State-Trait Anger Expression Inventory (STAXI-2)

Abstract

The State-Trait Anger Expression Inventory-2 (STAXI-2) is a standardized, multidimensional psychometric instrument developed by Charles D. Spielberger to assess the experience, expression, and control of anger across clinical, forensic, occupational, and research settings. Originating as an expansion and refinement of the original STAXI, the 57-item self-report questionnaire distinguishes between anger as a transient emotional state (state-trait distinction) and anger as a relatively stable personality disposition, while further delineating the cognitive and behavioral mechanisms through which anger is expressed or suppressed. The STAXI-2 comprises six primary scales and five subscales: State Anger (further subdivided into Feeling Angry, Feel Like Expressing Anger Verbally, and Feel Like Expressing Anger Physically), Trait Anger (subdivided into Angry Temperament and Angry Reaction), Anger Expression-Out (AX-O), Anger Expression-In (AX-I), Anger Control-Out (AC-O), and Anger Control-In (AC-I). An overarching metric, the Anger Expression Index (AX Index), integrates these dimensions to provide an overall index of anger expression. State items are evaluated on a 4-point Likert scale reflecting current intensity (1 = Not at all to 4 = Very much so), whereas Trait and Anger Expression/Control items are scored on a 4-point Likert frequency scale (1 = Almost never to 4 = Almost always). Extensive psychometric evaluations demonstrate high internal consistency (Cronbach’s alpha coefficients frequently ranging from .73 to .95 across normative and clinical samples), robust factorial validity supported by exploratory and confirmatory factor analyses, and sound convergent and discriminant validity across psychiatric, medical (e.g., cardiovascular disease), and correctional populations.

Keywords

STAXI-2, State-Trait Anger Expression Inventory, Charles D. Spielberger, State Anger, Trait Anger, Anger Expression-Out, Anger Expression-In, Anger Control, Psychometrics, Factor Structure, Assessment of Aggression, Hostility, Cardiovascular Reactivity

Authors

The State-Trait Anger Expression Inventory-2 was authored by Charles Donald Spielberger, Ph.D., ABPP (1927–2013), an internationally recognized pioneer in clinical psychology and psychometrics. During the conceptualization and development of the STAXI-2, Dr. Spielberger served as a Distinguished Research Professor of Psychology and Director of the Center for Research in Behavioral Medicine and Health Psychology at the University of South Florida, Tampa, USA.

Throughout his prolific career, Spielberger served as President of the American Psychological Association (APA, 1991–1992) and the International Assessment Council. His work fundamentally transformed psychological assessment through the formal operationalization of state and trait distinctions, which originated with the State-Trait Anxiety Inventory (STAI) and culminated in the comprehensive measurement of anger, anxiety, curiosity, and depression. Inquiries regarding official psychometric testing kits, scoring software, and licensing rights are managed through the publisher, Psychological Assessment Resources (PAR), located in Lutz, Florida.

Purpose

The primary purpose of the STAXI-2 is to provide an empirically grounded, nuanced assessment of the multidimensional experience of anger. Anger represents a functional yet potentially maladaptive primary emotion characterized by subjective feelings that vary in intensity from mild irritation to intense fury, typically accompanied by autonomic nervous system activation. Despite its evolutionary significance, chronic or poorly regulated anger is implicated in diverse psychiatric disorders, interpersonal violence, workplace disruption, and somatic conditions—most notably essential hypertension and coronary artery disease. The STAXI-2 was designed to disentangle the phenomenological experience of anger from its behavioral expression and voluntary self-regulation.

Clinical Applications

In clinical psychology and psychiatry, the STAXI-2 is employed to detect emotional dysregulation across various diagnostic presentations, such as intermittent explosive disorder, borderline personality disorder, post-traumatic stress disorder (PTSD), and substance use disorders. Clinicians use the scale’s profile to ascertain whether a patient suffers from heightened physiological-affective arousal (high State Anger), a dispositional hypersensitivity to perceived provocations (high Trait Anger), an outward aggressive behavioral trajectory (high AX-O), or internalizing rumination (high AX-I). Consequently, the inventory informs targeted psychotherapy, distinguishing clients who require cognitive restructuring and perceptual defusion (for high Angry Temperament) from those needing dialectical behavioral distress tolerance or assertiveness training (for high AX-I or low AC-O/AC-I).

Health Psychology and Behavioral Medicine

In health psychology and behavioral medicine, the STAXI-2 serves as an indispensable tool for investigating the cardiovascular reactivity hypothesis. Research demonstrates that whereas anger is an inevitable emotional response, chronic internal suppression (Anger-In) and unchecked hostile confrontations (Anger-Out) exert detrimental effects on vascular tone, endothelial function, platelet aggregation, and sympathetic-adrenomedullary (SAM) axis activity. Researchers utilize the STAXI-2 to isolate the toxic components of the historically debated “Type A” behavior pattern, showing that suppressed hostility is a potent independent risk factor for arterial hypertension and stroke.

Forensic and Correctional Assessment

Within forensic and correctional contexts, the STAXI-2 is routinely administered to evaluate violent offenders, individuals undergoing court-mandated domestic violence intervention, and candidates for parole. Because the instrument incorporates scales measuring both overt behavioral propulsion (Feel Like Expressing Anger Physically) and intentional self-restraint (Anger Control), it assists forensic psychologists in conducting violence risk assessments, determining institutional adjustment potential, and evaluating post-intervention reductions in emotional hostility.

Psychological Construct

The psychological construct evaluated by the STAXI-2 is anger, conceptualized by Spielberger as a multifaceted psychobiological emotional state accompanied by skeletal muscle tension, autonomic arousal, and cognitive appraisals of threat, injustice, or frustration. The STAXI-2 divides this construct into six primary scales and five subscales:

1. State Anger (S-Ang)

State Anger is defined as a transitory emotional condition characterized by subjective feelings of tension, annoyance, irritation, fury, or rage, accompanied by activation of the autonomic nervous system. The intensity of S-Ang fluctuates over time as a function of situational stressors and cognitive appraisals. The STAXI-2 captures the multifaceted nature of state anger through 15 items grouped into three distinct 5-item subscales:

• Feeling Angry (S-Ang/F): Assesses the intensity of pure, subjective affective feelings of anger at the moment of evaluation (e.g., “I am furious”, “I feel irritated”).
• Feel Like Expressing Anger Verbally (S-Ang/V): Captures the momentary motoric and subjective impulses to engage in verbal hostility (e.g., “I feel like yelling at somebody”, “I feel like cursing out loud”).
• Feel Like Expressing Anger Physically (S-Ang/P): Measures the immediate bodily action tendencies oriented toward overt destruction or physical confrontation (e.g., “I feel like breaking things”, “I feel like hitting someone”).

2. Trait Anger (T-Ang)

Trait Anger reflects enduring dispositional differences in the frequency, duration, and ease with which an individual experiences state anger across diverse situations. Individuals high in T-Ang perceive a wider array of circumstances as unfair, intentional provocations, or frustrating barriers. The 10-item Trait Anger scale is divided into two 4-item subscales (along with two additional general items):

• Angry Temperament (T-Ang/T): Measures a generalized, spontaneous disposition to experience and express anger quickly without specific provocation or identifiable external justification (e.g., “I am quick-tempered”, “I have a fiery temper”).
• Angry Reaction (T-Ang/R): Measures the dispositional tendency to experience intense anger specifically in response to situations involving perceived slights, frustration, unfair evaluation, or negative interpersonal criticism (e.g., “I get angry when I am slowed down by others’ mistakes”, “It makes me furious when I am criticized in front of others”).

3. Anger Expression-Out (AX-O)

The 8-item Anger Expression-Out scale gauges the frequency with which an individual directs angry feelings outward toward other people or objects through aggressive physical or verbal behaviors. High scorers display low behavioral inhibition, manifested through screaming, slamming doors, using insulting or sarcastic language, arguing, or physical destruction (e.g., “I slam doors”, “I say nasty things”).

4. Anger Expression-In (AX-I)

The 8-item Anger Expression-In scale assesses the frequency with which angry feelings are suppressed, channeled inward, or held back from interpersonal disclosure. Individuals with elevated AX-I scores simmer internally, experience subjective rumination, harbor lingering grudges, and withdraw interpersonally without resolving the provoking event (e.g., “I keep things in”, “I boil inside, but I don’t show it”).

5. Anger Control-Out (AC-O)

The 8-item Anger Control-Out scale measures the degree of intentional effort and self-regulatory energy an individual devotes to preventing the external manifestation of anger toward people or objects. It assesses conscious behavioral modification strategies, such as monitoring verbal tone, practicing patience, keeping one’s cool, and actively resisting violent or insulting impulses (e.g., “I control my temper”, “I can stop myself from losing my temper”).

6. Anger Control-In (AC-I)

The 8-item Anger Control-In scale assesses how often an individual invests psychological energy into calming down, soothing internal physiological arousal, and adopting constructive cognitive attitudes to dissipate anger internally. Techniques measured include relaxation, conscious cognitive reappraisal, and psychological acceptance (e.g., “I take a deep breath and relax”, “I try to simmer down”, “I try to accept the situation”).

7. Anger Expression Index (AX Index)

The AX Index is a cumulative metric summarizing overall anger expression patterns. Calculated via the formula AX Index = (AX-O + AX-I) - (AC-O + AC-I) + 48, it yields a unified metric indicating whether a person’s anger management is skewed toward pathological expression/suppression or healthy emotional containment and modulation.

Theoretical Framework

The STAXI-2 is embedded in Charles Spielberger’s overarching State-Trait Theory of Emotion, an integrative paradigm initially formulated to differentiate acute situational anxiety from personality-based neurotic vulnerabilities. Spielberger extended this architecture to the anger-hostility-aggression nexus, responding to pervasive historical ambiguities wherein the three constructs were frequently conflated.

The State-Trait Distinction

Spielberger maintained that emotions must be parsed into two distinct conceptual levels: transient states and enduring traits. State anger ($S_{Anger}$) represents an experiential, time-limited event triggered by internal or external stimuli, manifesting as sympathetic arousal (tachycardia, elevated galvanic skin response, heightened systolic blood pressure) alongside cognitive interpretations of threat or insult. In contrast, Trait anger ($T_{Anger}$) denotes an individual difference construct—a latent personality predisposition that dictates an individual’s threshold for state anger elicitation. Individuals with elevated trait anger exhibit cognitive schemas characterized by low frustration tolerance, selective attention to hostile intent, and hyper-reactivity, resulting in more frequent, prolonged, and intense state anger spikes.

The STAXI-2 framework also incorporates elements of Leonard Berkowitz’s Cognitive-Neoassociation Theory of aggression. Berkowitz posited that aversive events automatically activate negative affect, which rapidly primes both “fight” (anger, approach, attack) and “flight” (fear, avoidance) cognitive-motor networks. Spielberger’s paradigm maps how this initial undifferentiated affective flash transforms into behavioral trajectories: individuals can intentionally modulate the experience via self-regulatory feedback loops (Anger Control) or surrender to immediate motor scripts (Anger Expression).

Cardiovascular Reactivity and Psychosomatic Pathways

A second theoretical pillar undergirding the development of the STAXI-2 is the Psychophysiological Reactivity Hypothesis, rooted in the classic Cannon-Bard model of fight-or-flight and modified by modern behavioral medicine. Early cardiovascular literature identified the Type A behavior pattern as a predictor of myocardial infarction, but subsequent meta-analyses demonstrated that broad Type A characteristics were inconsistent predictors; instead, the specific affective-behavioral component of cynical hostility and anger expression drove pathogenic outcomes.

Spielberger recognized that neither high state nor high trait anger alone accounted for somatic morbidity. Rather, the pathway to somatic pathology depends on expression style: chronic internal suppression (high AX-I, low AC-I) induces prolonged vascular resistance, elevated peripheral vascular tone, and hypercortisolemia without the cathartic release of motor action. Conversely, high outward aggression (high AX-O) causes repeated hemodynamic surges and shear stress on coronary endothelium. Thus, the STAXI-2’s four-factor expression model (AX-O, AX-I, AC-O, AC-I) provides a complete matrix for testing psychosomatic hypotheses.

Validity

The construct, convergent, discriminant, and predictive validity of the STAXI-2 have been thoroughly verified across multinational clinical, forensic, and non-clinical cohorts.

Construct and Factorial Validity

Construct validity is evidenced by the consistent differentiation between emotional experience (State/Trait) and anger regulation (Expression/Control). Factor analytic examinations have repeatedly confirmed that the scales represent empirically distinct yet correlated latent dimensions. For example, Lievaart et al. (2016) demonstrated that confirmatory factor analysis (CFA) across psychiatric outpatients and healthy controls confirmed the theoretical 6-factor structure of the expression/control domain and the multi-subscale architecture of the state and trait dimensions.

Convergent and Discriminant Validity

Convergent validity is robustly supported through correlations with established personality and psychopathology scales:

• Buss-Perry Aggression Questionnaire (BPAQ): The STAXI-2 AX-O scale exhibits strong positive correlations ($r = .65$ to $.78$) with the BPAQ Physical and Verbal Aggression subscales, whereas Trait Anger correlates strongly ($r = .70$ to $.81$) with the BPAQ Hostility and Anger scales.
• Cook-Medley Hostility Scale (Ho): Trait Anger and Angry Reaction correlate significantly ($r = .52$ to $.68$) with the Ho scale, confirming its sensitivity to cynical and antagonistic cognitive styles.
• State-Trait Anxiety Inventory (STAI): The State and Trait Anger scales display moderate correlations ($r = .35$ to $.48$) with the STAI, capturing shared negative affectivity while preserving discriminant uniqueness (anger specifically drives approach-oriented action tendencies, whereas anxiety triggers avoidance).
• Beck Depression Inventory-II (BDI-II): Anger Expression-In (AX-I) demonstrates specific, heightened correlations ($r = .45$ to $.60$) with depressive symptomatology and ruminative tendencies, supporting the psychoanalytic and cognitive premise that internalizing anger contributes to depressive etiology.

Predictive and Clinical Validity

In medical settings, STAXI-2 profiles prospectively predict cardiovascular morbidity. Individuals presenting with elevated AX-I scores coupled with low AC-I exhibit significantly higher ambulatory systolic blood pressure and higher rates of post-angioplasty restenosis. In forensic settings, AX-O and the State Anger Physical subscale (S-Ang/P) successfully differentiate violent from non-violent offenders, demonstrating significant predictive utility in violent recidivism modeling.

Reliability

The STAXI-2 possesses exemplary reliability indices across various linguistic adaptations, cultures, and populations.

Internal Consistency

Extensive normative studies published by Spielberger (1999), encompassing over 1,600 healthy adults, university undergraduates, and adolescent cohorts, established high Cronbach’s alpha internal consistency coefficients across all scales:

• State Anger (S-Ang): $\alpha = .91$ to $.95$; Subscales: S-Ang/F ($\alpha = .88$), S-Ang/V ($\alpha = .86$), S-Ang/P ($\alpha = .84$).
• Trait Anger (T-Ang): $\alpha = .84$ to $.89$; Subscales: T-Ang/T ($\alpha = .84$ to $.87$), T-Ang/R ($\alpha = .74$ to $.78$).
• Anger Expression-Out (AX-O): $\alpha = .73$ to $.81$.
• Anger Expression-In (AX-I): $\alpha = .74$ to $.80$.
• Anger Control-Out (AC-O): $\alpha = .79$ to $.86$.
• Anger Control-In (AC-I): $\alpha = .82$ to $.88$.
• Anger Expression Index (AX Index): Composite internal consistency typically surpasses $.85$.

Test-Retest Reliability

Test-retest stability coefficients reflect the theoretical state-trait dichotomy. Over intervals spanning from several weeks to several months:

• State Anger: Exhibits intentionally low to modest stability ($r_{tt} = .18$ to $.35$), as state emotions naturally fluctuate based on current situational stimuli and environmental demands.
• Trait Anger & Expression Scales: Demonstrate substantial stability across 2- to 12-week test-retest intervals ($r_{tt} = .70$ to $.84$), confirming that dispositional temperament, internal suppression styles, and control mechanisms operate as stable, enduring personality attributes over time.

Factor Analysis

The structural validity of the STAXI-2 has been validated through rigorous exploratory factor analyses (EFA) and structural equation modeling using confirmatory factor analyses (CFA).

Part 1: State Anger Factor Structure

During scale development, principal axis factoring with promax (oblique) and varimax (orthogonal) rotations on the 15 State Anger items consistently revealed three robust first-order factors, accounting for over 60% of the total variance:

1. Feeling Angry: High loadings ($lambda = .65$ to $.86$) for items 1, 2, 3, 6, 9, and 11.
2. Feel Like Expressing Anger Verbally: Primary loadings ($lambda = .61$ to $.82$) for items 4, 10, 13, and 14.
3. Feel Like Expressing Anger Physically: High loadings ($lambda = .64$ to $.85$) for items 5, 7, 8, 12, and 15.

Second-order hierarchical CFA demonstrates that these three sub-factors load onto a singular higher-order State Anger latent factor, with good fit indices: $chi^2/df < 2.5$, Root Mean Square Error of Approximation ($text{RMSEA}) = .048$ ($90% \text{ CI } [.041, .055]$), Comparative Fit Index ($\text{CFI}) = .97$, and Tucker-Lewis Index ($\text{TLI}) = .96$.

Part 2: Trait Anger Factor Structure

Factor analyses of the 10 Trait Anger items yield two correlated dimensions ($r \approx .45$ to $.60$):

• Angry Temperament: Items 16, 17, 18, and 19 load strongly ($lambda > .70$) on this factor, reflecting generalized emotional reactivity without explicit external triggers.
• Angry Reaction: Items 20, 21, 22, and 24 capture stimulus-dependent irritation (e.g., frustration with delays, negative evaluations), with loadings between $.60$ and $.78$.

Part 3: Anger Expression and Control Structure

The 32 items evaluating expression and control (Items 26–57) were designed to evaluate a four-factor orthogonal or oblique model. In extensive CFA investigations (e.g., Forgays et al., 1997; Spielberger, 1999), a four-factor solution consistently outperforms one-, two-, or three-factor alternatives:

• Anger Control-Out (AC-O): Items 26, 30, 36, 39, 42, 45, 51, and 56 load cleanly ($lambda = .55$ to $.76$).
• Anger Control-In (AC-I): Items 48, 49, 52, 53, 54, and 57 display substantial loadings on internal cognitive/physiological regulation.
• Anger Expression-Out (AX-O): Items 27, 29, 31, 33, 35, 38, 41, and 43 load robustly onto overt aggressive output.
• Anger Expression-In (AX-I): Items 28, 32, 34, 37, 40, 44, 47, and 55 define a dedicated latent factor of internal anger containment and social withdrawal.

Multigroup invariance testing indicates structural invariance (configural, metric, and scalar) across biological sex, age brackets (adolescents vs. adults), and language translations (e.g., German, Dutch, Italian, Spanish), demonstrating that the STAXI-2’s underlying construct architecture is cross-culturally stable.

Instrument / Measurement Tool

• Full Instrument Name: State-Trait Anger Expression Inventory-2
• Acronym: STAXI-2
• Author: Charles D. Spielberger, Ph.D.
• Publisher: Psychological Assessment Resources, Inc. (PAR)
• Format / Administration: Self-report questionnaire, available in paper-and-pencil, computer-administered, and online tele-assessment formats.
• Item Count: 57 items in total, organized into three distinct operational sections:
• Part 1 (State Anger): 15 items (Items 1–15) assessing current emotional status.
• Part 2 (Trait Anger): 10 items (Items 16–25) assessing dispositional emotional style.
• Part 3 (Anger Expression & Control): 32 items (Items 26–57) evaluating expression and self-regulatory behaviors.
• Scales and Subscales (6 Scales, 5 Subscales):
  • State Anger (S-Ang)
    • Feeling Angry (S-Ang/F) [Items 1, 2, 3, 6, 9]
    • Feel Like Expressing Anger Verbally (S-Ang/V) [Items 4, 10, 11, 13, 14]
    • Feel Like Expressing Anger Physically (S-Ang/P) [Items 5, 7, 8, 12, 15]
  • Trait Anger (T-Ang)
    • Angry Temperament (T-Ang/T) [Items 16, 17, 18, 19]
    • Angry Reaction (T-Ang/R) [Items 20, 21, 22, 24]
  • Anger Expression-Out (AX-O) [8 items: e.g., 27, 29, 31, 33, 35, 38, 43, 46]
  • Anger Expression-In (AX-I) [8 items: e.g., 28, 32, 34, 37, 40, 44, 47, 55]
  • Anger Control-Out (AC-O) [8 items: e.g., 26, 30, 36, 39, 42, 45, 51, 56]
  • Anger Control-In (AC-I) [8 items: e.g., 48, 49, 50, 52, 53, 54, 57]
  • Anger Expression Index (AX Index) [Composite index]
• Authentic Response Scale:
  • Part 1 (State Anger): 4-point Likert scale (1 = Not at all, 2 = Somewhat, 3 = Moderately so, 4 = Very much so).
  • Part 2 (Trait Anger) & Part 3 (Anger Expression and Control): 4-point Likert scale (1 = Almost never, 2 = Sometimes, 3 = Often, 4 = Almost always).
• Scoring and Transformation:
  • Raw scale scores are computed by summing the item response values (1 through 4) designated for each respective scale and subscale.
  • The Anger Expression Index (AX Index) is calculated using the algebraic formula: AX Index = (AX-O + AX-I) - (AC-O + AC-I) + 48.
  • Raw scores for all scales and subscales are converted to normative percentile ranks and normalized standardized T-scores ($M = 50, SD = 10$) using demographic tables stratified by gender and age (adolescents aged 16–19, adults aged 20–29, 30–39, 40–49, 50–59, and 60+).
• Completion Time: Approximately 12 to 15 minutes.

Permissions & Fee and Test Year

The State-Trait Anger Expression Inventory-2 was officially published in 1999 by Psychological Assessment Resources, Inc. (PAR). The STAXI-2 is a proprietary, copyrighted psychological assessment instrument. It is not in the public domain and cannot be reproduced, administered, or integrated into digital survey systems without obtaining licensing permissions or purchasing standardized test forms from the publisher.

To acquire the STAXI-2 manual, item booklets, reusable test sheets, or software scoring keys, researchers and practitioners must hold a minimum Qualification Level B (requiring master’s level training in psychology, testing, or related behavioral disciplines). Further purchasing details, translation licenses, and commercial usage fees can be accessed directly at Psychological Assessment Resources (PAR).

References

• Berkowitz, L. (1990). On the formation and regulation of anger and aggression: A cognitive-neoassociationistic analysis. American Psychologist, 45(4), 494–503. https://doi.org/10.1037/0003-066X.45.4.494
• Forgays, D. G., Forgays, D. K., & Spielberger, C. D. (1997). Factor structure of the State-Trait Anger Expression Inventory. Journal of Personality Assessment, 69(3), 497–507. https://doi.org/10.1207/s15327752jpa6903_5
• Lievaart, M., Franken, I. H. A., & Hovens, J. E. (2016). Anger assessment in clinical and nonclinical populations: Further validation of the State-Trait Anger Expression Inventory-2. Journal of Personality Assessment, 98(4), 437–447. https://doi.org/10.1080/00223891.2015.1118389
• Spielberger, C. D. (1988). State-Trait Anger Expression Inventory: Professional manual. Psychological Assessment Resources.
• Spielberger, C. D. (1999). STAXI-2: State-Trait Anger Expression Inventory-2. Professional manual. Psychological Assessment Resources.
• Spielberger, C. D., Krasner, S. S., & Solomon, E. P. (1988). The experience, expression, and control of anger. In M. P. Janisse (Ed.), Health Psychology: Individual Differences and Stress (pp. 89–108). Springer-Verlag. https://doi.org/10.1007/978-1-4612-3824-9_5
• Spielberger, C. D., & Sydeman, S. J. (1994). State-Trait Anxiety Inventory and State-Trait Anger Expression Inventory. In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcome assessment (pp. 292–321). Lawrence Erlbaum Associates.

Items of the Scale