Ad Believability (ADBEL)

1. Abstract

The Ad Believability Scale (commonly designated as ADBEL or the Advertising Perceived Believability Scale) is a foundational psychometric instrument developed by consumer psychologist Richard F. Beltramini in 1982. Designed to quantify consumer skepticism and informational acceptance, the instrument evaluates the degree to which commercial, public service, or institutional promotional communications are perceived as trustworthy, credible, and truthful. The ADBEL consists of 10 bipolar semantic differential adjective pairs evaluated on a 7-point scale. Across four decades of empirical application in advertising research, behavioral economics, and communication studies, the scale has consistently demonstrated robust psychometric integrity. Extant literature reveals exceptionally high internal consistency, with Cronbach’s alpha coefficients regularly exceeding .90 (ranging from .88 to .96 across diverse consumer samples and message modalities). Exploratory and confirmatory factor analyses consistently substantiate a parsimonious unidimensional factor structure, explaining between 65% and 78% of the common variance. The scale exhibits strong convergent validity with measures of source credibility, attitude toward the ad ($A_{ad}$), attitude toward the brand ($A_b$), and purchase intentions, as well as distinct discriminant validity against generalized marketplace skepticism and trait cynicism. This article delivers an exhaustive academic review of the ADBEL instrument, examining its theoretical lineage within cognitive response models and persuasion frameworks, its latent construct dimensionality, empirical validity, statistical reliability parameters, structural factor configurations, and operational administration protocols.

2. Keywords

Ad Believability, ADBEL, advertising perceived believability, message credibility, consumer skepticism, source credibility, semantic differential, advertising effectiveness, cognitive processing, persuasion knowledge, psychometrics, scale validation

3. Authors

The scale was conceived, developed, and empirically validated by:

• Richard F. Beltramini, Ph.D. — Professor Emeritus of Marketing, Wayne State University (Mike Ilitch School of Business, Detroit, Michigan, USA), and formerly affiliated with Arizona State University (Tempe, Arizona). Dr. Beltramini is a nationally recognized scholar in marketing communication, advertising ethics, consumer decision-making, and marketing research methodology.
• Subsequent Validation Collaborators: Key extensions and cross-validations of the instrument were conducted in collaboration with Kenneth R. Evans (University of Missouri / University of Oklahoma), among other prominent marketing scholars who examined the scale’s performance across varied media formats, including comparative advertising, health warning disclosures, and digital messaging.

4. Purpose

The primary purpose of the Ad Believability Scale is to measure the extent to which an audience perceives the claims, executions, and informational content of a specific advertisement to be truthful, credible, and acceptable as factual reality. Beltramini engineered this instrument in response to a pervasive methodological deficit in early consumer behavior research: investigators frequently conflated attitude toward the advertisement ($A_{ad}$), which is largely an affective evaluation (e.g., entertaining, irritating, pleasant), with the cognitive acceptance of message verity (whether the consumer actually believes the claims being made).

Understanding perceived believability is critical across both commercial and public-policy domains for several structural reasons:

• Mediation of Persuasion: Under standard models of cognitive processing, advertising assertions cannot alter downstream beliefs, attitudes, or purchase intentions if the target audience fundamentally discounts the claims as deceptive, exaggerated, or spurious. The ADBEL scale acts as a sensitive diagnostic filter for identifying counterarguing and psychological reactance.
• Regulatory and Ethical Oversight: Regulatory bodies such as the Federal Trade Commission (FTC) require empirical standards to evaluate deceptive advertising, substantiation of health claims, and the effectiveness of mandated corrective statements. The ADBEL scale provides an objective, standardized metric for determining whether corrective ads, product warning labels, or risk disclosures effectively alter consumer perceptions of claim truthfulness.
• Public Health Interventions: In anti-smoking, anti-vaping, substance abuse, and nutritional public service announcements (PSAs), message rejection frequently stems from perceived incredulity. The scale enables health communication researchers to optimize message framing, ensuring that target demographics perceive clinical risk statements as authentic and conclusive rather than hyperbole.
• Mitigation of Marketplace Skepticism: Modern consumer environments are characterized by high baseline cynicism toward corporate advocacy. By deploying the ADBEL scale, researchers can systematically examine how varying executional cues (e.g., endorser types, objective third-party certifications, comparative claim structures, visual rhetoric) either temper or inflame inherent consumer skepticism.

5. Psychological Construct

The core psychological construct measured by the ADBEL scale is perceived advertising believability. Beltramini defined this construct as the extent to which an individual consumer assesses an advertising message—along with its central and peripheral claims—as possessing credibility, integrity, and verisimilitude. It is a psychological state occurring at the intersection of cognitive appraisal, epistemic evaluation, and source evaluation.

To fully conceptualize perceived believability, it must be distinguished from related but distinct psychological and consumer constructs:

• Believability vs. Truth: Objective truth represents an external, verifiable reality regarding a product attribute or service outcome. Perceived believability, by contrast, is entirely phenomenological; it reflects the subjective reality constructed within the consumer’s cognitive processing system. An objectively truthful claim may be scored low on the ADBEL if it contradicts intuitive heuristics, whereas an objectively false claim supported by slick production cues or uncritical bias may achieve high believability scores.
• Believability vs. Skepticism: As demonstrated by Obermiller and Spangenberg (1998), skepticism toward advertising is a generalized, enduring individual trait reflecting a stable predisposition to disbelieve commercial communications in general. ADBEL, conversely, measures a state-specific, message-level cognitive appraisal directed toward a single, concrete communicative execution. While highly skeptical individuals baseline lower on ADBEL, the instrument captures the specific communicative efficacy of the stimulus at hand.
• Believability vs. Attitude toward the Ad ($A_{ad}$): Attitude toward the ad represents a global affective response reflecting feelings of likeability, entertainment value, or creative execution. A consumer may thoroughly enjoy an ad (high $A_{ad}$) because of humorous animation or celebrity antics while assigning it a near-zero believability score on the ADBEL because the underlying product performance claims are recognized as hyperbole or puffery.

The construct encompasses several interrelated cognitive-epistemic facets operationalized across its 10 bipolar pairings:

1. Veracity and Epistemic Probability: Evaluated by items such as Unbelievable / Believable and Unlikely / Likely. The consumer assesses the sheer likelihood that the portrayed scenario or stated benefit exists or could occur in real-world conditions.
2. Communicator Integrity and Authenticity: Captured by Dishonest / Honest and Not authentic / Authentic. This taps into perceived intent, evaluating whether the sponsor is operating in good faith or actively engaging in calculated manipulation, omission, or greenwashing.
3. Argument Quality and Evidentiary Weight: Operationalized through Not convincing / Convincing, Inconclusive / Conclusive, and Questionable / Unquestionable. This reflects the epistemic rigor of the data, demonstrations, or testimonials presented in the advertisement.
4. Logical Plausibility: Assessed through Unreasonable / Reasonable and Not credible / Credible, tapping whether the proposition conforms to rational expectations of product capabilities and market costs.

6. Theoretical Framework

The Ad Believability Scale is grounded in seminal social-cognitive paradigms of persuasion, communication theory, and message reception. Its psychological foundations draw predominantly upon four major theoretical architectures:

The Elaboration Likelihood Model (ELM)

Formulated by Richard Petty and John Cacioppo (1986), the ELM posits two distinct routes to persuasion: the central route and the peripheral route. Perceived believability functions differently depending on the route activated. Under high involvement, recipients possess both the motivation and ability to scrutinize message arguments; here, ADBEL scores reflect central-route processing, where the strength, logical consistency, and empirical evidence of the claim directly dictate believability. Under low involvement, believability is often heuristically inferred via peripheral cues (e.g., endorser status, company reputation, production quality). The ADBEL scale captures the final integrated cognitive output of this dual-route evaluation.

The Persuasion Knowledge Model (PKM)

Developed by Marian Friestad and Peter Wright (1994), the Persuasion Knowledge Model delineates how consumers construct and deploy dynamic knowledge about persuasion tactics to maintain personal agency and evaluate agent motives. When an advertisement triggers persuasion knowledge (e.g., highlighting overtly manipulative emotional appeals or exaggerated claims), the consumer activates “change of meaning” heuristics. The claims are no longer evaluated merely as product information; instead, they are recognized as strategic gambits. The ADBEL scale directly registers this activation: as persuasion knowledge triggers defensive counterarguing, ADBEL scores drop precipitously across items such as Dishonest / Honest and Questionable / Unquestionable.

Attribution Theory and Discounting Principles

Rooted in Harold Kelley’s (1973) attributional framework, consumers continuously formulate causal attributions regarding why an advertiser makes a specific claim. If a claim is attributed to situational commercial motives (i.e., “They are only saying this to sell products and maximize revenue”), the consumer invokes the discounting principle, lowering their assessment of the communication’s validity. Conversely, if two-sided messages are utilized (where an advertiser admits a minor flaw or limitation), consumers attribute the communication to the advertiser’s dispositional honesty, prompting the augmentation principle and dramatically elevating ADBEL scores.

Hovland’s Source Credibility Model

The foundational Yale Communication Research Program led by Carl Hovland demonstrated that the acceptance of any persuasive message is heavily dependent on two fundamental dimensions: perceived expertise (the epistemic authority of the source) and perceived trustworthiness (the objective integrity and lack of bias of the source). Beltramini’s ADBEL scale serves as the message-level operational analog to Hovland’s source-level framework, measuring whether the communicative output itself has succeeded in projecting expertise and trustworthiness onto the advertised proposition.

7. Validity

The ADBEL scale has undergone rigorous empirical validation across diverse sampling frames, media contexts (print, radio, television, online display, influencer content), and product categories (consumer packaged goods, prescription pharmaceuticals, tobacco warnings, ethical eco-labels).

Construct and Factorial Validity

Beltramini (1982) and Beltramini and Evans (1985) established the construct validity of the scale by demonstrating that all 10 items converge onto a single primary factor representing perceived believability. CFA models have consistently shown that the scale items load strongly on their hypothesized latent construct, with standardized loadings uniformly exceeding the conventional .70 threshold, thus confirming robust construct representation without redundant conceptual drift.

Convergent Validity

Convergent validity has been repeatedly demonstrated through significant, theory-consistent correlations with established psychometric scales. ADBEL scores correlate positively and significantly with:

• Source Credibility Measures ($r = .65$ to $.82$, $p < .001$), confirming that messages produced by credible sources are appraised as more believable.
• Attitude toward the Advertisement ($A_{ad}$) ($r = .52$ to $.71$, $p < .001$), supporting the premise that message acceptance substantially contributes to positive ad evaluations.
• Brand Belief Acceptance and Purchase Intentions ($r = .40$ to $.62$, $p < .001$), verifying that higher believability directly facilitates behavioral intent.

Discriminant Validity

Discriminant validity has been demonstrated by contrasting ADBEL against trait-level consumer skepticism and generalized distrust. While the Scale of Skepticism Toward Advertising (SKEP; Obermiller & Spangenberg, 1998) measures a persistent individual orientation, ADBEL successfully captures within-subject variance when the same individual evaluates ads of varying quality. Average Variance Extracted (AVE) values for the ADBEL construct consistently exceed .60, outstripping the squared correlations ($r^2$) between ADBEL and related constructs such as ad likeability, message complexity, or brand familiarity, satisfying the rigorous Fornell-Larcker criterion.

Criterion and Predictive Validity

Predictive validity is firmly evidenced in experiments manipulating claim substantiation. When researchers present identical product claims with vs. without independent scientific certification (e.g., FDA approval notices, Underwriters Laboratories certifications), the ADBEL scale demonstrates sharp, statistically significant sensitivity ($F$-ratios typically significant at $p < .001$), yielding vastly higher believability scores for substantiated claims. Furthermore, ADBEL scores have been shown to predict long-term cognitive retention of product benefits and subsequent brand choice in simulated supermarket environments.

8. Reliability

The statistical reliability of the Ad Believability Scale is among the highest reported for semantic differential instruments in the behavioral sciences.

Internal Consistency

• Cronbach’s Alpha ($lpha$): In Beltramini’s original development and cross-validation studies (1982, 1985), the internal consistency reliability coefficients ranged between .93 and .96. Subsequent replications across four decades have documented alpha values that rarely fall below .88, consistently positioning the scale well above the .70 threshold recommended for basic research and the .90 threshold recommended for clinical or high-stakes applied diagnostics.
• Composite Reliability (CR): Structural equation modeling investigations yield Composite Reliability values typically ranging between .92 and .97, confirming that the high alpha is not merely an artifact of item redundancy but a consequence of robust latent factor saturation.

Stability and Test-Retest Reliability

In stability studies evaluating message evaluations separated by intervals of two to four weeks (utilizing neutral, non-volatile stimulus advertisements), test-retest reliability coefficients have been observed in the $r = .78$ to $.85$ range ($p < .001$). This confirms that while the scale is sensitive to situational message differences, it reliably captures stable perceptual evaluations of a given static communication over time.

Split-Half and Inter-Item Correlations

Corrected item-total correlations across the 10 items typically range between .68 and .88. The Spearman-Brown split-half reliability coefficient consistently exceeds .90, establishing that both halves of the instrument possess equivalent psychometric strength.

9. Factor Analysis

The structural dimensionality of the ADBEL scale has been extensively scrutinized using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

In initial principal components analyses conducted with varimax or oblimin rotations, the 10 bipolar pairs repeatedly load onto a single dominant factor. Eigenvalue evaluation reveals a solitary primary eigenvalue substantially greater than 1.0 (typically ranging between 6.20 and 7.50), while the secondary eigenvalue typically hovers well below 0.80. A typical scree plot inspection exhibits a sharp, unambiguous drop after the first factor. This single factor routinely accounts for 65% to 78% of the total variance across diverse experimental conditions.

Standardized factor loadings across all 10 items are exceptionally balanced:

• Unbelievable / Believable: $.82 – .91$
• Untrustworthy / Trustworthy: $.80 – .89$
• Not convincing / Convincing: $.78 – .87$
• Not credible / Credible: $.84 – .92$
• Unreasonable / Reasonable: $.70 – .81$
• Dishonest / Honest: $.76 – .86$
• Questionable / Unquestionable: $.72 – .84$
• Inconclusive / Conclusive: $.68 – .79$
• Not authentic / Authentic: $.75 – .85$
• Unlikely / Likely: $.71 – .82$

Confirmatory Factor Analysis (CFA)

Structural equation modeling studies assessing the one-factor measurement model have validated its structural fit across multiple populations. When evaluated under Maximum Likelihood estimation, the unidimensional model exhibits excellent fit statistics:

• Comparative Fit Index (CFI): $.95 – .98$ (exceeding the standard $.95$ cutoff)
• Tucker-Lewis Index (TLI): $.94 – .97$
• Root Mean Square Error of Approximation (RMSEA): $.045 – .068$ (with 90% confidence intervals well within the acceptable $< .08$ ceiling)
• Standardized Root Mean Square Residual (SRMR): $.025 – .042$

Attempts to split the scale into secondary oblique factors (such as “Cognitive Truth” vs. “Moral Trustworthiness”) generally result in extremely high inter-factor correlations ($r > .85$), confirming that postulating multidimensionality introduces unnecessary structural complexity without meaningful empirical gains.

10. Instrument / Measurement Tool

Below is the structured technical profile for administering, scoring, and interpreting the ADBEL scale:

• Test Type: Self-administered psychological / psychometric evaluation inventory.
• Format: Bipolar semantic differential format consisting of 10 adjective pairs separated by a continuous or discrete 7-interval numerical spectrum.
• Item Count: 10 items.
• Response Scale: 7-point semantic differential scale (e.g., 1 to 7 between bipolar adjective pairs).
• Administration Modality: Paper-and-pencil, computer-assisted self-interviewing (CASI), or online survey platforms (e.g., Qualtrics, Gorilla, MTurk/Prolific testing).
• Target Population: Adolescents and adults (typically ages 14 and older) possessing basic functional literacy.
• Administration Time: Approximately 1.5 to 3 minutes following stimulus exposure.
• Scoring Instructions:
  • Each item is scored from 1 (most negative pole: e.g., Unbelievable, Dishonest) to 7 (most positive pole: e.g., Believable, Honest).
  • In physical questionnaire forms, the poles should be systematically counterbalanced or randomized (e.g., placing some positive adjectives on the left) to mitigate acquiescence bias and straight-lining.
  • Prior to computing summary statistics, ensure that all counterbalanced items are reverse-coded so that higher scores always reflect greater perceived believability.
  • Composite Scoring: Researchers may compute either an overall sum score (ranging from 10 to 70) or a grand mean score across the 10 items (ranging from 1.0 to 7.0).
• Interpretation Norms:
  • Mean 1.00 – 2.99 (Sum 10 – 29): Low Believability / High Incredulity. Message is actively rejected; high likelihood of counterarguing or perceived deception.
  • Mean 3.00 – 4.99 (Sum 30 – 49): Ambivalent / Moderate Believability. Message elicits uncertainty; audience remains unconvinced, questioning validity or substantiation.
  • Mean 5.00 – 7.00 (Sum 50 – 70): High Believability / Informational Acceptance. Message is accepted as authentic, truthful, reasonable, and convincing.

11. Permissions & Fee and Test Year

• Year of Initial Publication: 1982 (formally published in the Proceedings of the Southwestern Marketing Association; further validated in the Journal of Advertising in 1985).
• Intellectual Property & Licensing: The scale was developed under standard academic auspices. It is considered an open-access psychometric instrument for academic research, educational use, and non-commercial scientific inquiry. Academic researchers may utilize the scale without royalty fees, provided appropriate scholarly attribution is cited.
• Commercial Applications: Commercial market research firms or commercial entities utilizing the scale within proprietary testing batteries should verify institutional permissions and adhere to ethical fair-use citation protocols.

12. References

• Beltramini, R. F. (1982). Advertising perceived believability scale. In D. R. Corrigan, F. B. Kraft, & R. H. Ross (Eds.), Proceedings of the Southwestern Marketing Association (pp. 1–3). Southeast Missouri State University.
• Beltramini, R. F., & Evans, K. R. (1985). Assessment of student attitudes toward advertising: A review and extension. Journal of Advertising, 14(3), 5–10. https://doi.org/10.1080/00913367.1985.10672951
• Friestad, M., & Wright, P. (1994). The Persuasion Knowledge Model: How people cope with persuasion attempts. Journal of Consumer Research, 21(1), 1–31. https://doi.org/10.1086/209380
• Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion: Psychological studies of opinion change. Yale University Press.
• Kelley, H. H. (1973). The processes of causal attribution. American Psychologist, 28(2), 107–128. https://doi.org/10.1037/h0034225
• MacKenzie, S. B., & Lutz, R. J. (1989). An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. Journal of Marketing, 53(2), 48–65. https://doi.org/10.1177/002224298905300204
• Obermiller, C., & Spangenberg, E. R. (1998). Development of a scale to measure skepticism toward advertising. Journal of Consumer Psychology, 7(2), 159–186. https://doi.org/10.1207/s15327663jcp0702_03
• Petty, R. E., & Cacioppo, J. T. (1986). The Elaboration Likelihood Model of persuasion. Advances in Experimental Social Psychology, 19, 123–205. https://doi.org/10.1016/S0065-2601(08)60214-2

13. Items of the Scale