PAD Emotional State Scale

1. Abstract

The PAD Emotional State Scale is a foundational psychometric instrument developed by Albert Mehrabian and James A. Russell (1974) to capture the multifaceted spectrum of human emotional responses to physical, social, and virtual environments. Constructed on a dimensional paradigm of affective experience, the scale posits that any emotional state can be parsimoniously yet comprehensively characterized through three orthogonal, continuous dimensions: Pleasure–Displeasure (measuring the valence or hedonic tone of the affective state), Arousal–Nonarousal (measuring the physiological activation, alertness, or neurovegetative motor mobilization), and Dominance–Submissiveness (measuring perceived agency, mastery, autonomy, or environmental control). In its classic operationalization, the instrument comprises 18 bipolar semantic differential item pairs (six items per dimension), evaluated on a 9-point response continuum traditionally scored from −4 to +4 or from 1 to 9. Decades of empirical inquiry in environmental psychology, architectural design, consumer marketing, human-computer interaction (HCI), and neuromarketing have established robust psychometric properties for the PAD scale. Reliability analyses across diverse cultural and operational settings consistently demonstrate high internal consistency, with Cronbach’s alpha coefficients typically ranging from .80 to .93 for Pleasure, .75 to .88 for Arousal, and .70 to .85 for Dominance. Construct, convergent, and discriminant validities have been corroborated against non-verbal measures (such as the Self-Assessment Manikin, SAM) and discrete emotion inventories, cementing the PAD framework as a central theoretical bridge within the Stimulus-Organism-Response (S-O-R) paradigm.

2. Keywords

PAD Emotional State Scale, Pleasure-Arousal-Dominance, Albert Mehrabian, James A. Russell, environmental psychology, affective dimensional model, Stimulus-Organism-Response (S-O-R), consumer behavior, semantic differential, emotion measurement, psychometrics, valence, physiological arousal, perceived dominance.

3. Authors

The PAD Emotional State Scale was conceived and validated by:

• Albert Mehrabian, Ph.D.: Professor Emeritus of Psychology at the University of California, Los Angeles (UCLA). Renowned for his pioneering investigations into nonverbal communication, implicit communication of affect, personality traits, and spatial environmental psychology.
• James A. Russell, Ph.D.: Professor of Psychology at Boston College (formerly at the University of British Columbia). A leading figure in affective science, widely recognized for his theoretical contributions to the circumplex model of affect, core affect theory, and the cross-cultural universality of emotional experience.

4. Purpose

The primary objective of the PAD Emotional State Scale is to quantitatively capture momentary affective states induced by ambient, social, and structural environmental stimuli. Before the formalization of the PAD model in 1974, researchers lacked a standardized, domain-general framework capable of bridging external physical attributes (e.g., lighting, acoustic levels, crowding, architectural layout) with immediate subjective psychological states and subsequent behavioral trajectories.

In environmental psychology and architectural design, the scale provides an empirical diagnostic tool to determine how spatial configurations, institutional spaces (e.g., psychiatric wards, acute hospitals, correctional facilities, higher education classrooms), and ambient conditions elicit stress or restoration. Rather than relying on categorical taxonomies of emotion (such as sadness, joy, anger, or fear), the PAD scale views emotional states as dynamic positions across three fundamental, continuous axes, offering greater sensitivity to subtle, sub-clinical, and ambient-driven emotional shifts.

In consumer behavior and sensory marketing, the PAD framework achieved widespread prominence through the pioneering work of Donovan and Rossiter (1982), who integrated the scale into the Stimulus-Organism-Response (S-O-R) paradigm. In retail environments, physical and digital atmospherics—including background music tempo, chromatic warmth, ambient olfaction, spatial density, and website navigation architecture—operate as environmental stimuli (S) that modulate the internal emotional organism (O) represented by Pleasure, Arousal, and Dominance. These internal states, in turn, directly govern behavioral responses (R), such as dwell time, in-store exploration, interpersonal interaction with sales personnel, impulse spending, and repurchase intentions.

Beyond traditional physical environments, the instrument is widely applied in human-computer interaction (HCI), digital user experience (UX) research, virtual reality (VR) immersion studies, and affective computing. By assessing user states along Pleasure, Arousal, and Dominance, software architects and game designers quantify emotional engagement, cognitive load, user frustration, and virtual presence, allowing programmatic optimization of digital interfaces and autonomous agent responsiveness.

5. Psychological Construct

The psychological architecture of the PAD Emotional State Scale rests on the premise that human affective experience is continuous and multidimensional. Mehrabian and Russell demonstrated that three bipolar, orthogonal factors account for the vast majority of variance in emotional reactions:

Pleasure – Displeasure (P)

The Pleasure dimension represents the qualitative, evaluative, or hedonic valence of an emotional state. It captures the degree to which an individual feels good, joyful, contented, gratified, and positively valenced, as opposed to feeling unhappy, melancholic, annoyed, distressed, or dissatisfied. Pleasure corresponds to the primary evaluative system of the brain—the neurological reward versus aversion circuitry. In ambient environments, elevated pleasure is elicited by aesthetic coherence, thermal comfort, soothing auditory tones, and appealing visual symmetry. Low pleasure is provoked by noxious odors, auditory dissonance, physical discomfort, or environmental deprivation.

Arousal – Nonarousal (A)

The Arousal dimension represents the degree of energetic mobilization, cortical alertness, neurovegetative excitation, and motor readiness of the individual. Spanning a continuum from deep sleep, lethargy, or stupor to frantic excitation, frenzied alertness, or panic, arousal reflects the activation level of the reticular activating system (RAS) and sympathetic nervous system. Crucially, arousal is non-evaluative; it conveys bodily readiness rather than hedonic tone. When combined with high pleasure, high arousal yields sensations of ecstasy, adventure, excitement, and elation. Conversely, when combined with low pleasure (displeasure), elevated arousal manifests as acute anxiety, panic, distress, or rage. In environmental applications, high arousal is triggered by loud music, vibrant chromatic saturation (e.g., deep reds and yellows), high spatial complexity, and unpredictable stimuli.

Dominance – Submissiveness (D)

The Dominance dimension reflects the extent to which an individual perceives personal agency, situational mastery, behavioral autonomy, and freedom of choice within a given environment. At the dominant anchor, an individual feels influential, in control, autonomous, and capable of imposing their will on the physical or social surroundings. At the submissive anchor, an individual feels constrained, restricted, manipulated, powerless, guided, or psychologically dwarfed by the environment. Dominance is particularly sensitive to spatial layout, social density, bureaucratic complexity, and architectural scale. For example, monumental, cavernous architecture or crowded transit hubs can induce acute submissiveness, whereas private, easily navigable, highly customizable spaces restore psychological dominance.

6. Theoretical Framework

The theoretical roots of the PAD Emotional State Scale lie in the intersection of neobehaviorism, cognitive psychology, and physiological arousal models. The conceptual spine of the instrument is derived from the Stimulus-Organism-Response (S-O-R) paradigm, initially formalized by Robert S. Woodworth (1929) and later synthesized into environmental psychology by Mehrabian and Russell (1974).

Classical behaviorism (S-R) posited that human actions could be modeled exclusively as direct functional relationships between environmental inputs and outward behavioral outputs. Mehrabian and Russell rejected this direct determinism, arguing that environmental stimuli (S) are fundamentally mediated by the internal, neuroaffective state of the organism (O) before manifesting as behavioral responses (R). Within the PAD framework, emotional states constitute the primary, indispensable mediating mechanism of human adaptation:

Stimulus (Physical/Social Environment) → Organism (PAD Emotional State) → Response (Approach vs. Avoidance)

The ultimate behavioral outcomes (R) are conceptualized as an overarching continuum of Approach – Avoidance behaviors, which Mehrabian and Russell operationalized across four distinct behavioral dimensions:

• Physical Approach/Avoidance: The inclination to enter, remain in, explore, or conversely flee and avoid an environment.
• Affiliative Approach/Avoidance: Willingness to interact with, communicate with, and establish social contact with others present within the space.
• Work/Performance Approach/Avoidance: Willingness to concentrate, expend effort, solve problems, and optimize cognitive or manual performance.
• Consumption/Expenditure Approach/Avoidance: In retail settings, willingness to spend money, browse merchandise, and execute commercial transactions.

A central theoretical postulation of the Mehrabian-Russell framework is the interactive hypothesis: while Pleasure exerts a direct monotonic positive effect on approach behaviors, Arousal acts as an amplifier. Under pleasurable conditions, elevated arousal markedly amplifies approach tendencies, promoting exploration and spending. However, in an unpleasurable environment, elevated arousal intensifies negative valence, escalating avoidance, stress reactions, and rapid exit behaviors.

The inclusion of Dominance was a critical theoretical divergence from contemporaneous two-dimensional affect models (such as Wilhelm Wundt’s historical tridimensional system or James Russell’s later two-dimensional circumplex model of affect). Mehrabian contended that without Dominance, the affective state cannot fully differentiate between emotions that share identical coordinates on the Pleasure and Arousal planes. For instance, both anger and fear are characterized by low pleasure and high arousal; yet, anger entails high dominance (perceived agency, confrontation), whereas fear is governed by profound submissiveness (perceived vulnerability, powerlessness). Thus, Dominance provides essential diagnostic power in dynamic and socially contested spaces.

7. Validity

The psychometric validity of the PAD Emotional State Scale has been rigorously investigated across more than four decades of continuous empirical application:

Construct and Structural Validity

Construct validity was established in Mehrabian and Russell’s (1974) original psychometric batteries, wherein the 18 semantic differential pairs were demonstrated to account for the primary variance across hundreds of narrative descriptions of real and simulated environments. Subsequent structural analyses across various cross-cultural populations have confirmed the stability of the three-dimensional manifold. Multitrait-multimethod (MTMM) evaluations confirm that Pleasure, Arousal, and Dominance exhibit distinct divergent trajectories from trait personality measures (such as the Big Five or Eysenck’s Extraversion/Neuroticism), affirming that PAD reliably indexes transient affective states rather than static behavioral traits.

Convergent and Discriminant Validity

Convergent validity has been repeatedly demonstrated against physiological markers and alternative validated psychometric tools:

• The Pleasure subscale strongly correlates with the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988), exhibiting positive correlations with Positive Affect ($r \approx .65$ to $.78$) and robust negative correlations with Negative Affect ($r \approx -.55$ to $-.70$).
• The Arousal subscale correlates directly with autonomic biomarkers, including skin conductance response (galvanic skin response/electrodermal activity, $r \approx .50$ to $.68$), heart rate variability (HRV) spectral power, and pupil dilation metrics.
• The scale demonstrates high convergent validity ($r > .75$) with pictorial instruments such as the Self-Assessment Manikin (SAM; Bradley & Lang, 1994), which operationalizes the exact three PAD dimensions via non-verbal pictorial representations.
• Discriminant validity is supported by low inter-factor correlations between Pleasure and Arousal in normative environments, confirming the theoretical orthogonality of the dimensions under resting conditions (though weak correlations occasionally emerge in extreme consumer or survival contexts).

Predictive and Criterion-Related Validity

Predictive validity is extensively documented across real-world applied environments. In Donovan and Rossiter’s (1982) landmark investigation in department stores, Pleasure alone accounted for a substantial proportion of variance in in-store spending ($R^2 = .23$) and overall store patronage. In healthcare environments, high Dominance ratings by hospitalized patients prospectively predict lowered biological markers of distress (serum cortisol) and heightened subjective pain tolerance. In e-commerce environments, high Pleasure and moderate Arousal scores reliably predict conversion rates and prolonged dwell times.

8. Reliability

The PAD Emotional State Scale demonstrates high internal consistency and measurement precision across laboratory and naturalistic settings.

Internal Consistency

Original validation cohorts documented by Mehrabian and Russell (1974) yielded solid reliability estimates, which have been consistently replicated and surpassed in subsequent literature using modernized psychometric guidelines:

• Pleasure Subscale: Cronbach’s alpha coefficients regularly range from $\alpha = .85$ to $.93$, reflecting high item homogeneity. Average inter-item correlations typically fall within the optimal range of $.50$ to $.68$.
• Arousal Subscale: Cronbach’s alpha values typically span $\alpha = .78$ to $.88$. Occasional variations are observed when translating specific arousal adjectives (e.g., “frenzied” vs. “sluggish”) into non-English languages, necessitating careful linguistic validation.
• Dominance Subscale: Cronbach’s alpha values generally span $\alpha = .74$ to $.85$. While historically displaying slightly lower consistency than Pleasure, contemporary revisions and purified 4-item or 6-item versions routinely achieve $\alpha > .80$.
• Composite Reliability (CR) & McDonald’s Omega: Structural equation modeling studies routinely report composite reliabilities exceeding the standard $.70$ threshold, with $\omega$ values frequently matching or exceeding Cronbach’s alpha values across all three latent factors.

Test-Retest Reliability and State Sensitivity

Because the PAD scale measures transient, dynamic state affect, classic test-retest reliability across long temporal intervals is theoretically contraindicated—the instrument must demonstrate acute sensitivity to environmental shifts. However, in controlled laboratory settings where environmental conditions remain invariant over brief intervals (e.g., 10 to 30 minutes), test-retest reliability coefficients remain high ($r_{tt} > .80$). When experimental conditions are intentionally altered, the PAD subscales exhibit immediate, statistically significant sensitivity, confirming the scale’s responsiveness to situational manipulation.

9. Factor Analysis

The factor structure of the 18-item PAD Emotional State Scale has been extensively evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

In original and early replication studies utilizing Principal Component Analysis (PCA) and Common Factor Analysis with orthogonal (Varimax) and oblique (Promax) rotations, three distinct eigenvalues greater than 1.0 reliably emerge (Kaiser-Guttman criterion). These three factors account for approximately 58% to 68% of the total variance across diverse experimental settings:

• Factor 1 (Pleasure): Items such as Happy–Unhappy, Pleased–Annoyed, and Contented–Melancholic consistently display primary factor loadings between $.72$ and $.89$, with cross-loadings onto Arousal and Dominance typically falling well below $.25$.
• Factor 2 (Arousal): Items such as Stimulated–Relaxed, Excited–Calm, and Wide-awake–Sleepy load heavily onto the second factor (.68 to .84). Minor cross-loadings occasionally appear between the Relaxed–Bored item and Pleasure, leading some researchers to drop or substitute this specific pair.
• Factor 3 (Dominance): Items such as Controlling–Controlled, Influential–Influenced, and Autonomous–Guided load between $.62$ and $.83$ on the third factor, confirming a clear tripartite structure.

Confirmatory Factor Analysis (CFA)

Contemporary structural equation modeling routinely tests the hypothesized three-factor oblique model against alternative nested configurations (e.g., one-factor unidimensional models or two-factor Pleasure/Arousal models). The three-factor model consistently demonstrates superior goodness-of-fit indices:

• Comparative Fit Index (CFI): Values routinely exceed $.92$, frequently reaching $.95$ or higher in purified iterations.
• Tucker-Lewis Index (TLI): Typically ranges between $.91$ and $.95$.
• Root Mean Square Error of Approximation (RMSEA): Generally falls within acceptable parameters, ranging between $.045$ and $.068$ (with 90% confidence intervals below $.08$).
• Standardized Root Mean Square Residual (SRMR): Values consistently settle below $.055$.

CFA findings conclusively corroborate that allowing weak correlations among the three latent constructs provides optimal structural fit, while preserving the practical conceptual autonomy of Pleasure, Arousal, and Dominance.

10. Instrument / Measurement Tool

• Instrument Name: PAD Emotional State Scale (often referred to as the Mehrabian-Russell PAD Scale).
• Original Authors: Albert Mehrabian and James A. Russell (1974).
• Constructs Assessed: Three continuous dimensions of momentary emotional state:
  • Pleasure – Displeasure (P)
  • Arousal – Nonarousal (A)
  • Dominance – Submissiveness (D)
• Total Item Count: 18 bipolar adjective pairs (6 items per subscale).
  • Pleasure: 6 item pairs
  • Arousal: 6 item pairs
  • Dominance: 6 item pairs
• Response Format: 9-point semantic differential scale (scored numerically as −4 to +4, or 1 to 9). Respondents mark a point along the nine-interval continuum between opposing polar adjectives.
• Scoring Guidelines:
  • Reverse coding is applied to negatively oriented poles to ensure that higher numeric scores uniformly represent higher levels of Pleasure, Arousal, and Dominance.
  • Subscale scores are computed by summing or averaging the six individual item scores corresponding to that dimension.
  • Dimension indices can range from −4.0 to +4.0 (or 1.0 to 9.0), establishing an overall coordinate in three-dimensional affective space: $(P, A, D)$.
• Administration Modality: Self-administered paper-and-pencil or digital survey format (web/mobile/laboratory computer). Pictorial derivatives (e.g., SAM) can be employed for populations with language or literacy constraints.
• Estimated Completion Time: Approximately 2 to 4 minutes.

11. Permissions, Fee, and Test Year

• Year of Initial Publication: 1974 (published in *An Approach to Environmental Psychology* by MIT Press).
• Copyright & Ownership: The conceptual framework, original validation data, and initial item sets were published by Albert Mehrabian and James A. Russell in their 1974 academic monograph (MIT Press). Commercial software adaptations and proprietary consulting implementations are held by Albert Mehrabian / Mediax Corporation.
• Academic and Non-Commercial Research Use: The scale is widely considered an open academic measurement paradigm. Scholars, university researchers, and students routinely utilize and adapt the 18 semantic differential items for non-profit academic research, scientific inquiry, and educational theses without licensing fees, provided formal scholarly citation is credited to Mehrabian and Russell (1974).
• Commercial Applications: Commercial organizations and market research firms planning proprietary software integration or branded enterprise diagnostics should consult current intellectual property holders or obtain permissions through standard publisher clearance channels.

12. References

• Bakker, I., van der Voordt, T., Vink, P., & de Boon, J. (2014). Pleasure, Arousal, Dominance: Mehrabian and Russell revisited. Current Psychology, 33(3), 405–421. https://doi.org/10.1007/s12144-014-9219-4
• 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
• Donovan, R. J., & Rossiter, J. R. (1982). Store atmosphere: An environmental psychology approach. Journal of Retailing, 58(1), 34–57.
• Mehrabian, A. (1995). Framework for a comprehensive description and measurement of emotional states. Genetic, Social, and General Psychology Monographs, 121(3), 339–361.
• Mehrabian, A., & Russell, J. A. (1974). An approach to environmental psychology. MIT Press.
• 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., & Pratt, G. (1980). A description of the affective quality attributed to environments. Journal of Personality and Social Psychology, 38(2), 311–322. https://doi.org/10.1037/0022-3514.38.2.311
• 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

13. Items of the Scale