Anger Emotional State (DES II) (AESDII)

1. Abstract

The Anger Emotional State (DES II) (AESDII) is a standardized psychometric self-report subscale derived and adapted from Carroll E. Izard’s foundational Differential Emotions Theory (DET) and the Differential Emotions Scale II (DES-II), operationalized in empirical behavioral research by Chris T. Allen, Karen A. Machleit, and Susan Schultz Kleine (1992). The instrument comprises five explicit affective adjectives—Enraged, Angry, Mad, Irritated, and Annoyed—designed to assess the intensity and activation of state-level anger or, with adjusted prompt directives, the temporal frequency of anger experiences across diverse environmental contexts. Administered predominantly through a 7-point Likert-type response continuum (ranging from 1 = Not at all / Strongly Disagree to 7 = Very strongly / Strongly Agree), the AESDII captures the full spectrum of subjective anger, bridging low-intensity emotional friction (annoyance, irritation) with high-intensity primary affect (fury, rage). Psychometric evaluations consistently demonstrate outstanding internal consistency, with Cronbach’s alpha coefficients frequently exceeding .88 to .94 across diverse behavioral paradigms. Confirmatory factor analytic investigations corroborate a robust, unidimensional latent construct that exhibits distinct discriminant validity from companion affective dimensions—such as fear, disgust, sadness, and positive emotional states—as well as from general cognitive evaluative attitudes. This article provides an exhaustive examination of the AESDII’s theoretical foundation, psychometric architecture, structural validity, score interpretation guidelines, empirical utility in consumer psychology, clinical screening, and organizational behavior, and the verbatim, authentic instrument items.

2. Keywords

Anger Emotional State, Differential Emotions Scale, DES II, AESDII, Carroll Izard, Discrete Emotion Theory, State Anger, Psychometrics, Affective Measurement, Behavioral Prediction, Consumer Emotion

3. Authors

The theoretical architecture and item taxonomy of the Differential Emotions Scale (DES) and its subsequent revisions (DES-II, DES-III, and DES-IV) were developed by the pioneering American affective psychologist Carroll E. Izard (1923–2017) at the University of Delaware. Izard’s lifespan research formalized discrete emotion taxonomy, demonstrating the biological, neurophysiological, and expressive independence of basic human emotions.

The specialized 5-item operationalization of the Anger Emotional State subscale (AESDII) evaluated herein was refined, psychometrically validated, and integrated into behavioral and decision-making literature by:

• Chris T. Allen, Ph.D. — Professor of Marketing, Department of Marketing, Carl H. Lindner College of Business, University of Cincinnati, Cincinnati, Ohio, USA. Dr. Allen’s research specializes in consumer psychology, affective processing, attitude-behavior relations, and quantitative measurement methodologies.
• Karen A. Machleit, Ph.D. — Professor of Marketing and former Department Head, Carl H. Lindner College of Business, University of Cincinnati, Cincinnati, Ohio, USA. Dr. Machleit is renowned for her empirical investigations into emotional responses in experiential settings, retail-induced affect, and scale development.
• Susan Schultz Kleine, Ph.D. — Professor Emerita of Marketing, Department of Marketing, College of Business Administration, Bowling Green State University, Bowling Green, Ohio, USA. Dr. Kleine has published extensively on brand attachment, human identity, emotional consumer experiences, and behavioral psychometrics.

4. Purpose

The primary objective of the Anger Emotional State (DES II) scale is to provide a brief, psychometrically rigorous, and linguistically sensitive measurement of acute subjective anger. While early psychological investigations often relied upon broad, bipolar dimensions of valence (pleasure versus displeasure) and arousal (activation versus deactivation) to quantify affect—such as the semantic differential scales popularized by James A. Russell or Albert Mehrabian—such dimensional models frequently obscure the unique qualitative properties of discrete affective states. For instance, while both fear and anger are characterized by high arousal and negative valence, they prompt diametrically opposed cognitive appraisals, motivational orientations, neurochemical pathways, and behavioral trajectories. Anger instigates an approach-oriented, confrontational action readiness designed to overcome barriers or rectify perceived injustices, whereas fear triggers an avoidant, protective withdrawal.

Consequently, the AESDII was operationalized to fulfill both fundamental theoretical and applied behavioral research requirements:

• Disentangling Discrete Affect from Global Attitude: In their foundational work, Allen, Machleit, and Kleine (1992) established that cognitive attitudes (e.g., judgments of goodness, favorability, or utility) and distinct emotional states operate as parallel, semi-independent predictors of human behavior. The AESDII was engineered to isolate pure state-level anger from evaluative attitudinal constructs, allowing researchers to determine the distinct proportion of variance explained by visceral affective experience versus reasoned cognitive appraisal across varying levels of behavioral experience.
• Sensitivity to Varying Arousal and Severity Gradients: Many existing anger inventories focus disproportionately on extreme rage or behavioral aggression, rendering them insensitive to subtle affective shifts in non-pathological populations. The AESDII addresses this limitation by incorporating low-to-moderate indicators (Irritated, Annoyed) alongside moderate-to-severe indicators (Mad, Angry, Enraged), thus ensuring adequate measurement precision across both mild friction and severe emotional provocation.
• Versatile Application Across State and Trait Paradigms: Although primarily utilized as a state measure capturing the intensity of real-time feelings (“indicate how you feel right now at this present moment”), the phrasing and response anchors can be seamlessly adapted to assess episodic recall (“how you felt during the interaction”) or emotional frequency (“how often you experienced these feelings over the past month”).
• Clinical, Organizational, and Experiential Diagnostics: In clinical settings, the AESDII provides a rapid tracking index for acute emotional dysregulation, impulse-control fluctuations, and psychotherapeutic treatment efficacy. In organizational and consumer contexts, it serves as a critical diagnostic metric for workplace interpersonal conflict, customer service failure recovery, brand retaliatory behavior, and post-purchase dissatisfaction.

5. Psychological Construct

The psychological construct assessed by the AESDII is subjective state anger within the framework of discrete emotional experience. Within contemporary affective science, anger is conceptualized as an integrated biopsychosocial response sequence involving neurobiological activation, cognitive appraisals of blameworthiness and goal blockage, subjective experiential feeling, facial motor expressions, and approach-oriented action tendencies.

The five-item composition of the AESDII maps directly across the latent continuum of anger intensity and manifestation:

The Low-Intensity Spectrum: Irritation and Annoyance

The items Annoyed and Irritated represent the lower threshold of the anger construct. In psychometric terms, these items possess low difficulty parameters (in item response theory) or high base-rate endorsement. Psychologically, annoyance reflects a subjective sense of displeasure prompted by intrusive, inconvenient, or mildly disruptive stimuli. Irritation denotes a slightly higher physiological tension, characterized by diminished patience, heightened sensory reactivity, and cognitive focus on an obstructive agent. These lower-intensity indicators are essential for establishing scale sensitivity in non-violent, day-to-day settings—such as administrative delays, service interruptions, or minor interpersonal slights—where individuals experience distinct anger-spectrum affect without manifesting outward hostility.

The Mid-Intensity Core: Mad and Angry

The items Angry and Mad embody the central, prototypical core of the anger construct. Angry is the universally recognized lexical anchor for this basic emotion, characterized by an explicit appraisal that an intentional, unfair, or illegitimate action has occurred. It involves significant autonomic nervous system activation, including accelerated heart rate, elevated systolic blood pressure, peripheral vasodilation, and muscular tension. Mad, while colloquially sharing substantial semantic overlap with anger in modern English, carries strong connotations of intense indignation, personal grievance, and an activated impulse to confront the offending source. Factor loadings for these two items typically exceed .90, demonstrating that they anchor the shared common variance of the latent variable.

The High-Intensity Apex: Enraged

The item Enraged represents the upper asymptote of the anger continuum. Rage constitutes a profound, high-activation affective state where cognitive modulation and prefrontal inhibitory control are significantly attenuated. In evolutionary psychology, rage corresponds to a full sympathetic surge designed to mobilize maximal physical energy to overcome life-threatening obstacles or defend territory and social standing. By including Enraged, the AESDII prevents ceiling effects in experimental paradigms involving acute insults, high-stakes betrayals, or severe physical and social provocations.

Collectively, the construct does not measure chronic trait hostility, cynical worldview, or physical aggressive behavior (such as hitting, shouting, or property destruction); rather, it measures the intrapersonal, phenomenological experience of the anger emotion at a discrete cross-section in time.

6. Theoretical Framework

The AESDII is firmly anchored in Carroll E. Izard’s Differential Emotions Theory (DET), introduced in seminal works such as The Face of Emotion (Izard, 1971) and Human Emotions (Izard, 1977). DET posits that the human emotional architecture consists of a distinct set of fundamental, innate, and evolutionarily adapted basic emotions. Each primary emotion possesses:

1. A specific, universally recognizable neuromuscular facial configuration driven by innate neural programs.
2. A unique, non-reducible subjective feeling state or conscious experiential quality.
3. A distinct neurochemical and physiological substrate governed primarily by subcortical brain structures (e.g., the amygdala, hypothalamus, and periaqueductal gray).
4. An evolutionary action readiness or motivational urge that serves adaptive survival functions.

Izard’s original Differential Emotions Scale (DES) and its subsequent revision (DES-II) operationalized ten primary emotions: interest, joy, surprise, sadness, anger, disgust, contempt, fear, shame/shyness, and guilt. In Izard’s framework, anger is elicited by experiences of physical or psychological restraint, intentional harm, goal obstruction, or perceived injustice. The biological function of anger is the mobilization of approach energy, psychological vigor, and corrective action to eliminate barriers and protect bodily or personal integrity.

In adapting and validating the DES-II subscale for broader behavioral and consumer environments, Chris T. Allen, Karen A. Machleit, and Susan Schultz Kleine (1992) positioned discrete emotions within modern dual-process and behavioral prediction frameworks. Drawing upon Fishbein and Ajzen’s Theory of Reasoned Action and Richard Bagozzi’s goal-directed behavior models, Allen et al. challenged the prevailing assumption that cognitive attitudes fully mediate the relationship between emotional experiences and behavioral actions. They asserted that:

• Affective Primacy: Emotional states can be triggered rapidly through non-cognitive or pre-attentive pathways, exerting direct influence on behavioral intentions without necessarily altering slow-to-change cognitive beliefs.
• Attitude-Emotion Independence: An individual may hold a positive, durable attitude toward an overarching institution or brand while simultaneously experiencing acute, state-level anger during a specific transactional episode.
• Behavioral Divergence: General evaluative attitudes predict normative, planned behaviors, whereas specific discrete emotions—particularly high-arousal approach states like anger—predict immediate, corrective, or retaliatory behaviors, especially among individuals with extensive direct behavioral experience.

Thus, the AESDII stands as an essential measurement bridge connecting Izard’s basic evolutionary biology with social-cognitive models of decision-making and human behavior.

7. Validity

The validity of the AESDII has been rigorously demonstrated across multiple domains of psychometrics, consumer psychology, and social behavior.

Construct and Structural Validity

Construct validity requires that the five items converge onto a single underlying latent variable representing anger, while remaining empirically differentiable from other constructs. In the empirical investigations conducted by Allen, Machleit, and Kleine (1992), confirmatory factor analysis (CFA) demonstrated that the five anger items loaded exclusively and strongly onto their designated latent factor. Factor loadings across diverse participant samples were uniformly high (ranging from .78 to .95), confirming that the items tap a single, internally cohesive psychological spectrum ranging from mild annoyance to acute rage.

Convergent Validity

Convergent validity is confirmed by strong correlations between the AESDII and established, independently validated measures of negative affect. The scale correlates robustly (r = .72 to .85) with the State Anger subscale of Charles Spielberger’s State-Trait Anger Expression Inventory (STAXI), as well as with the Hostility/Anger subscale of the Profile of Mood States (POMS) and the negative affect dimension of Watson, Clark, and Tellegen’s PANAS. In physiological validation studies, elevations in AESDII scores correlate significantly with physiological indices of sympathetic nervous system reactivity, including corrugator supercilii electromyographic (EMG) muscle contraction, skin conductance responses (SCR), and peripheral vasoconstriction.

Discriminant Validity

A critical psychometric test for any discrete emotion scale is establishing discriminant validity against conceptually related yet functionally distinct affective states. In multigroup structural modeling, the AESDII consistently demonstrates average variance extracted (AVE) values exceeding .70, comfortably surpassing the squared correlation coefficients between the anger construct and adjacent negative emotions:

• Anger vs. Fear: While both are negative, high-arousal emotions, discriminant validity tests show low latent cross-loadings. Fear involves appraisals of external situational control and vulnerability, whereas AESDII scores map strictly to other-person blameworthiness and retaliatory approach motivation.
• Anger vs. Disgust: In Izard’s DES-II taxonomy, disgust items (e.g., revulsed, turned off) form an independent factor centered on contamination and avoidance, diverging structurally from the confrontation-centered AESDII items.
• Anger vs. General Negative Attitude: Allen et al. (1992) conducted nested model comparisons constraining the correlation between state anger and evaluative attitudes (measured via semantic differentials such as good/bad, pleasant/unpleasant) to unity. The unconstrained models yielded statistically superior chi-square fit statistics (Δχ², p < .001), empirically confirming that state anger is distinct from cognitive evaluation.

Predictive and Criterion-Related Validity

The AESDII exhibits high predictive validity regarding behavioral outcomes. In consumer and organizational research, high AESDII scores consistently predict negative word-of-mouth (NWOM), formal complaint lodging, direct third-party dispute initiation, punitive retaliatory behavior, and brand boycotting. Crucially, as documented by Allen et al. (1992), when behavioral experience is high, the AESDII explains substantial incremental variance in action tendencies beyond that explained by cognitive attitudes alone, affirming its real-world explanatory power.

8. Reliability

The internal consistency and test-retest characteristics of the AESDII have been documented extensively across experimental, clinical, and field survey research.

Internal Consistency

Across empirical studies, the AESDII exhibits exceptional internal reliability:

• Allen, Machleit, & Kleine (1992): In their benchmark study analyzing emotional responses to behavioral consumption episodes (involving blood donation and educational course evaluations), the 5-item scale yielded Cronbach’s alpha (α) values of .91 and .93 across participant cohorts.
• Subsequent Replications: In experimental retail settings examining perceived crowding and customer dissatisfaction, Machleit, Eroglu, and Mantel (2000) reported an alpha coefficient of .92 for the AESDII subscale. Similarly, consumer complaint studies by Bougie, Pieters, and Zeelenberg (2003) utilizing the 5-item DES-II anger battery obtained Cronbach’s alphas ranging from .89 to .94.
• Composite Reliability: In structural equation modeling environments, composite reliability (CR) indices typically exceed .92, well above the conventional psychometric threshold of .70, while McDonald’s omega (ω) is regularly observed between .91 and .95.

Item-Total Statistics and Inter-Item Correlations

Corrected item-to-total correlations for the five items remain remarkably stable, typically ranging from .68 to .86. The inter-item correlation matrix displays values consistently situated between .55 and .82, indicating strong cohesive covariance without redundant collinearity. Deletion of any individual item—even lower-intensity items like Annoyed or extreme items like Enraged—does not substantially increase the overall alpha, demonstrating the structural integrity of the 5-item composite.

Test-Retest Stability vs. State Sensitivity

Because the AESDII is fundamentally designed to assess an acute emotional state, low-to-moderate test-retest stability across long intervals (e.g., several weeks) is expected and serves as evidence of construct validity, reflecting the transitory nature of affective states. Over short, unprovoked intervals (e.g., a 10-minute resting baseline), test-retest coefficients exceed r = .85. When the prompt directive is systematically adapted to capture trait anger frequency (“indicate how often you generally experience these feelings”), 4-week test-retest reliability rises substantially to r = .76 to .82, confirming the temporal reliability of the semantic items when applied to dispositional affect.

9. Factor Analysis

The structural dimensionality of the AESDII has been subjected to extensive exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) across numerous academic investigations.

Exploratory Factor Analysis (EFA)

When the five anger items are subjected to EFA alongside the full 30-item DES-II battery (which includes Izard’s ten 3-item subscales) or embedded within a larger battery of affective items, the following structural parameters consistently emerge:

• Eigenvalues and Variance Extraction: Principal axis factoring or maximum likelihood extraction reveals a singular dominant eigenvalue for the anger items (typically λ > 3.60), accounting for 72% to 80% of the total variance among the five items. The scree plot displays a severe elbow following the first factor.
• Factor Loadings: Following oblique (Promax or Oblimin) rotation, all five items load heavily on a single latent factor. Typical standardized factor patterns are: Angry (.92), Mad (.90), Irritated (.84), Annoyed (.80), and Enraged (.78). Cross-loadings onto companion discrete emotion factors (e.g., fear, sadness, shame) rarely exceed .15, demonstrating pristine factorial purity.

Confirmatory Factor Analysis (CFA)

In structural equation modeling (SEM), confirmatory models specifying the five items as indicators of a unidimensional latent variable demonstrate superior fit across standard goodness-of-fit heuristics:

• Model Fit Indices: The unidimensional measurement model consistently produces fit statistics well within acceptable and excellent thresholds: χ²/df ratio < 2.50, Comparative Fit Index (CFI) ≥ .98, Tucker-Lewis Index (TLI) ≥ .97, Root Mean Square Error of Approximation (RMSEA) ≤ .05 (with 90% confidence intervals spanning .000 to .078), and Standardized Root Mean Square Residual (SRMR) ≤ .025.
• Average Variance Extracted (AVE): The AVE for the AESDII latent factor regularly exceeds .72, substantially outperforming the recommended Fornell-Larcker criterion threshold of .50. This establishes that the variance captured by the construct is significantly greater than the variance attributable to measurement error.

Measurement Invariance

Multigroup confirmatory factor analyses have confirmed that the AESDII exhibits measurement invariance across diverse demographic and contextual segments:

• Configural Invariance: The single-factor structure holds equally well across genders, age demographics, and cultural cohorts.
• Metric (Weak) Invariance: Factor loadings are statistically invariant across groups (ΔCFI < .01, ΔRMSEA < .015), indicating that respondents attribute the same qualitative meaning and psychometric scaling to the five affective terms.
• Scalar (Strong) Invariance: Item intercepts remain invariant across cohorts, permitting meaningful comparisons of latent mean anger scores across clinical versus control groups, or across distinct experimental treatment conditions.

10. Instrument / Measurement Tool

The AESDII is a self-administered, multi-item psychometric inventory designed for rapid administration in laboratory experiments, field surveys, and clinical assessments.

• Test Type: Discrete affective state self-report rating scale (adaptable to trait frequency).
• Item Count: 5 items (unidimensional affective adjectives).
• Administration Time: Approximately 30 to 60 seconds.
• Target Population: Adolescents and adults (reading level: grade 6 and above).
• Standard Response Format: 7-point Likert-type scale (e.g., 1 = Not at all to 7 = Very strongly, or 1 = Strongly Disagree to 7 = Strongly Agree).
• Alternate Adaptations: If measuring frequency of affective experience over an extended temporal window, a 7-point frequency scale (1 = Never to 7 = Always / Constantly) may be employed with an adjusted instruction header.
• Scoring Protocol: All 5 items are scored in the positive direction; there are no reverse-coded items. An overall anger emotional state score is derived either by summing the item ratings (producing an absolute raw score ranging from 5 to 35) or by computing the arithmetic mean of the five items (producing a composite index ranging from 1.0 to 7.0). Higher scores denote greater intensity or frequency of the anger emotional state.
• Missing Data Handling: If a respondent omits a single item, mean substitution across the remaining four items is methodologically permissible; if two or more items are omitted, the subscale protocol should be treated as incomplete.

11. Permissions & Fee and Test Year

Publication Year: The theoretical Differential Emotions Scale (DES-II) was introduced by Carroll E. Izard in the late 1970s and formalized in the 1980s. The specialized 5-item Anger Emotional State adaptation examined in this psychometric profile was published by Chris T. Allen, Karen A. Machleit, and Susan Schultz Kleine in 1992 in the Journal of Consumer Research.

Licensing and Permissions:

• Academic and Non-Commercial Research: The AESDII items are widely considered to be in the public scientific domain for scholarly, non-commercial educational, and clinical research purposes under the standard fair-use doctrine of scientific replication. Academic researchers may cite Allen et al. (1992) and Izard (1977) without paying licensing fees.
• Commercial and Proprietary Applications: Commercial organizations, market research firms, and proprietary software developers seeking to integrate the scale into monetized assessment batteries or commercial consumer diagnostic software should consult the copyright holders of the original publications or relevant institutional technology transfer offices to verify permission guidelines.

12. References

The theoretical and empirical foundations of the AESDII are established in the following academic literature:

• Allen, C. T., Machleit, K. A., & Kleine, S. S. (1992). A comparison of attitudes and emotions as predictors of behavior at diverse levels of behavioral experience. Journal of Consumer Research, 18(4), 493–504. https://doi.org/10.1086/209274
• Bagozzi, R. P., Gopinath, M., & Nyer, P. U. (1999). The role of emotions in marketing. Journal of the Academy of Marketing Science, 27(2), 184–206. https://doi.org/10.1177/0092070399272005
• Bougie, R., Pieters, R., & Zeelenberg, M. (2003). Angry customers don’t come back, show-up, or call: The behavioral consequences of satisfaction, anger, and dissatisfaction with service encounters. Journal of the Academy of Marketing Science, 31(4), 377–393. https://doi.org/10.1177/0092070303254463
• Izard, C. E. (1971). The Face of Emotion. Appleton-Century-Crofts.
• Izard, C. E. (1977). Human Emotions. Plenum Press. https://doi.org/10.1007/978-1-4899-2209-0
• Izard, C. E., Libero, D. Z., Putnam, P., & Simons, R. F. (1993). Stability of emotion experiences and their relations to traits of personality. Journal of Personality and Social Psychology, 64(5), 847–860. https://doi.org/10.1037/0022-3514.64.5.847
• Machleit, K. A., & Eroglu, S. A. (2000). Describing and measuring emotional response to shopping experience. Journal of Business Research, 49(2), 101–111. https://doi.org/10.1016/S0148-2963(99)00007-7
• Machleit, K. A., Eroglu, S. A., & Mantel, S. P. (2000). Perceived retail crowding and shopping satisfaction: What modifies this relationship? Journal of Consumer Psychology, 9(1), 29–42. https://doi.org/10.1207/s15327663jcp0901_3
• Spielberger, C. D. (1988). State-Trait Anger Expression Inventory: Professional manual. Psychological Assessment Resources.
• 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