Aggression Questionnaire (AQ)

1. Abstract

The Aggression Questionnaire (AQ), developed by Arnold H. Buss and W. L. Warren (2000), represents a seminal psychometric instrument engineered to evaluate multifaceted dimensions of aggressive tendencies across the lifespan. Evolving directly from the foundational Buss-Durkee Hostility Inventory (BDHI; Buss & Durkee, 1957) and the widely cited Buss-Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992), the Buss and Warren standard commercial revision comprises 34 self-report items mapped across five robust clinical and research dimensions: Physical Aggression, Verbal Aggression, Anger, Hostility, and Indirect Aggression. Formulated at a third-grade reading level to ensure developmental and educational accessibility, the AQ employs a 5-point Likert response scale ranging from 1 (“Not at all like me”) to 5 (“Completely like me”). The standardized instrument yields individual subscale raw scores, a Total Aggression composite score, and a specialized Inconsistent Responding Index designed to identify response sets, malingering, or random endorsement patterns.

Standardized on a normative sample of 2,138 individuals spanning ages 9 to 88 years, the AQ provides developmentally stratified normative profiles across three primary age brackets: 9–18 years (children and adolescents), 19–39 years (young adults), and 40–88 years (middle to older adults), with sex-stratified normative adjustments for the Physical and Verbal Aggression dimensions. Extensive psychometric evaluations demonstrate high internal consistency (Cronbach’s alpha coefficients typically ranging between α = .72 and .88 for subscales, and exceeding .90 for the Total Score) alongside robust test-retest reliability across multi-week intervals (r = .80 to .90). Confirmatory factor analytic investigations widely support the five-factor structural architecture, confirming measurement invariance across clinical, correctional, and community cohorts. The questionnaire successfully predicts clinically significant violent behavior, domestic assault recidivism, institutional infractions, and psychiatric symptom exacerbation, making it a gold standard in clinical psychology, forensic evaluation, correctional intake, psychiatric epidemiology, and social-cognitive personality assessment.

2. Keywords

Aggression Questionnaire, Buss and Warren, Buss-Perry Aggression Questionnaire, Physical Aggression, Verbal Aggression, Anger, Hostility, Indirect Aggression, Psychometrics, Factor Analysis, Forensic Psychology, Measurement Invariance, Violent Recidivism, Self-Report Inventory, Impulsive Aggression.

3. Authors

The Aggression Questionnaire (AQ; 2000 revision) was developed and standardized by Arnold H. Buss, Ph.D., and W. L. Warren, Ph.D., and published by Western Psychological Services (WPS). Dr. Arnold H. Buss (1924–2021) was an internationally distinguished professor of psychology at the University of Texas at Austin. Renowned for his foundational contributions to behavioral genetics, temperament theory, personality psychology, and social evolution, Buss pioneered modern aggression measurement through the landmark publication of the Buss-Durkee Hostility Inventory with Ann Durkee in 1957. Later, in collaboration with Mark Perry, he refined this construct into the 29-item Buss-Perry Aggression Questionnaire in 1992. Dr. W. L. Warren collaborated with Buss under the auspices of Western Psychological Services to formalize extensive normative stratification, optimize reading comprehension levels, integrate an Indirect Aggression subscale, and construct valid validity indices, yielding the standardized 34-item commercial edition in 2000.

Their research was conducted across extensive clinical, school, correctional, and community settings across the United States. Inquiries regarding official test materials, commercial licensing, manual distribution, computerized scoring systems, and standardized normative tables are managed by Western Psychological Services (625 Alaska Avenue, Torrance, CA 90503, USA).

4. Purpose

The primary purpose of the Aggression Questionnaire (AQ) is to provide an objective, empirically validated, and developmentally comprehensive psychometric measurement of aggressive behaviors, affective arousal, and cognitive hostility in children, adolescents, and adults. Aggression represents an extraordinarily complex, transdiagnostic behavioral construct implicated in numerous clinical diagnoses outlined in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR), such as Conduct Disorder, Oppositional Defiant Disorder, Intermittent Explosive Disorder, Antisocial Personality Disorder, and Borderline Personality Disorder. Prior to the formalization of multidimensional models, clinical assessment frequently conflated the behavioral execution of physical violence with subjective emotional anger and enduring cognitive cynicism. The AQ overcomes this clinical limitation by cleanly demarcating distinct behavioral, affective, and cognitive components.

In clinical practice, the AQ is widely utilized to establish baseline aggressive symptom severity, design targeted cognitive-behavioral interventions (such as anger management or dialectical behavior therapy skills training), and track therapeutic outcomes longitudinally. For instance, individuals presenting with elevated Anger subscale scores alongside normative Physical Aggression profiles require clinical strategies focused primarily on autonomic downregulation and affective tolerance (e.g., physiological calming, progressive relaxation). Conversely, clients exhibiting high Hostility scores alongside elevated Indirect or Physical Aggression necessitate cognitive restructuring targeting persecutory attributional biases, social-cognitive schemas, and instrumental problem-solving deficiencies.

In forensic and correctional settings, the AQ serves as a rapid, empirically grounded intake screening tool. Correctional psychologists utilize the instrument to estimate institutional management risk, evaluate readiness for parole or community reintegration, identify candidates for specialized domestic violence intervention programs (e.g., Batterer Intervention Programs), and document post-treatment behavioral change. The inclusion of the Inconsistent Responding Index offers forensic evaluators an essential mechanism to detect uncritical acquiescence, extreme defensiveness, random responding, or reading comprehension failure among court-mandated individuals. Furthermore, the questionnaire is extensively deployed in psychiatric epidemiology, developmental psychology, organizational psychology (e.g., assessing workplace deviance and counterproductive work behavior), and sports psychology to evaluate individual differences in reactive versus proactive aggressive reactivity.

5. Psychological Construct

The Aggression Questionnaire operationalizes aggression as a five-dimensional construct encompassing overt behavioral acts, covert and relational expressions, emotional activation, and enduring cognitive belief systems. Rather than viewing aggression as a unitary drive, Buss and Warren (2000) conceptualized aggressive behavior as a multifaceted output governed by distinct psychological systems:

5.1. Physical Aggression (Behavioral Component)

Physical Aggression reflects the overt, motoric behavioral inclination to inflict bodily harm, injury, or physical pain upon other individuals or to utilize physical force to destroy objects or coerce others. This dimension captures instrumental and reactive physical behaviors, ranging from direct physical altercations, hitting, and brawling to threatening posturing and weapon utilization. Items tapping this dimension evaluate historical engagement in violence, loss of physical behavioral control (e.g., Item 23: “At times I can’t control the urge to hit someone”), and retaliatory violent responses (e.g., Item 25: “If somebody hits me, I hit back”). Elevated scores on this subscale serve as one of the strongest psychometric predictors of interpersonal assault, domestic abuse, violent crime, and institutional violence in correctional facilities.

5.2. Verbal Aggression (Expressive/Communicative Component)

Verbal Aggression represents the expressive, communicative dimension of aggression characterized by the delivery of noxious verbal stimuli, argumentative defiance, verbal hostility, and verbal threats. Unlike physical violence, verbal aggression manifests through language, voice tone, and explicit argumentative conflict. The subscale evaluates habitual argumentative patterns (e.g., Item 1: “My friends say that I argue a lot”), uncontrollable debate tendencies (e.g., Item 6: “I can’t help getting into arguments when people disagree with me”), and candid, abrasive confrontational habits (e.g., Item 26: “I tell my friends openly when I disagree with them”). While socially less physically devastating than overt violence, high verbal aggression profoundly disrupts occupational functioning, marital stability, and interpersonal relationships.

5.3. Anger (Affective/Physiological Component)

The Anger dimension captures the emotional, physiological, and autonomic arousal associated with aggressive behavior. Anger serves as the psychological bridge connecting cognitive evaluations of provocation with motoric behavioral aggression. This subscale measures physiological activation, rapid arousal onset, autonomic lability, and subjective difficulty regulating affective temperament. Items evaluate rapid temper escalation (e.g., Item 3: “I flare up quickly, but get over it quickly”), unprovoked emotional outbursts (e.g., Item 7: “At times I get very angry for no good reason”), subjective loss of emotional equilibrium (e.g., Item 12: “I have trouble controlling my temper”), and extreme subjective pressure (e.g., Item 29: “At times I feel like a bomb ready to explode”). Elevation on the Anger subscale is intimately linked to cardiovascular reactivity, impulse control pathology, and affective disorders.

5.4. Hostility (Cognitive/Evaluative Component)

Hostility represents the cognitive, attitudinal, and evaluative component of the aggressive construct. It consists of enduring feelings of ill will, bitter resentment, pervasive cynicism, suspiciousness, paranoia, and perceived injustice. Unlike anger, which is an episodic, intense affective state, hostility constitutes a stable, trait-like cognitive schema through which social information is perceived. Items assess perceived victimization (e.g., Item 2: “Other people always seem to get the breaks”; Item 5: “At times I feel I have gotten a raw deal out of life”), chronic interpersonal bitterness (e.g., Item 9: “I wonder why sometimes I feel so bitter about things”), and paranoid attributional styles (e.g., Item 16: “I wonder what people want when they are nice to me”; Item 28: “I do not trust strangers who are too friendly”). High hostility operates as a potent cognitive vulnerability that predisposes individuals to interpret ambiguous social cues as intentionally hostile, fueling reactive violence.

5.5. Indirect Aggression (Relational/Covert Component)

Indirect Aggression captures covert, circuitous, or passive aggressive behaviors designed to inflict harm on another individual without direct face-to-face physical or verbal confrontation. This subscale incorporates passive-aggressive resistance, social and relational aggression (e.g., Item 18: “I sometimes spread gossip about people I don’t like”), malicious damage to an adversary’s property or occupational output (e.g., Item 13: “If I’m angry enough, I may mess up someone’s work”), intentional emotional withdrawal (e.g., Item 30: “When someone really irritates me, I might give him or her silent treatment”), and deliberate defiance of authority through intentional inefficiency (e.g., Item 15: “When people are bossy, I take my time doing what they want, just to show them”). The addition of this subscale in the Buss and Warren standardization addressed critical developmental and clinical findings showing that relational and covert behaviors are prominent modalities of interpersonal aggression across sexes and across organizational environments.

6. Theoretical Framework

The theoretical architecture underpinning the Aggression Questionnaire originates from multiple converging psychological paradigms, including behavioral genetics, social cognitive theory, the Frustration-Aggression Hypothesis, and modern integrative models such as the General Aggression Model (GAM) formulated by Anderson and Bushman (2002).

6.1. Historical Lineage: From BDHI to BPAQ and the AQ

The genesis of the AQ traces to Arnold Buss and Ann Durkee’s (1957) seminal assertion that aggression cannot be operationalized as an undifferentiated, global trait. Buss and Durkee designed the Buss-Durkee Hostility Inventory (BDHI), comprising 75 true/false items mapped onto seven putative subclasses (Assault, Indirect Aggression, Irritability, Negativism, Resentment, Suspicion, and Verbal Aggression). While the BDHI dominated psychometric aggression research for three decades, it faced severe methodological critiques regarding dichotomous response constraints, confounding factor structures, and psychometrically unstable item loadings. In 1992, Buss and Perry conducted comprehensive exploratory and confirmatory factor analyses on a revised pool of items, using 5-point Likert scales, which yielded the 29-item Buss-Perry Aggression Questionnaire (BPAQ). The BPAQ established the classic four-factor structural taxonomy: Physical Aggression, Verbal Aggression, Anger, and Hostility.

Despite its widespread adoption, the 1992 BPAQ possessed notable limitations: it lacked contemporary normative stratification across diverse age groups, omitted an explicit scale for covert/indirect behaviors (which had been lost during the BDHI-to-BPAQ reduction), and lacked validation for low-literacy or clinical populations. Buss and Warren (2000) resolved these limitations by recalibrating items to a third-grade reading level, re-introducing a psychometrically refined Indirect Aggression subscale, embedding an Inconsistent Responding Index, and executing a nationwide, stratified standardization across ages 9 to 88, formally creating the modern 34-item Aggression Questionnaire.

6.2. The General Aggression Model (GAM) and Cognitive Neoassociationism

The five dimensions of the AQ interface with the theoretical stages of Leonard Berkowitz’s (1989) Cognitive Neoassociation Theory and Anderson and Bushman’s (2002) General Aggression Model. According to the GAM, an aggressive episode is initiated by the interaction of person inputs (e.g., biological predispositions, hostile personality traits) and situation inputs (e.g., interpersonal provocation, frustration). These inputs impact the individual’s internal state through three interconnected routes: affect (emotional anger), cognition (hostile thoughts, schemas, and attributions), and physiological arousal.

Within this framework, the AQ subscales represent distinct components of the internal appraisal and behavioral execution matrix:

• Cognitive Routes: The Hostility subscale evaluates deeply entrenched semantic associative networks containing hostile attributional biases, beliefs about human malevolence, and feelings of cynicism.
• Affective Routes: The Anger subscale reflects physiological arousal, visceral emotional reactivity, and autonomic activation mediated by the limbic system (notably the amygdala and anterior insular cortex).
• Appraisal & Behavioral Manifestations: The behavioral outcomes of these cognitive-affective states are manifested either overtly through the motor systems tapped by Physical Aggression and Verbal Aggression or covertly through calculated strategies assessed by Indirect Aggression.

6.3. Reactive versus Proactive Aggression Paradigms

The AQ also aligns with developmental taxonomies distinguishing reactive (impulsive, affective, provocation-driven) from proactive (instrumental, cold-blooded, goal-directed) aggression (Dodge & Coie, 1987). Reactive aggression is characterized by high scores on the Anger and Hostility subscales, reflecting intense autonomic arousal and emotional dysregulation in response to perceived slights. In contrast, purely proactive aggression may present with elevated Physical and Indirect Aggression scores accompanied by normative or suppressed Anger scores, capturing instrumental manipulation without emotional distress.

7. Validity

The Aggression Questionnaire has undergone psychometric validation across thousands of research studies globally, confirming its construct, convergent, discriminant, and criterion-related validity across clinical, general, and forensic populations.

7.1. Convergent Validity

Extensive convergent validity analyses demonstrate statistically significant correlations between AQ subscales and other well-validated measures of psychological aggression, anger, and impulsivity. Buss and Warren (2000), as well as independent validation researchers, reported strong positive correlations between the AQ Total Score and the State-Trait Anger Expression Inventory-2 (STAXI-2; Spielberger, 1999). Specifically, the AQ Anger subscale correlates strongly with the STAXI-2 Trait Anger scale (r values typically ranging between .70 and .82), demonstrating that the affective dimension of the AQ reliably measures trait emotional arousal.

In forensic and intimate partner violence populations, such as the investigation by García-León et al. (2002) and subsequent studies of partner-assaultive men, AQ Physical Aggression scores demonstrated robust positive correlations with the Conflict Tactics Scale-2 (CTS-2; Straus et al., 1996) Physical Assault scale (r = .52 to .68). The Hostility subscale correlates significantly with the Cook-Medley Hostility Scale (r = .64) and measures of generalized paranoia and external locus of control, while the Verbal Aggression subscale correlates positively with scales evaluating assertiveness, argumentativeness, and interpersonal dominance.

7.2. Discriminant Validity

Discriminant validity is supported by modest or near-zero correlations between the AQ and unrelated constructs, including social desirability, intelligence, and somatic anxiety. Research utilizing the Marlowe-Crowne Social Desirability Scale (MCSDS) demonstrates low negative correlations with AQ subscales (ranging from r = -.12 to -.28), confirming that while aggressive disclosure is modestly influenced by social desirability, the scale successfully extracts authentic variance independent of impression management. Furthermore, the distinctiveness of the subscales is confirmed by divergent clinical patterns: whereas the Hostility subscale correlates moderately with depressive symptomatology on the Beck Depression Inventory (BDI-II) due to shared elements of dysphoria and perceived injustice (r = .40 to .48), the Physical Aggression subscale exhibits negligible correlations with depression (r = .06 to .14), demonstrating clear separation between cognitive distress and violent behavior.

7.3. Criterion and Predictive Validity

The AQ exhibits exceptional criterion-related validity in distinguishing known aggressive and non-aggressive cohorts. In clinical and forensic validation studies, individuals incarcerated for violent offenses score significantly higher on the Physical Aggression, Anger, and Total Aggression scales than non-violent offenders and matched community controls (Cohen’s d effect sizes ranging from 0.75 to 1.35). In institutional correctional settings, baseline AQ Physical Aggression scores reliably predict future disciplinary infractions, weapon possession, and verbal assaults on staff over 6- and 12-month follow-up periods. In psychiatric settings, elevated Anger and Hostility scores at admission accurately predict acute violent episodes and medication non-compliance.

8. Reliability

The Aggression Questionnaire exhibits high internal consistency, standard error of measurement (SEM) precision, and temporal stability across diverse clinical and non-clinical cohorts.

8.1. Internal Consistency

In the nationwide standardization sample of 2,138 individuals (Buss & Warren, 2000), the AQ demonstrated robust internal consistency across all age brackets and sexes. The Cronbach’s alpha coefficients reported across the normative sample and verified across independent validation studies are summarized below:

• Physical Aggression (8 items): α = .82 to .88
• Verbal Aggression (5 items): α = .72 to .80
• Anger (7 items): α = .78 to .85
• Hostility (8 items): α = .80 to .86
• Indirect Aggression (6 items): α = .71 to .79
• Total Aggression Score (34 items): α = .90 to .94

Across special populations, including partner-assaultive men, adolescent psychiatric inpatients, and military recruits, the composite reliability of the Total Score consistently exceeds α = .90, satisfying the rigorous psychometric threshold (α ≥ .90) mandated for high-stakes individual clinical decision-making.

8.2. Test-Retest Reliability and Temporal Stability

Test-retest stability was evaluated across various time intervals ranging from two weeks to several months. In the original standardization sample, a subset of 115 participants retested over a 2-week interval yielded exceptional stability coefficients: Physical Aggression (r = .86), Verbal Aggression (r = .78), Anger (r = .80), Hostility (r = .82), Indirect Aggression (r = .76), and Total Aggression (r = .88). Over a 9-week retest interval among young adults, stability coefficients remained elevated (ranging from r = .72 for Verbal Aggression to r = .84 for Physical Aggression), confirming that the AQ assesses stable, trait-like characteristics rather than ephemeral mood states.

9. Factor Analysis

The underlying dimensionality of the Aggression Questionnaire has been scrutinized through exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) across numerous linguistic adaptations and demographic cohorts.

9.1. Exploratory and Confirmatory Factor Structures

Historically, Buss and Perry’s (1992) principal component analyses extracted four correlated factors (Physical Aggression, Verbal Aggression, Anger, Hostility), which formed the core of the BPAQ. However, Buss and Warren (2000) incorporated items assessing indirect, passive, and relational aggression, deriving a five-factor solution. When evaluated via Confirmatory Factor Analysis (CFA), the hypothesized five-factor model demonstrates superior model fit compared to unifactorial, two-factor (behavioral vs. cognitive/affective), or traditional four-factor solutions.

Across diverse psychometric investigations (e.g., Tremblay & Ewart, 2005; García-León et al., 2002), the five-factor structural architecture consistently exhibits acceptable to excellent goodness-of-fit indices:

• Comparative Fit Index (CFI): .91 to .96
• Tucker-Lewis Index (TLI): .90 to .95
• Root Mean Square Error of Approximation (RMSEA): .042 to .058 (with 90% confidence intervals rarely exceeding .065)
• Standardized Root Mean Square Residual (SRMR): .045 to .055

Alternative structural evaluations, such as bifactor modeling, have demonstrated that while a broad general “Aggression” factor accounts for substantial shared variance across items (supporting the use of the Total Aggression Score), the five specific group factors retain significant, unique, and reliable variance that warrants individual subscale interpretation in clinical and diagnostic settings.

9.2. Factor Loadings and Measurement Invariance

Standardized factor loadings across the 34 items generally range from λ = .42 to .81, with the vast majority loading above λ = .55 onto their respective target constructs. Cross-loadings are minimal, although slight non-pathological covariance frequently emerges between items on the Anger and Physical Aggression dimensions (reflecting the affective escalation preceding physical acts) and between Anger and Hostility (reflecting the interaction of affect and cognition).

Multigroup confirmatory factor analyses (MGCFA) have rigorously established configural, metric (weak), and scalar (strong) measurement invariance across sex and age groups. Metric invariance confirms that the factor loadings are equivalent across males and females, confirming that the questionnaire measures the exact same underlying constructs in both sexes. Scalar invariance permits meaningful cross-group latent mean comparisons, which reveal consistent developmental patterns: physical aggression and anger generally decline with advancing age across the adult lifespan, whereas hostility and indirect aggression demonstrate greater longitudinal stability.

10. Instrument / Measurement Tool

• Official Title: Aggression Questionnaire (AQ)
• Authors: Arnold H. Buss, Ph.D., and W. L. Warren, Ph.D.
• Publication Year: 2000 (standardized commercial manual and norms; foundational versions in 1957 [BDHI] and 1992 [BPAQ])
• Publisher: Western Psychological Services (WPS)
• Format: Paper-and-pencil self-report inventory, AutoScore answer form, or computerized administration via the AQ CD / WPS Online Evaluation System
• Target Population: Children, adolescents, and adults aged 9 to 88 years
• Reading Level: Third-grade reading level (Flesch-Kincaid grade level ≤ 3.5)
• Administration Time: Approximately 10 minutes for the full 34-item form; 3 to 5 minutes for the 15-item short form
• Number of Items: 34 items (Full Scale); 15 items (Short Form)
• Response Scale: 5-point Likert-type scale scored from 1 to 5:
  • 1 = Not at all like me
  • 2 = A little like me
  • 3 = Somewhat like me
  • 4 = Very much like me
  • 5 = Completely like me
• Subscale Allocation (34 Items):
  • Physical Aggression (8 items): Items 8, 10, 11, 17, 23, 24, 25, 27
  • Verbal Aggression (5 items): Items 1, 4, 6, 20, 26
  • Anger (7 items): Items 3, 7, 12, 14, 19 [Reverse Scored], 22, 29
  • Hostility (8 items): Items 2, 5, 9, 16, 21, 28, 31, 33
  • Indirect Aggression (6 items): Items 13, 15, 18, 30, 32, 34
• Reverse-Scored Item: Item 19 (“I am a calm person.”) is reverse scored: 1 → 5, 2 → 4, 3 → 3, 4 → 2, 5 → 1.
• Scoring and Metrics:
  • Subscale raw scores are calculated by summing item responses within each scale.
  • Total Score: Calculated by summing all 34 item responses (theoretical range: 34 to 170).
  • Standardized Scores: Raw scores are converted to normalized T-scores (Mean = 50, Standard Deviation = 10) and percentile ranks based on the appropriate age- and sex-stratified normative table.
  • Inconsistent Responding Index (INC): Computed by calculating absolute difference scores between pairs of highly correlated items throughout the test; scores exceeding established clinical cutoffs indicate invalid, careless, or random responding.
• Short Form (15 Items): Administered via the first 15 items on the AQ AutoScore form, consisting of exactly three representative items per subscale (Total raw score range: 15 to 75).

11. Permissions & Fee and Test Year

The Aggression Questionnaire (AQ; 2000 revision) is a proprietary, copyrighted psychological assessment instrument owned and published exclusively by Western Psychological Services (WPS). Test kits, manual handbooks, AutoScore answer sheets, and computer scoring licenses require formal commercial acquisition through WPS (www.wpspublish.com).

The instrument is classified under Qualification Level B. Under this classification, the purchaser or administrator must hold a master’s degree in psychology, education, counseling, social work, or a related healthcare field, or have documented formal coursework in the administration, scoring, and clinical interpretation of standardized psychological assessment tools. Academic researchers wishing to utilize the AQ in non-commercial, empirical research must obtain written authorization, reproduction permissions, or research licenses directly from the WPS Rights and Permissions Department. Unlicensed public reproduction, mass digital dissemination, or unauthorized online hosting of copyrighted test forms violates international intellectual property law.

12. References

• Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53(1), 27–51. https://doi.org/10.1146/annurev.psych.53.100901.135231
• Berkowitz, L. (1989). Frustration-aggression hypothesis: Examination and reformulation. Psychological Bulletin, 106(1), 59–73. https://doi.org/10.1037/0033-2909.106.1.59
• Buss, A. H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility. Journal of Consulting Psychology, 21(4), 343–349. https://doi.org/10.1037/h0046900
• Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality and Social Psychology, 63(3), 452–459. https://doi.org/10.1037/0022-3514.63.3.452
• Buss, A. H., & Warren, W. L. (2000). Aggression Questionnaire: Manual. Los Angeles, CA: Western Psychological Services.
• Dodge, K. A., & Coie, J. D. (1987). Social-information-processing factors in reactive and proactive aggression in children’s peer groups. Journal of Personality and Social Psychology, 53(6), 1146–1158. https://doi.org/10.1037/0022-3514.53.6.1146
• Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustration and Aggression. New Haven, CT: Yale University Press. https://doi.org/10.1037/10022-000
• García-León, A., Reyes, G. A., Vila, J., Pérez, N., Hébert, M., & Sánchez-Barrera, M. B. (2002). The Aggression Questionnaire: A validation study in student samples. The Spanish Journal of Psychology, 5(1), 45–53. https://doi.org/10.1017/s1138741600005822
• Spielberger, C. D. (1999). State-Trait Anger Expression Inventory-2: Professional Manual. Odessa, FL: Psychological Assessment Resources.
• Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996). The Revised Conflict Tactics Scales (CTS2): Development and preliminary psychometric data. Journal of Family Issues, 17(3), 283–316. https://doi.org/10.1177/019251396017003001
• Tremblay, P. F., & Ewart, L. A. (2005). The Buss–Perry Aggression Questionnaire and its scales: Construct validity, invariance, and prediction of violence. European Journal of Personality, 19(6), 481–499. https://doi.org/10.1002/per.550

13. Items of the Scale