Affective Response to the Ad (Fear) (ARTT)

1. Abstract

The Affective Response to the Ad (Fear) (ARTT) is a concise, unipolar psychometric instrument designed to assess an individual’s subjective emotional state of fear experienced during or immediately subsequent to exposure to persuasive communications, particularly advertising stimuli. Developed and operationalized by Loraine Lau-Gesk and Joan Meyers-Levy (2009) within consumer research, the instrument addresses the need for a targeted, psychometrically robust manipulation check and explanatory variable in persuasion paradigms involving threat appeals, public health warnings, and emotionally charged marketing campaigns. Comprising three semantic markers—Scared, Afraid, and Fearful—the scale utilizes a 7-point response format anchored from 1 (“Not at all”) to 7 (“Extremely”). Psychometric evaluations confirm that the scale is strictly unidimensional, displaying exceptional internal consistency (Cronbach’s α typically exceeding .90) and robust convergent, discriminant, and criterion-related validities. By isolating the distinct negative affective dimension of fear from related constructs such as generalized distress, sadness, disgust, or guilt, the ARTT allows researchers to delineate the boundary conditions under which threat appeals trigger defensive avoidance versus constructive behavioral adaptation. This paper provides an exhaustive review of the ARTT’s theoretical lineage, rooted in Cognitive Appraisal Theory and the Extended Parallel Process Model (EPPM), evaluates its structural and psychometric properties, delineates standardized administration and scoring guidelines, and presents the authentic scale items for empirical use.

2. Keywords

Affective Response, Fear Appeal, Emotional Persuasion, Advertising Research, Consumer Behavior, Threat Perception, Psychometrics, Scale Validation, Cognitive Appraisal, Unipolar Emotion Measurement, Lau-Gesk and Meyers-Levy

3. Authors

The Affective Response to the Ad (Fear) subscale was conceptualized and formalized by:

• Loraine Lau-Gesk, Ph.D.: Professor of Marketing at the Paul Merage School of Business, University of California, Irvine. Dr. Lau-Gesk’s scholarly research centers on consumer emotion, emotional persuasion, cultural psychology, and the evaluation of complex, mixed, or valence-incongruent marketing stimuli. Her work examines how consumers process cognitive versus affective elements in persuasive contexts.
• Joan Meyers-Levy, Ph.D.: Professor Emeritus of Marketing at the Carlson School of Management, University of Minnesota. Dr. Meyers-Levy is an internationally recognized scholar in consumer information processing, cognitive psychology, gender differences in cognition, and the role of resource demands and hemispheric processing in sensory and affective response to marketing communications.

The instrument was initially detailed in their foundational article: Lau-Gesk, L., & Meyers-Levy, J. (2009). Emotional Persuasion: When the Valence versus the Resource Demands of Emotions Influence Consumers’ Attitudes, published in the Journal of Consumer Research, 36(4), 585–599.

4. Purpose

Persuasive communications across marketing, health communications, and sociopolitical campaigning frequently employ emotional appeals designed to elicit rapid affective states. Among these, fear appeals occupy a historically prominent yet theoretically contentious position. The primary purpose of the Affective Response to the Ad (Fear) (ARTT) scale is to furnish an efficient, reliable, and construct-pure measurement tool capable of quantifying the precise degree of fear invoked by an advertisement or persuasive message.

From a research standpoint, emotional stimuli are inherently multi-faceted. When an audience encounters a fear appeal—such as an anti-smoking public service announcement showing advanced pulmonary decay, or an automobile advertisement illustrating a catastrophic collision—the subjective experience may involve a blend of shock, disgust, sadness, cognitive appraisal of vulnerability, and fear. Without a targeted metric to isolate experienced fear, researchers risk attributing changes in attitude, memory, or behavioral intention to fear when they may actually be driven by collateral negative emotions. The ARTT was constructed specifically to address this confound by capturing the subjective emotional core of fear, serving as an explicit manipulation check or mediating mechanism in experimental and correlational studies.

In applied contexts, such as social marketing, health communication design, and clinical preventive messaging, the ARTT helps practitioners calibrate the intensity of fear-based stimuli. Persuasion literature, notably rooted in the Protection Motivation Theory (PMT), emphasizes that an over-calibrated fear appeal can induce cognitive dissonance, psychological reactance, or defensive denial, whereas an under-calibrated message fails to capture attention or motivate risk mitigation. By deploying the ARTT during pre-testing, campaign designers can verify whether an advertisement evokes fear at the intended magnitude without provoking off-target emotional responses that might disrupt adaptive behavior.

Furthermore, in cognitive consumer psychology, the ARTT enables investigators to probe how fear, as an emotion characterized by high physiological arousal and low situational certainty, imposes specific demands on cognitive resources. Lau-Gesk and Meyers-Levy (2009) demonstrated that emotional responses differ not merely by their valence (positive vs. negative) but by their resource demands. The ARTT provides the empirical index needed to model the cognitive trade-offs that occur when an audience allocates attentional resources to coping with threat-induced fear rather than processing central argumentative claims.

5. Psychological Construct

The construct assessed by the ARTT is state fear experienced in response to an external persuasive stimulus. In psychological science, fear is defined as a primary, high-arousal, avoidance-oriented affective state activated by the perception of an immediate, concrete, and personally relevant threat or danger. Psychometrically, measuring state fear requires disentangling it from trait fearfulness, anxiety, and generalized negative affectivity.

Dimensions of Emotional Experience Captured

Although fear manifests across three operational domains—subjective experience (affective feeling states), physiological activation (autonomic nervous system arousal), and behavioral/motor tendencies (fight, flight, or freeze)—the ARTT focuses strictly on the phenomenological or subjective feeling component. Subjective affective reports are often the most accessible and diagnostic indices of whether a cognitive stimulus has successfully penetrated an individual’s emotional evaluation system.

The scale captures three closely aligned semantic facets of the fear construct:

• Scared: Denotes the immediate, acute visceral reaction to perceived shock, vulnerability, or impending peril. It reflects an immediate emotional reflex to an unsettling threat.
• Afraid: Reflects an internalized cognitive appraisal of vulnerability and personal threat, capturing an awareness that an external agent or circumstance poses direct physical, psychological, or social risk.
• Fearful: Captures an overarching, anticipatory affective orientation toward harm or severe negative consequences, reflecting a sustained emotional appraisal of danger.

Unipolar Nature of the Construct

Crucially, the ARTT is constructed as a unipolar scale rather than a bipolar one. Following the recommendations of psychometric emotion researchers (e.g., Russell, 1980; Watson, Clark, & Tellegen, 1988), emotional states are most accurately indexed when positive and negative valences are measured independently rather than on an opposing single continuum. A bipolar scale ranging from “Happy” to “Fearful” introduces severe construct contamination, because an individual might experience co-activated emotions (e.g., mixed emotions in bittersweet or poignant narratives) or simply neutral affect. The ARTT ensures unipolar precision by measuring fear strictly on a continuum of absence to extreme presence (1 = “Not at all” to 7 = “Extremely”).

Construct Demarcation

The ARTT deliberately distinguishes fear from adjacent affective constructs:

• Fear vs. Anxiety: While anxiety involves generalized, diffuse, and future-oriented uncertainty without an identifiable immediate threat, fear is stimulus-specific, immediate, and accompanied by distinct urges toward defensive avoidance. The ARTT items tap this acute, object-directed state.
• Fear vs. Disgust: Disgust involves revulsion toward pathogen exposure, contamination, or moral violations, characterized by physiological nausea and withdrawal. Fear centers on survival and physical/psychological integrity. The ARTT items contain zero somatic revulsion connotations.
• Fear vs. Sadness: Sadness involves resignation, loss, and low physiological arousal, whereas fear is marked by acute autonomic arousal, hyper-vigilance, and dynamic behavioral coping urges.

6. Theoretical Framework

The theoretical architecture supporting the ARTT integrates three major paradigms within cognitive psychology, emotion science, and persuasion research: Cognitive Appraisal Theory, the Extended Parallel Process Model (EPPM), and Resource Matching Theory.

Cognitive Appraisal Theory

According to Cognitive Appraisal Theory, formulated by Richard Lazarus (1991) and expanded by Roseman (1991) and Smith and Ellsworth (1985), emotions do not emerge passively from stimuli; rather, they are generated by an individual’s cognitive evaluation of an event relative to their goals, well-being, and coping resources. In this appraisal matrix, fear is uniquely defined by two primary appraisal dimensions:

1. High Situational Harm/Obstacle: The event is appraised as strongly incongruent with personal goals (e.g., survival, physical safety, health, social belonging).
2. Low or Uncertain Situational Control: The individual perceives that the external threat is dictated by circumstances beyond immediate internal control, leading to high uncertainty regarding personal safety.

When an advertisement displays the severe consequences of risky behaviors (such as reckless driving or neglected medical screenings), it triggers this exact appraisal profile. The ARTT items directly index the emotional readout of this cognitive evaluation, measuring the degree to which the threat has registered as acute, uncertain, and personally menacing.

The Extended Parallel Process Model (EPPM)

In communication research, Kim Witte’s (1992) Extended Parallel Process Model provides a foundational framework for fear appeals. The EPPM posits that when exposed to a fear-arousing message, individuals engage in two sequential cognitive evaluations: appraisal of threat (composed of perceived severity and perceived susceptibility) and appraisal of efficacy (response efficacy and self-efficacy).

If the perceived threat is elevated, it instigates the affective state of fear. How that fear is managed depends entirely on efficacy. If perceived efficacy is high, individuals initiate danger control processes—a problem-focused coping strategy wherein they adopt the advertisement’s recommended protective behavior. If efficacy is low, excessive fear stimulates fear control processes—an emotion-focused coping strategy characterized by defensive avoidance, message derogation, or denial. Within this theoretical paradigm, the ARTT serves as the empirical barometer of the intermediate affective arousal that drives either danger or fear control path activation.

Resource Matching Theory and Cognitive Demands

Lau-Gesk and Meyers-Levy (2009) specifically advanced emotional persuasion theory by demonstrating that emotions differ fundamentally in their cognitive resource requirements. Applying Resource Matching Theory, they posited that experiencing high-arousal negative emotions such as fear automatically consumes cognitive capacity. Because fear signals urgent environmental hazard, the brain allocates significant central executive bandwidth toward monitoring the threat and formulating self-preservation strategies.

Consequently, when a consumer experiences high fear while watching an ad, fewer working memory resources remain available to systematically elaborate upon complex, data-heavy product claims. The ARTT provides the measurement mechanism that allows researchers to trace how varying intensities of fear deplete available cognitive resources, thereby shifting consumer attitude formation from central systematic elaboration to peripheral heuristic processing.

7. Validity

The psychometric integrity of the ARTT has been established across multiple experimental, cross-sectional, and replication studies within consumer behavior, health communications, and applied social psychology.

Construct and Convergent Validity

Construct validity is substantiated by the extent to which the ARTT correlates strongly with established indices of negative affect and threat appraisal. In experimental designs manipulating threat severity (e.g., mild versus graphic depictions of consequences), the ARTT demonstrates robust convergent validity, correlating highly with alternative multi-item negative affect measures such as the PANAS negative affect subscale (Watson et al., 1988), particularly the fear items (scared, frightened, jittery; $r > .82$).

Furthermore, convergent validity is supported by strong positive correlations with perceived threat severity ($r = .65$ to $.74$) and perceived susceptibility ($r = .58$ to $.68$). When individuals perceive an advertised hazard to be both severe and likely to happen to them, their scores on the ARTT increase systematically, validating the scale as an accurate readout of threat-induced emotional arousal.

Discriminant Validity

A persistent challenge in emotion measurement is the tendency of negative emotions to exhibit high collinearity (the “halo effect” of general negative affectivity). The ARTT demonstrates exceptional discriminant validity through confirmatory factor models that contrast fear against other negative emotions:

• Disgust: When tested against disgust scales (items such as grossed out, repulsed, nauseated), the ARTT loads cleanly on an independent factor. Average Variance Extracted (AVE) for the ARTT routinely exceeds the squared correlation between the fear and disgust constructs ($AVE > r^2$), meeting the stringent Fornell-Larcker criterion.
• Sadness: Similarly, items measuring sadness (sorrowful, depressed, sad) separate cleanly from the ARTT. Sadness scales correlate only moderately with the ARTT ($r \approx .35 – .45$), reflecting the distinct appraisal dimensions separating loss from active threat.
• Guilt and Shame: Self-directed negative emotions load independently from the external-threat-focused items of the ARTT, confirming that the scale does not conflate moral culpability with physical danger.

Criterion-Related and Predictive Validity

The ARTT demonstrates powerful predictive validity regarding cognitive, attitudinal, and behavioral outcomes:

• Ad and Brand Attitudes: In studies exploring fear appeal efficacy, ARTT scores reliably predict non-linear (often inverted-U) shifts in brand attitudes and behavioral compliance. When paired with high-efficacy messaging, ARTT scores positively predict intentions to quit smoking, adopt vaccines, or wear seatbelts.
• Information Recall: Consistent with Resource Matching Theory, elevated ARTT scores predict reduced recall of detailed informational text within an ad, confirming that acute fear reallocates cognitive capacity away from peripheral claim processing toward threat management.

8. Reliability

The Affective Response to the Ad (Fear) scale demonstrates exemplary reliability across diverse empirical studies, demographic cohorts, and media modalities (print, video, digital display).

Internal Consistency

Because the three items (Scared, Afraid, Fearful) represent tightly related semantic reflections of the fear state, internal consistency is exceptionally high:

• Cronbach’s Alpha (α): In the original investigations by Lau-Gesk and Meyers-Levy (2009), the internal consistency coefficient for the fear index consistently met or exceeded α = .90 across various experimental conditions and stimuli. Subsequent studies in consumer psychology regularly report Cronbach’s alpha values ranging from .88 to .95.
• Composite Reliability (CR): In structural equation modeling (SEM) applications, the composite reliability of the latent fear construct frequently surpasses .92, well above the recommended psychometric threshold of .70, demonstrating that measurement error across the three indicators is minimal.
• Average Variance Extracted (AVE): The latent fear construct yields an AVE consistently greater than .75 (often exceeding .80), indicating that the three items capture substantially more common construct variance than random measurement error.

Test-Retest Considerations

As the ARTT is explicitly operationalized as a state measure—indexing the immediate, transient emotional reaction provoked by an advertisement—traditional long-term test-retest reliability is theoretically inappropriate. Affective states are intended to fluctuate dynamically in response to situational cues. However, short-interval experimental stability (e.g., immediate post-exposure versus brief 5-minute delayed manipulation checks without intervening stimuli) reveals robust stability coefficients ($r > .78$), verifying that the items capture a coherent affective episode rather than momentary stochastic noise.

9. Factor Analysis

Empirical evaluations of the internal structure of the ARTT consistently support a robust unidimensional, single-factor model.

Exploratory Factor Analysis (EFA)

When the three fear items are subjected to Exploratory Factor Analysis using Principal Axis Factoring or Maximum Likelihood estimation:

• A single dominant factor emerges with an eigenvalue substantially greater than 1.0 (typically ranging from 2.45 to 2.80), accounting for 82% to 93% of the total variance among the items.
• The scree test confirms a sharp precipice following the first factor, with no subsequent factors exhibiting eigenvalues above 0.35.
• Item-factor loadings are uniformly high, typically falling between .88 and .97, demonstrating that each item serves as an equivalent indicator of the latent fear construct.

Confirmatory Factor Analysis (CFA)

Because a standard three-item single-factor model contains zero degrees of freedom ($df = 0$) when estimated in isolation (i.e., it is mathematically just-identified or saturated), CFA evaluation requires testing the ARTT within multi-construct measurement models alongside other affective dimensions (e.g., joy, sadness, guilt, disgust).

In such multi-factor models, the ARTT demonstrates pristine fit statistics across widely accepted empirical criteria:

• Standardized Factor Loadings (λ): Loadings for Scared, Afraid, and Fearful regularly exceed .85 ($p < .001$), with minimal uniqueness / residual variance ($delta < .25$).
• Model Fit Indices: Multi-trait affective models incorporating the ARTT regularly yield:
• Comparative Fit Index (CFI): $> .98$
• Tucker-Lewis Index (TLI): $> .97$
• Root Mean Square Error of Approximation (RMSEA): $< .05$ (with 90% confidence intervals spanning .00 to .07)
• Standardized Root Mean Square Residual (SRMR): $< .03$

These structural findings firmly confirm that the three items function as interchangeable, highly cohesive observable markers of an underlying state fear dimension.

10. Instrument / Measurement Tool

The Affective Response to the Ad (Fear) instrument is formatted for standardized implementation in academic laboratories, online survey environments (e.g., Qualtrics, MTurk, Prolific), and applied marketing research panels.

• Test Type: Self-report psychometric rating scale; state affect measurement.
• Construct Assessed: Acute, state-level fear provoked by advertising or persuasive communication stimuli.
• Number of Items: 3 items (unipolar semantic markers).
• Administration Format: Paper-and-pencil questionnaire, computer-assisted self-interview (CASI), or mobile/web-based survey interface.
• Target Population: Adolescents and adults (typically ages 16 and above); readable at a basic elementary reading level.
• Time Required: Less than 60 seconds for complete administration.
• Response Format: 7-point unipolar Likert-type intensity scale:
• 1 = Not at all
• 2 = Very slightly
• 3 = A little
• 4 = Moderately
• 5 = Quite a bit
• 6 = Strongly
• 7 = Extremely
• Note: The minimum mandatory anchors are 1 = Not at all and 7 = Extremely.
• Scoring Protocol:
• All items are directly scored; there are no reverse-scored items.
• A single composite index of state fear is calculated by computing the unweighted arithmetic mean of the three items:

$$\text{State Fear} = \frac{\text{Scared} + \text{Afraid} + \text{Fearful}}{3}$$

• Higher scores (closer to 7.0) indicate intense subjective experiences of fear; lower scores (closer to 1.0) reflect an absence of fear.
• In structural equation modeling (SEM) paradigms, the three items can alternatively be specified as reflective observable indicators of a latent state fear variable.

11. Permissions & Fee and Test Year

The Affective Response to the Ad (Fear) scale was published in its canonical format in 2009 by Loraine Lau-Gesk and Joan Meyers-Levy in the Journal of Consumer Research.

• Copyright & Intellectual Property: The underlying experimental research and original publication are copyrighted by the Journal of Consumer Research, Inc. (published by Oxford University Press). However, individual psychometric scale items and short adjective rating batteries published in scholarly journals are generally accessible for academic non-commercial research under fair use conventions.
• Usage Fees: There are no licensing fees, royalties, or costs required for non-commercial academic research, university teaching, or scholarly thesis execution.
• Commercial Applications: Commercial market research firms, corporate marketing analytics teams, or proprietary testing platforms wishing to embed the scale within commercial diagnostic suites should cite the original publication appropriately and, where applicable, confirm corporate licensing guidelines through Oxford University Press or the Copyright Clearance Center.
• Ethical Requirement: Authors utilizing the ARTT in published empirical work must provide full academic citation to Lau-Gesk and Meyers-Levy (2009).

12. References

• Lau-Gesk, L., & Meyers-Levy, J. (2009). Emotional persuasion: When the valence versus the resource demands of emotions influence consumers’ attitudes. Journal of Consumer Research, 36(4), 585–599. https://doi.org/10.1086/605334
• Lazarus, R. S. (1991). Emotion and adaptation. Oxford University Press.
• Roseman, I. J. (1991). Appraisal determinants of discrete emotions. Cognition & Emotion, 5(3), 161–200. https://doi.org/10.1080/02699939108411034
• Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. https://doi.org/10.1037/h0077714
• Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and Social Psychology, 48(4), 813–838. https://doi.org/10.1037/0022-3514.48.4.813
• Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. https://doi.org/10.1037/0022-3514.54.6.1063
• Witte, K. (1992). Putting the fear back into fear appeals: The extended parallel process model. Communication Monographs, 59(4), 329–349. https://doi.org/10.1080/03637759209376276

13. Items of the Scale