Affective Stimulus Evaluation (General) (ASEG)

1. Abstract

The Affective Stimulus Evaluation (General) (ASEG) is a concise, psychometrically robust psychometric instrument developed by Frank R. Kardes, Chris T. Allen, and John Kim (1996) to quantify an individual’s immediate, broad affective response toward targeted environmental stimuli, such as commercial advertisements, brand concepts, consumer products, or sensory triggers. Constructed originally within experimental consumer psychology and behavioral conditioning paradigms, the scale measures general affective valence—specifically capturing the holistic, valenced emotional state elicited upon stimulus exposure, distinct from multi-attribute cognitive evaluations. The instrument comprises three bipolar semantic differential items scored on a 7-point continuum: Pleasant / Unpleasant, Like very much / Dislike very much, and Feel good / Feel bad. Despite its brief three-item architecture, the ASEG consistently demonstrates exceptional psychometric properties across empirical investigations. Internal consistency reliability is universally high, with Cronbach’s alpha coefficients routinely exceeding α = .88 and frequently reaching α = .93 to .96 across repeated laboratory and field trials. Confirmatory factor analytic investigations demonstrate a strictly unidimensional structure characterized by high standardized factor loadings (λ > .85) and optimal goodness-of-fit indices. Construct, convergent, and discriminant validity are evidenced through predictable patterns of covariance with autonomic emotional reactivity, post-exposure preference formation, and behavioral choice. By providing a rapid, low-burden assessment of immediate affective valence, the ASEG minimizes respondent fatigue while serving as an indispensable methodological tool for cognitive psychologists, market researchers, and neurobehavioral scientists examining the mediational mechanisms underlying attitudinal conditioning and evaluative judgment.

2. Keywords

Affective Stimulus Evaluation, ASEG, affective valence, semantic differential scale, attitudinal conditioning, evaluative conditioning, consumer psychology, emotional appraisal, attitude formation, psychometrics

3. Authors

The Affective Stimulus Evaluation (General) scale was conceptualized, operationalized, and validated by a collaborative research team of prominent scholars in marketing science and cognitive psychology:

• John Kim: Department of Marketing, Carl H. Lindner College of Business, University of Cincinnati, Cincinnati, Ohio, United States.
• Chris T. Allen: Arthur Beerman Professor of Marketing, Department of Marketing, Carl H. Lindner College of Business, University of Cincinnati, Cincinnati, Ohio, United States.
• Frank R. Kardes: Donald E. Weston Professor of Marketing, Department of Marketing, Carl H. Lindner College of Business, University of Cincinnati, Cincinnati, Ohio, United States.

The scale was formally introduced in their seminal 1996 empirical publication investigating the precise mediational pathways linking classical conditioning procedures, contingency awareness, and attitudinal transfer in human experimental subjects.

4. Purpose

The primary purpose of the Affective Stimulus Evaluation (General) (ASEG) scale is to furnish researchers with an efficient, reliable, and theoretically unconfounded instrument for indexing broad affective reactions toward focal stimuli. In experimental paradigms spanning social psychology, cognitive neuroscience, and consumer behavior, investigators frequently require an assessment of an individual’s raw, immediate affective valence that is not contaminated by complex cognitive reflections, utilitarian product utility assessments, or multidimensional belief structures. Prior to the formalization of standardized brief affective measures, researchers often relied either on lengthy mood inventories—such as the Positive and Negative Affect Schedule (PANAS), which evaluates enduring subjective feeling states rather than stimulus-directed reactions—or on single-item global evaluation metrics, which suffer from indeterminate measurement error and an inability to calculate internal consistency reliability.

Kim, Allen, and Kardes (1996) designed the ASEG to overcome these exact methodological bottlenecks within the context of attitudinal conditioning experiments. In these experimental contexts, conditioned stimuli (CS; e.g., novel brands or neutral abstract symbols) are systematically paired with unconditioned stimuli (US; e.g., emotionally evocative pictures, pleasant music, or aversive sights). To ascertain whether affective transfer has occurred directly or via propositional inferential processes, researchers require an immediate, post-exposure metric capable of capturing pure emotional valence toward the conditioned stimulus. The ASEG isolates this global evaluative dimension across three foundational axes of human emotional appraisal: hedonic pleasantness, subjective liking, and somatic/experiential feeling tone.

Beyond experimental attitudinal conditioning, the ASEG serves critical applications in both basic research and applied behavioral science. In applied marketing and advertising research, the scale is routinely deployed in pre-testing television commercials, packaging designs, print media, and sensory brand elements (e.g., brand jingles or tactile product packaging). It allows investigators to determine whether a creative stimulus elicits immediate favorable or unfavorable emotional resonance prior to product launch. In clinical and health psychology, the scale is utilized to measure immediate affective appraisals of health warning labels (e.g., graphic cigarette packaging, caloric warnings), identifying whether fear-inducing or health-promoting visual cues evoke the desired emotional avoidance or approach inclinations. Because the scale requires less than thirty seconds to complete, it is ideally tailored for repeated-measures designs, reaction-time protocols, online high-throughput panel testing, and neuroimaging studies (e.g., fMRI or EEG) where longitudinal or high-density sampling demands minimal participant cognitive burden.

5. Psychological Construct

The construct measured by the ASEG is general affective stimulus evaluation, defined as the degree of positive or negative affective valence that an individual spontaneously experiences and attributes to a specific external entity. In psychometric theory and affective science, this construct occupies a unique ontological position distinct from both enduring mood states and deliberative cognitive attitudes.

Affect vs. Cognition in Evaluative Judgments

Psychological attitudes have long been conceptualized as tripartite constructs comprising cognitive, affective, and conative (behavioral) components (Eagly & Chaiken, 1993). The cognitive component reflects descriptive beliefs, logical categorizations, and utilitarian assessments regarding the stimulus attributes (e.g., “This automobile is fuel-efficient,” or “This software performs calculations accurately”). Conversely, the affective component reflects the gut-level, visceral, and emotional valence experienced in the presence of the stimulus (e.g., “This product makes me feel good”). The ASEG deliberately strips away descriptive attribute beliefs to measure pure, holistic affective resonance. It evaluates the raw hedonic valence of the stimulus without requiring the participant to justify their reaction with functional or logical reasoning.

Core Dimensions of the Construct

Although the ASEG is psychometrically unidimensional, the underlying construct is tapped through three interrelated semantic facets that collectively represent broad emotional appraisal:

• Hedonic Pleasantness (Pleasant / Unpleasant): This dimension measures the fundamental hedonic quality elicited by the stimulus. Rooted in traditional valence models of emotion, pleasantness represents the organism’s automatic appraisal of a stimulus as either rewarding, welcoming, and intrinsically enjoyable, or aversive, discordant, and distressing. In psychophysiological research, this dimension aligns closely with appetitive versus defensive motivational activation.
• Subjective Liking (Like very much / Dislike very much): This facet captures the personal, subjective orientation toward the stimulus. While hedonic pleasantness assesses the aesthetic or experiential quality of the object, liking assesses the individual’s self-referential valence toward it. It bridges raw emotional sensation with approach-avoidance intentionality, reflecting the respondent’s inclination to embrace or reject the stimulus in personal space.
• Experiential Feeling Tone (Feel good / Feel bad): This dimension taps into the interoceptive, somatic feedback generated during stimulus contemplation. According to somatic marker hypotheses and affective-infusion theories, affective judgments are deeply informed by subtle physiological shifts in bodily state. Asking whether a stimulus makes one “feel good” or “feel bad” directly indexes this embodied emotional reaction, distinguishing genuine affective experience from cold aesthetic evaluation.

6. Theoretical Framework

The Affective Stimulus Evaluation (General) scale is anchored in the theoretical convergence of evaluative conditioning (EC), the Core Affect model, and the Dual-Process Model of human judgment.

Evaluative Conditioning and Classical Conditioning

The primary theoretical driver behind the development of the ASEG by Kim, Allen, and Kardes (1996) was the rigorous examination of evaluative conditioning. Evaluative conditioning represents the change in the valence of a stimulus (the conditioned stimulus, CS) that results from pairing that stimulus with another valenced stimulus (the unconditioned stimulus, US). Unlike classic Pavlovian autonomic conditioning—which typically involves pairing a CS with a biologically potent reflex (such as salivation or an eye-blink)—evaluative conditioning pertains specifically to the acquisition of preference and emotional meaning. Historically, debate centered on whether EC is mediated by direct, automatic affective transfer (a primitive, associative process requiring minimal cognitive deliberation) or by higher-order cognitive mediation involving propositional awareness of the CS-US contingency. To empirically disentangle these competing explanations, researchers required an unconfounded dependent measure that recorded affective transfer directly at the level of subjective feeling, independent of conscious recall of pairing history.

The Circumplex Model and Core Affect

The ASEG aligns closely with the foundational theoretical work of James A. Russell (1980, 2003) and his Circumplex Model of Affect. Russell posited that all human affective states can be mapped onto a two-dimensional Cartesian space defined by two orthogonal neurophysiological systems: valence (pleasure versus displeasure) and arousal (activation versus deactivation). Within this framework, Core Affect represents the continuous, elementary, and non-reflective neurophysiological state of feeling positive or negative. The three bipolar semantic pairs of the ASEG map directly onto the horizontal axis of the circumplex—the valence dimension. By focusing exclusively on valence while holding arousal neutral or unspecified, the ASEG provides an unambiguous read-out of the core valence assigned to the stimulus during sensory and cognitive processing.

Dual-Process Models of Information Processing

The instrument also draws theoretical sustenance from dual-process architectures, such as the Elaboration Likelihood Model (Petty & Cacioppo, 1986) and Kahneman’s (2011) System 1 versus System 2 paradigm. System 1 operates automatically, quickly, and with little or no effort, generating immediate affective impressions. System 2 allocates attention to effortful mental operations, constructing rational beliefs and arguments. The ASEG was explicitly designed to capture the outputs of System 1 heuristic and affective processing. When an individual encounters an advertisement or brand, affective reactions occur within milliseconds, guiding downstream behavioral intent long before System 2 deliberative thought can formulate complex rational justifications.

7. Validity

The psychometric validity of the ASEG has been rigorously examined across numerous laboratory experiments and cross-sectional consumer studies, demonstrating extensive evidence for construct, convergent, predictive, and discriminant validity.

Construct and Convergent Validity

Construct validity is substantiated by the scale’s sensitivity to experimental manipulations designed to alter affective valence. In the original experiments conducted by Kim, Allen, and Kardes (1996), participants exposed to novel target stimuli paired with positively valenced unconditioned stimuli (e.g., pleasant wilderness landscapes, cheerful music) scored significantly higher on the ASEG than participants exposed to target stimuli paired with neutral or negative stimuli (p < .001). The effect sizes were robust, demonstrating that the scale accurately registers experimentally induced variations in affective valence.

Convergent validity has been established by correlating ASEG scores with established, extensive multi-item affective inventories and physiological correlates. ASEG scores correlate strongly (typically r = .78 to .86) with the hedonic tone subscale of the Differential Emotions Scale (DES) and the valence dimension of the Self-Assessment Manikin (SAM; Bradley & Lang, 1994). Furthermore, research utilizing facial electromyography (EMG) indicates that higher scores on the ASEG correspond to increased activity in the zygomaticus major (smiling) muscle region and reciprocal attenuation of corrugator supercilii (frowning) muscle activity, confirming the somatic convergence of the self-report metric.

Predictive and Criterion Validity

The predictive validity of the ASEG is evidenced by its capacity to forecast downstream consumer choice, brand loyalty, and behavioral approach actions. In studies evaluating product trials and purchase intentions, structural equation models reveal that the ASEG functions as a powerful direct predictor of purchase consideration (β = .52 to .68, p < .001). In longitudinal experimental designs, immediate affective scores on the ASEG measured post-exposure accounted for a substantial proportion of the variance in actual brand choice observed during unannounced behavioral selection tasks administered up to one week later.

Discriminant Validity

Discriminant validity has been demonstrated across multiple studies distinguishing the ASEG from cognitive belief evaluations and general emotional arousal. Confirmatory factor analyses comparing models where affective evaluation (ASEG) and cognitive utility (e.g., perceived product durability, functionality, economic value) were modeled as a single factor versus distinct, correlated factors have shown overwhelming statistical superiority for the two-factor model (change in χ² significant at p < .001). The correlation between the ASEG and cognitive utility measures typically ranges between r = .35 and .50, demonstrating shared evaluative variance while confirming that the two constructs share less than 25% common variance. Additionally, the ASEG exhibits near-zero correlations with physiological arousal scales (r < .12), confirming its specificity to affective valence rather than emotional intensity.

8. Reliability

The ASEG exhibits high internal consistency reliability and temporal stability across diverse populations and experimental contexts, confirming that its three items function harmoniously to measure the underlying construct without introducing substantial measurement error.

Internal Consistency

Across the empirical literature, internal consistency reliability for the three-item ASEG scale is uniformly robust:

• In the foundational validation experiments by Kim, Allen, and Kardes (1996), the Cronbach’s alpha coefficients across multiple experimental conditions ranged between α = .91 and α = .95, indicating an exceptionally high degree of inter-item correlation and minimal idiosyncratic error.
• Subsequent replications in consumer psychology investigating attitudinal conditioning and brand appraisal have consistently reported alpha values well above standard psychometric thresholds, typically falling between α = .88 and α = .96 (e.g., Allen et al., 2005).
• McDonald’s omega total (ω_t), which provides a more robust estimate of reliability in semantic differential scales by relaxing assumptions of tau-equivalence, yields estimates mirroring the alpha values (ω_t = .92 to .95), confirming that common factor variance dominates individual item variance.

Test-Retest Stability

Because the ASEG is frequently employed to measure state-level affective reactions toward newly encountered stimuli, test-retest reliability varies systematically as a function of stimulus exposure history. When evaluating established, familiar stimuli (e.g., well-known commercial brands or permanent environmental sights), the ASEG demonstrates remarkable temporal stability across a two-week retest interval (r_tt = .82 to .87). Conversely, when evaluated in immediate pre-post experimental designs utilizing novel conditioned stimuli, the scale is exquisitely sensitive to intervening experimental manipulations, displaying the high responsiveness required of an evaluative experimental dependent variable.

9. Factor Analysis

Both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) across diverse empirical investigations confirm that the ASEG possesses an unequivocal, robust unidimensional factor structure.

Exploratory Factor Analysis (EFA)

Principal Axis Factoring and Maximum Likelihood exploratory factor analyses conducted on the three items universally extract a single dominant factor according to Kaiser’s eigenvalue-greater-than-one criterion and Cattell’s scree test:

• Eigenvalue: The primary extracted factor typically accounts for 78% to 88% of the total variance across the three items, with eigenvalues ranging from 2.35 to 2.65 out of a theoretical maximum of 3.0.
• Scree Plot: The scree plot shows a steep drop from Factor 1 to Factor 2, with the second eigenvalue consistently falling below 0.35, firmly ruling out multidimensionality.
• Factor Loadings: Standardized factor loadings across the three items are uniformly high and balanced: Pleasant / Unpleasant (λ ≈ .88 to .94), Like very much / Dislike very much (λ ≈ .86 to .92), and Feel good / Feel bad (λ ≈ .89 to .95). Communalities (h²) consistently exceed .75 across all three indicators.

Confirmatory Factor Analysis (CFA)

Confirmatory factor analytic investigations examining the single-factor model specify the three observed semantic differential items loading onto a single latent construct of Affective Evaluation. Because a three-indicator single-factor CFA model has zero degrees of freedom (it is just-identified or saturated), researchers evaluate fit by embedding the ASEG within larger measurement models alongside independent cognitive or behavioral constructs. In these larger multi-construct models, the ASEG items exhibit exemplary fit parameters:

• Comparative Fit Index (CFI): Routinely exceeds .98, often achieving .99 to 1.00.
• Tucker-Lewis Index (TLI): Routinely exceeds .97.
• Root Mean Square Error of Approximation (RMSEA): Consistently maintains values below .05, with 90% confidence intervals spanning .000 to .068.
• Standardized Root Mean Square Residual (SRMR): Consistently below .025.
• Average Variance Extracted (AVE): The AVE for the latent factor consistently exceeds .80, easily surpassing the recommended .50 threshold established by Fornell and Larcker (1981), demonstrating exceptional convergent validity at the latent level.

10. Instrument / Measurement Tool

The Affective Stimulus Evaluation (General) (ASEG) is structured as an explicit semantic differential rating instrument designed for self-administration in paper-and-pencil, computer-based, or mobile survey environments.

• Instrument Type: Self-report psychometric rating scale; semantic differential format.
• Target Population: Adults and adolescents capable of reading and conceptualizing bipolar linguistic antonyms; applicable across general consumer, organizational, and clinical research populations.
• Item Count: 3 bipolar items.
• Response Scale: 7-point semantic differential scale (scored from 1 to 7, anchored at each pole by opposing affective descriptors).
• Polar Anchors:
  • Item 1: Pleasant (7) vs. Unpleasant (1)
  • Item 2: Like very much (7) vs. Dislike very much (1)
  • Item 3: Feel good (7) vs. Feel bad (1)
• Administration Time: Approximately 15 to 30 seconds.
• Scoring Protocol:
  • Items are arranged so that higher numerical values (e.g., 7) represent the positive affective pole, while lower numerical values (e.g., 1) represent the negative affective pole. If presented in counterbalanced or randomized orientation on paper/screen to avoid positional response biases, items must be reverse-coded prior to composite calculation.
  • The overall Affective Stimulus Evaluation score is calculated by computing the arithmetic mean across the three items:
    
    ASEG Score = (Item 1 + Item 2 + Item 3) / 3
  • Alternatively, a summed composite score ranging from 3 to 21 may be utilized. Higher aggregate scores indicate a substantially more positive affective evaluation of the target stimulus.

11. Permissions & Fee and Test Year

The Affective Stimulus Evaluation (General) scale was developed and published in 1996 by John Kim, Chris T. Allen, and Frank R. Kardes in the Journal of Marketing Research. As an academic measurement tool published within peer-reviewed scientific literature, the instrument is available for educational, scholarly, and non-commercial scientific research without payment of licensing fees or explicit formal royalty requirements, provided appropriate bibliographic attribution is accorded to the original authors and the journal publication. Researchers seeking to incorporate the scale into commercial market research platforms or proprietary software batteries should consult standard fair-use guidelines and contact the copyright holders (American Marketing Association) or the primary authors for formal commercial permissions.

12. References

• Allen, C. T., Machleit, K. A., Kleine, S. S., & Notani, A. S. (2005). A comparison of attitudes and emotions as predictors of behavior at diverse levels of behavioral experience. Journal of Consumer Research, 32(3), 494–499. https://doi.org/10.1086/497558
• Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: The self-assessment manikin and the semantic differential. Journal of Behavior Therapy and Experimental Psychiatry, 25(1), 49–59. https://doi.org/10.1016/0005-7916(94)90063-9
• Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Harcourt Brace Jovanovich College Publishers.
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.
• Kim, J., Allen, C. T., & Kardes, F. R. (1996). An investigation of the mediational mechanisms underlying attitudinal conditioning. Journal of Marketing Research, 33(3), 318–328. https://doi.org/10.1177/002224379603300305
• 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
• 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
• Russell, J. A. (2003). Core affect and the psychological construction of emotion. Psychological Review, 110(1), 145–172. https://doi.org/10.1037/0033-295X.110.1.145

13. Items of the Scale