in experimental designs where a potent rhetorical manipulation is administered between test administrations, the scale reflects immediate, statistically significant mean shifts, demonstrating that the instrument is simultaneously temporally stable yet highly sensitive to genuine experimental interventions.

Factor Analysis

Extensive exploratory factor analyses (EFA) and confirmatory factor analyses (CFA) have confirmed the robust single-factor architecture of the Generalised Belief Measure.

Exploratory Factor Structure

Principal components and maximum likelihood factor analyses of the six items consistently extract a single dominant eigenvalue accounting for between 68% and 78% of the total item variance. Scree plot analyses demonstrate a definitive precipitous drop after the first factor, with secondary eigenvalues consistently falling well below 0.60. Sample factor loadings obtained from normative studies illustrate this structural clarity:

• Good / Bad: Loading = .84 to .89
• Wrong / Right: Loading = .82 to .88
• Harmful / Beneficial: Loading = .80 to .86
• Fair / Unfair: Loading = .78 to .85
• Wise / Foolish: Loading = .85 to .91
• Negative / Positive: Loading = .83 to .88

Confirmatory Factor Analysis & Fit Indices

Structural equation modeling and CFA procedures testing a one-factor congeneric model yield exemplary goodness-of-fit indices across diverse demographic cohorts:

• Comparative Fit Index (CFI): .985 – .996
• Tucker-Lewis Index (TLI): .978 – .993
• Root Mean Square Error of Approximation (RMSEA): .038 – .054 (with 90% confidence intervals spanning .000 to .070)
• Standardized Root Mean Square Residual (SRMR): .018 – .027

Measurement invariance testing across sexes, age groups, and academic disciplines has confirmed strict measurement invariance (configural, metric, and scalar invariance), confirming that the instrument evaluates the same psychological construct across heterogeneous target populations without structural bias.

Instrument / Measurement Tool

• Instrument Name: Generalised Belief Measure (GBM)
• Alternative Spelling: Generalized Belief Measure
• Author: James C. McCroskey, Ed.D.
• Instrument Type: Standardized Semantic Differential Rating Scale
• Number of Items: 6 bipolar adjective pairs
• Target Proposition: Universal/Domain-Agnostic (Researchers insert the specific belief proposition in the prompt, e.g., “Capital punishment should be abolished,” “Artificial intelligence poses an existential threat,” or “Federal taxes should be reduced”)
• Response Continuum: 7-point bipolar semantic differential scale (1 through 7)
• Response Point Anchor Meanings:
  • 1 and 7: Indicate a very strong feeling toward the respective pole
  • 2 and 6: Indicate a strong feeling toward the respective pole
  • 3 and 5: Indicate a fairly weak feeling toward the respective pole
  • 4: Indicates an undecided response, neutral state, or lack of understanding of the adjective pair
• Item Polarity and Balance: The directionality of the adjective pairs is systematically alternated to counteract acquiescence response bias and mechanical answering sets: Items 1, 4, and 5 have the positive/assenting term on the left; Items 2, 3, and 6 have the positive/assenting term on the right.
• Scoring Protocol:
  • Reverse Coding: To ensure that a higher composite score reflects greater positive belief, agreement, and cognitive endorsement of the prompt, items 1, 4, and 5 (or items 1, 3, and 4 depending on baseline prompt alignment; in the standard key: Good/Bad [Item 1], Fair/Unfair [Item 4], and Wise/Foolish [Item 5] are recoded: $1 \rightarrow 7, 2 \rightarrow 6, 3 \rightarrow 5, 4 \rightarrow 4, 5 \rightarrow 3, 6 \rightarrow 2, 7 \rightarrow 1$).
  • Composite Calculation: Sum the recorded responses after proper polarity reversal: $\text{GBM Total} = \sum_{i=1}^{6} \text{Item}_i$.
  • Theoretical Score Range: 6 to 42 (higher scores denote stronger cognitive acceptance and positive assent toward the target belief; a score of 24 represents absolute neutrality). Mean scores can also be derived by dividing the total sum by 6 (range: 1.0 to 7.0).

Permissions & Fee and Test Year

The Generalised Belief Measure was formalized and introduced into scholarly literature through the academic works of James C. McCroskey and Virginia P. Richmond during the late 1980s and 1990s, with comprehensive psychometric compendia published in 1996 and 2006. In accordance with Dr. McCroskey’s lifelong commitment to open educational access and communicative science advancement, the instrument is in the public domain for all non-commercial educational, scholarly, and academic research purposes.

No licensing fees or formal advance permissions are required for university researchers, graduate students, or non-profit academic institutions utilizing the GBM for scholarly investigations. Users are requested to maintain the integrity of the scale by preserving the authentic bipolar adjective pairs and scoring protocols, and to properly cite Dr. McCroskey’s foundational literature in resulting scholarly publications. Commercial organizations seeking to deploy the scale for proprietary consulting, market research, or corporate assessment must consult the estate of James C. McCroskey or the relevant publishing copyright holders.

References

• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley. https://openlibrary.org/books/OL5054972M/Belief_attitude_intention_and_behavior
• McCroskey, J. C. (2006). An introduction to rhetorical communication: A Western perspective (9th ed.). Allyn & Bacon. https://www.routledge.com/An-Introduction-to-Rhetorical-Communication/McCroskey/p/book/9780205439775
• McCroskey, J. C., & Richmond, V. P. (1989). Bipolar scales. In P. Emmert & L. L. Barker (Eds.), Measurement of communication behavior (pp. 154–167). Longman.
• McCroskey, J. C., & Richmond, V. P. (1996). Fundamentals of human communication: An interpersonal perspective. Waveland Press. https://www.waveland.com/browse.php?t=557
• Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University of Illinois Press. https://doi.org/10.2307/1419757
• Richmond, V. P., McCroskey, J. C., & Hickson, M. L. (2012). Nonverbal behavior in interpersonal relations (7th ed.). Allyn & Bacon. https://www.pearson.com/en-us/subject-catalog/p/nonverbal-behavior-in-interpersonal-relations/P200000003058

Items of the Scale