Ad Affective Evaluation (Semantic Differential) (AASD)

1. Abstract

The Ad Affective Evaluation (Semantic Differential) (AASD) is a foundational psychometric instrument designed to assess consumers’ affective, aesthetic, and emotional evaluations of commercial advertisements. Originally developed by Michael J. Baker and Gilbert A. Churchill, Jr. in their seminal 1977 investigation of model physical attractiveness in advertising, the AASD isolates the affective component of the broader Attitude toward the Advertisement ($A_{ad}$) construct. Comprising 11 bipolar adjective pairs formatted across a seven-point semantic differential scale, the instrument operationalizes hedonic valence, aesthetic appreciation, dynamic visual engagement, and perceived affective sincerity, deliberately decoupling emotional resonance from cognitive or informational appraisals of ad content.

Psychometrically, the AASD demonstrates outstanding reliability, consistently yielding internal consistency coefficients (Cronbach’s $\alpha$) exceeding .88 across diverse media modalities, including print, broadcast, and interactive digital advertising. Extensive construct, convergent, discriminant, and predictive validity analyses have confirmed that affective evaluations captured by this scale directly mediate downstream marketing variables, notably brand attitudes ($A_b$) and consumer purchase intentions ($PI$). The scale represents an indispensable measurement standard within consumer psychology, advertising effectiveness research, sensory marketing, and applied neuromarketing studies.

2. Keywords

Attitude toward the ad, affective evaluation, semantic differential, consumer psychology, advertising effectiveness, Baker and Churchill, hedonic response, emotional appraisal, psychometrics, marketing communications, model attractiveness, brand attitude

3. Authors

The Ad Affective Evaluation scale was formulated by two pioneering scholars in marketing research methodology and consumer behavior:

• Michael J. Baker, TD, PhD, FCIM, FRSA: Emeritus Professor of Marketing at the University of Strathclyde, Glasgow, Scotland. Professor Baker was a founding figure in European marketing education, the founder of the Journal of Marketing Management, and author or editor of over 50 foundational texts on marketing theory, strategy, and buyer behavior.
• Gilbert A. Churchill, Jr., PhD: Emeritus Professor of Marketing at the University of Wisconsin–Madison. A recipient of the American Marketing Association’s highest honors (including the Paul D. Converse Award and the Charles Coolidge Parlin Award), Dr. Churchill revolutionized psychometric measurement in business disciplines through his landmark 1979 paper on developing better measures of marketing constructs.

4. Purpose

The primary purpose of the Ad Affective Evaluation (Semantic Differential) is to quantify the immediate, non-cognitive, emotional, and aesthetic reactions that consumers experience upon exposure to an advertising stimulus. In advertising theory, consumer reactions to marketing communications were historically conceptualized through strictly rational, cognitive-hierarchy models (such as the traditional AIDA framework: Attention, Interest, Desire, Action), which presumed that consumers systematically parse product claims, evaluate brand attributes, and calculate utility.

Baker and Churchill (1977) challenged this narrow paradigm by examining how extrinsic visual cues—specifically the physical attractiveness and gender of models—alter consumer evaluations irrespective of the cognitive message conveyed. To capture these sensory-driven, hedonic impressions, the authors engineered the AASD to measure affective response as an independent psychological entity. The scale answers critical theoretical and managerial questions, such as:

• How does the visual presentation of an advertisement influence immediate emotional and aesthetic valence?
• Does positive affective appreciation of an advertisement spill over into brand appraisal even when product claims are minimal or ambiguous?
• How do changes in artistic direction, color grading, casting, and music alter consumers’ visceral reactions to commercial messaging?

In contemporary research, the AASD is extensively deployed in laboratory and field experiments examining cross-cultural ad reactions, the impact of virtual influencers versus human models, interactive multimedia consumer engagement, and affective priming in digital interfaces. In applied industry settings, market researchers utilize the scale for pre-testing ad campaigns, evaluating copy variations, and identifying unintended negative or irritating consumer responses prior to nationwide media deployment.

5. Psychological Construct

The AASD operationalizes the affective dimension of Attitude toward the Advertisement ($A_{ad}$). The American Psychological Association (APA) conceptualizes an attitude as a relatively enduring organization of beliefs, feelings, and behavioral tendencies toward socially significant objects, groups, events, or symbols. Within advertising contexts, $A_{ad}$ represents a predisposed tendency to respond in a favorable or unfavorable manner to a particular advertising stimulus during a specific exposure occasion.

The AASD captures the affective domain across four closely related, mutually reinforcing facets:

Hedonic Valence

Hedonic valence reflects the fundamental pleasure-displeasure continuum. This sub-dimension captures whether the exposure experience generates feelings of subjective enjoyment, comfort, and positive warmth versus unpleasantness or distress. Anchored by pairs such as Unpleasant / Pleasant and Irritating / Enjoyable, this facet assesses the basic reward signals triggered in the central nervous system during sensory consumption of the advertisement.

Aesthetic Appreciation

Aesthetic evaluation targets the perceived beauty, elegance, artistic craftsmanship, and visual harmony of the execution. Operationalized via items such as Unattractive / Attractive and Unappealing / Appealing, this dimension captures whether the audience perceives the advertisement as an attractive visual artifact, which triggers positive aesthetic appraisals independent of the utility of the advertised brand.

Dynamic Arousal and Visual Engagement

Affective evaluations are intrinsically tied to activation and physiological arousal. The AASD assesses dynamic engagement through adjective pairs such as Dull / Dynamic and Boring / Interesting. This facet tracks whether the ad captures sensory energy, creates intrigue, and activates the orienting reflex, contrasting monotonous promotional formats with stimulating, emotionally charged creative executions.

Perceived Affective Sincerity

Emotional resonance depends heavily on authenticity. Consumers rapidly form affective judgments regarding the truthfulness of the emotional tone depicted on screen. Evaluated through pairs such as Phony / Genuine and Dishonest / Honest, this facet determines whether the affective appeal evokes feelings of genuine human connection or triggers defensive skepticism and emotional dissonance.

6. Theoretical Framework

The conceptual architecture of the AASD rests upon three interconnected theoretical paradigms in psychometrics, social psychology, and consumer information processing:

1. The Semantic Differential Paradigm of Meaning

The methodological backbone of the AASD is Charles E. Osgood, George Suci, and Percy Tannenbaum’s (1957) Measurement of Meaning. Osgood and colleagues demonstrated that human affective and connotative judgments across cultures consistently map onto a three-dimensional semantic space: Evaluation (good-bad, positive-negative), Potency (strong-weak, tough-fragile), and Activity (active-passive, fast-slow). The AASD strategically emphasizes the Evaluation dimension, calibrating bipolar adjective anchors specifically to tap emotional and aesthetic connotations elicited by commercial visual art.

2. The Dual-Mediation Hypothesis (DMH)

Formulated by MacKenzie, Lutz, and Belch (1986), the Dual-Mediation Hypothesis posits that affective evaluations of an advertisement ($A_{ad}$) exert both a direct influence on brand attitudes ($A_b$) via affective transfer (classical conditioning) and an indirect influence through cognitive brand attribute evaluations ($C_b$). When consumers form positive affective responses to an ad, they become more receptive to brand claims, lowering counterargumentation. The AASD serves as the canonical operationalization for the affective $A_{ad}$ locus within structural equation models testing the DMH.

3. The Elaboration Likelihood Model (ELM)

According to Petty and Cacioppo’s (1986) Elaboration Likelihood Model, persuasion occurs through two distinct pathways: the central route (deliberate, message-based cognitive processing) and the peripheral route (reliance on heuristic and affective cues). Under low-involvement processing conditions—typical of casual media consumption—visual attractiveness, musical scoring, and emotional tone serve as peripheral cues that determine persuasion. The AASD specifically measures these peripheral affective appraisals, proving critical for predicting attitude formation when consumers are neither motivated nor able to engage in deep cognitive elaboration.

7. Validity

The psychometric validity of the AASD has been rigorously confirmed across dozens of empirical investigations spanning multiple decades:

Construct and Factorial Validity

In Baker and Churchill’s (1977) initial validation study involving a factorial design ($2 \times 2 \times 2$) with male and female undergraduate participants evaluating print advertisements for coffee and perfume, the 11 affective items loaded heavily onto a single, dominant affective factor accounting for the vast majority of common variance. The scale exhibited distinct divergence from cognitive items (e.g., informative vs. uninformative, useful vs. useless) and conative measures (purchase intent).

Convergent Validity

Convergent validity is evidenced by high correlations between the AASD and alternative measures of emotional and evaluative response. Studies comparing the AASD with Mehrabian and Russell’s Pleasure-Arousal-Dominance (PAD) framework demonstrate strong positive correlations ($r = .65$ to $.78, p < .001$) between the AASD composite score and the PAD Pleasure dimension. Furthermore, the scale converges strongly with post-exposure facial electromyography (EMG) measurements, particularly zygomaticus major (smile muscle) activation during ad exposure.

Discriminant Validity

Discriminant validity has been consistently established using the Fornell-Larcker criterion and Multitrait-Multimethod (MTMM) matrices. The average variance extracted (AVE) of the AASD regularly exceeds $.60$, surpassing its shared variance with cognitive ad evaluation ($AVE > r^2_{Affective-Cognitive}$), prior brand familiarity, and general mood state. This demonstrates that the AASD captures ad-evoked affective valence rather than pre-existing baseline mood or pure informational utility.

Predictive and Criterion-Related Validity

The predictive validity of the scale has been demonstrated across hundreds of consumer behavior experiments. In structural equation modeling studies, affective scores derived from the AASD account for between $30%$ and $55%$ of the variance in overall attitude toward the brand ($A_b$) and indirectly predict up to $25%$ of the variance in behavioral purchase intentions ($PI$). Longitudinal ad-tracking studies show that ads scoring in the upper quartile of the AASD during pre-testing achieve significantly higher brand recall and lower commercial avoidance during live broadcasting.

8. Reliability

The AASD exhibits exemplary reliability across diverse sample populations, product categories, and cultural adaptations:

Internal Consistency

Baker and Churchill (1977) reported an overall internal consistency reliability coefficient (Cronbach’s alpha) of $\alpha = .89$ to $.92$ across their experimental conditions. Subsequent studies utilizing the full 11-item battery have replicated these findings:

• Print advertising replications: $\alpha = .90 – .94$
• Television and audiovisual commercial evaluations: $\alpha = .88 – .93$
• Digital, social media, and banner advertising contexts: $\alpha = .87 – .91$

Item-to-total correlations for each of the 11 pairs uniformly exceed $.55$, with central items like Unpleasant / Pleasant, Bad / Good, and Unappealing / Appealing frequently displaying corrected item-total correlations between $.70$ and $.84$.

Split-Half and Test-Retest Reliability

Guttman split-half coefficients routinely exceed $.88$. While test-retest reliability can be influenced by repeated exposure wear-out effects, short-interval test-retest stability (assessed with a 48-hour delay between exposures to identical static print advertisements) has yielded stability coefficients ranging from $r = .78$ to $r = .85$ ($p < .001$), demonstrating that the scale captures stable evaluative impressions rather than fleeting momentary measurement noise.

9. Factor Analysis

The underlying dimensionality of the AASD has been extensively scrutinized using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA):

Exploratory Factor Structure

In initial principal components analyses with varimax and oblimin rotations, the 11 bipolar pairs consistently produce a single primary eigenvalue exceeding $5.5$, accounting for $52%$ to $65%$ of the total variance. In some commercial contexts featuring highly complex audiovisual stimuli, a secondary minor factor emerges (eigenvalue $\approx 1.1$), separating pure Hedonic/Aesthetic Valence (e.g., Pleasant, Attractive, Appealing) from Dynamic Arousal/Authenticity (e.g., Dynamic, Genuine, Honest). However, due to high inter-factor correlations ($r > .60$), psychometricians generally advocate treating the scale as a unidimensional composite.

Confirmatory Factor Analysis (CFA) and Fit Indices

Modern CFA investigations evaluating the unidimensional 11-item model demonstrate strong global fit indices across large consumer samples ($N > 400$):

• Comparative Fit Index (CFI): $.95 – .98$
• Tucker-Lewis Index (TLI): $.94 – .97$
• Root Mean Square Error of Approximation (RMSEA): $.045 – .062$ (with $90%$ confidence intervals firmly below $.08$)
• Standardized Root Mean Square Residual (SRMR): $.031 – .042$

Standardized factor loadings ($lambda$) are uniformly high and statistically significant ($p < .001$):

• Unappealing / Appealing: $lambda = .84 – .89$
• Unpleasant / Pleasant: $lambda = .82 – .88$
• Bad / Good: $lambda = .80 – .85$
• Unattractive / Attractive: $lambda = .78 – .84$
• Irritating / Enjoyable: $lambda = .76 – .82$
• Boring / Interesting: $lambda = .73 – .80$
• Dull / Dynamic: $lambda = .68 – .77$
• Depressing / Cheerful: $lambda = .67 – .75$
• Phony / Genuine: $lambda = .64 – .73$
• Dishonest / Honest: $lambda = .62 – .71$
• Low quality / High quality: $lambda = .70 – .78$

10. Instrument / Measurement Tool

• Instrument Type: Self-administered psychometric rating scale.
• Format: Bipolar semantic differential scale with 7 response intervals between paired antonyms.
• Item Count: 11 bipolar adjective pairs.
• Target Population: Consumers, adolescents, and adults exposed to commercial advertising communications.
• Administration Time: Approximately 2 to 3 minutes.
• Scoring Protocol: Each bipolar pair is typically scored from 1 (most negative anchor) to 7 (most positive anchor), or alternatively from -3 to +3 (centered at 0). Negative anchor orientations are balanced during survey presentation to counteract acquiescence response bias and are reverse-scored before aggregation.
• Score Computation: Individual item scores are summed or averaged to generate a composite Affective Evaluation score (range: 11 to 77 for summed scoring, or 1.0 to 7.0 for mean scoring). Higher scores indicate more favorable affective, emotional, and aesthetic ad appraisals.

11. Permissions & Fee and Test Year

• Publication Year: 1977.
• Original Venue: Published in the Journal of Marketing Research (American Marketing Association).
• Intellectual Property & Licensing: The 11 bipolar semantic differential pairs are part of the open academic literature and may be utilized freely by researchers, students, and educators for non-commercial scientific research under standard scholarly fair use and citation conventions. Commercial market research platforms, copy-testing software suites, and corporate entities should ensure compliance with relevant journal publication copyright permissions held by the American Marketing Association (AMA).
• Usage Fee: Free for academic and scholarly empirical research.

12. References

• Baker, M. J., & Churchill, G. A., Jr. (1977). The impact of physically attractive models on advertising evaluations. Journal of Marketing Research, 14(4), 538–555. https://doi.org/10.1177/002224377701400411
• Batra, R., & Ray, M. L. (1986). Affective responses mediating acceptance of advertising. Journal of Consumer Research, 13(2), 234–249. https://doi.org/10.1086/209063
• Churchill, G. A., Jr. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(1), 64–73. https://doi.org/10.1177/002224377901600110
• Edell, J. A., & Burke, M. C. (1987). The power of feelings in understanding advertising effects. Journal of Consumer Research, 14(3), 421–433. https://doi.org/10.1086/209124
• Holbrook, M. B., & Batra, R. (1987). Assessing the role of emotions as mediators of consumer responses to advertising. Journal of Consumer Research, 14(3), 404–420. https://doi.org/10.1086/209123
• MacKenzie, S. B., Lutz, R. J., & Belch, G. E. (1986). The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of Marketing Research, 23(2), 130–143. https://doi.org/10.1177/002224378602300205
• Mitchell, A. A., & Olson, J. C. (1981). Are product attribute beliefs the only mediator of advertising effects on brand attitude? Journal of Marketing Research, 18(3), 318–332. https://doi.org/10.1177/002224378101800306
• Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The measurement of meaning. University of Illinois Press.
• 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