Buss-Perry Aggression Questionnaire (BPAQ)

1. Abstract

The Buss-Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992) represents the preeminent psychometric instrument for assessing trait aggression in contemporary social psychology, personality psychology, forensic assessment, and behavioral psychiatry. Constructed as a direct modernization and psychometric refinement of the classic Buss-Durkee Hostility Inventory (BDHI; Buss & Durkee, 1957), the BPAQ resolves longstanding structural and methodological limitations of its predecessor by replacing dichotomous true/false formats with a continuous five-point Likert scale and subjecting item pools to rigorous factor-analytic examination. Comprising 29 self-report items, the BPAQ is underpinned by a robust four-factor conceptual architecture that clearly delineates the motor/instrumental, emotional/affective, and cognitive components of aggressive human behavior. These dimensions manifest as four distinct subscales: Physical Aggression (9 items), Verbal Aggression (5 items), Anger (7 items), and Hostility (8 items).

Extensive psychometric investigations across diverse clinical, forensic, and non-clinical samples have demonstrated that the BPAQ exhibits commendable internal consistency, with Cronbach’s alpha coefficients typically ranging from .72 to .85 for the individual subscales and approximately .89 for the omnibus scale. Test-retest stability across intervals ranging from several weeks to months consistently yields coefficients between .72 and .80, indicating high longitudinal reliability. Construct, convergent, and discriminant validity have been repeatedly affirmed through empirical correlations with peer nominations, laboratory paradigms of physical retaliation, physiological indicators of autonomic arousal, and established personality frameworks such as the Five-Factor Model (specifically low Agreeableness and high Neuroticism) and Eysenck’s PEN model. The BPAQ has achieved universal adoption in clinical evaluations, violence risk appraisals, sports psychology, and experimental investigations examining the behavioral impacts of violent media and socio-cognitive provocation.

2. Keywords

Buss-Perry Aggression Questionnaire, BPAQ, trait aggression, physical aggression, verbal aggression, anger, hostility, psychometrics, factor analysis, violence risk assessment

3. Authors

The Buss-Perry Aggression Questionnaire was developed by Arnold H. Buss and Mark Perry. At the time of the instrument’s publication in 1992, both researchers were affiliated with the Department of Psychology at the University of Texas at Austin (Austin, Texas, United States). Arnold H. Buss (1924–2021) was an internationally renowned scholar in the fields of human aggression, temperament, social behavior, and experimental psychopathology. Prior to his seminal work on the BPAQ, Buss made foundational contributions to the scientific understanding of aggressive behavior through his 1961 monograph The Psychology of Aggression and his earlier collaborative design of the Buss-Durkee Hostility Inventory with Ann Durkee in 1957. Mark Perry collaborated closely with Buss on modernizing the measurement of aggression using contemporary statistical, computational, and structural equation modeling paradigms that were unavailable during mid-twentieth-century test construction.

4. Purpose

The primary purpose of the Buss-Perry Aggression Questionnaire is to provide an empirically grounded, multidimensional self-report inventory that operationalizes and measures individual differences in trait aggression. For several decades following its publication in 1957, the Buss-Durkee Hostility Inventory (BDHI) served as the standard assessment instrument in psychological research on hostile behavior. However, by the late 1980s, psychometricians and behavioral researchers had identified fundamental structural vulnerabilities in the BDHI. Chief among these shortcomings were its reliance on a dichotomous (True/False) response format—which artificially truncated variance, impaired statistical power, and heightened vulnerability to acquiescence response bias—and an unstable factor structure that frequently failed to replicate across exploratory factor analyses conducted by independent research laboratories.

Recognizing the acute necessity for an updated, theoretically coherent, and psychometrically defensible measurement tool, Buss and Perry designed the BPAQ to parse the overarching construct of aggression into its constituent biological, behavioral, emotional, and cognitive systems. Trait aggression is not a monolithic psychological entity; rather, an individual may experience intense cognitive grievances (hostility) without ever engaging in motoric outbursts (physical aggression), or conversely, may engage in tactical instrumental violence in the absence of acute emotional dysregulation (anger). The theoretical rationale driving the BPAQ posits that a complete psychometric profile of an individual’s aggressive propensity requires simultaneous quantification of overt behavioral acts, affective states, and cognitive appraisals.

In research applications, the BPAQ is deployed to investigate the etiology and behavioral consequences of aggression across diverse social and environmental contexts. It serves as an indispensable tool in experimental social psychology for categorizing participants into baseline aggression cohorts prior to laboratory stress, frustration induction, or social rejection tasks (such as the Taylor Aggression Paradigm). Furthermore, it is extensively utilized in media psychology to trace longitudinal fluctuations in aggressive orientation resulting from chronic exposure to violent video games, film, and digital interaction. In sports science, researchers utilize the BPAQ to evaluate competitive aggression, athletic performance, and rule-violating conduct across collision versus non-collision sports.

Clinically, the BPAQ provides vital diagnostic and prognostic utility. In psychiatric and forensic environments, the scale assists in the assessment of intermittent explosive disorder, antisocial personality disorder, borderline personality disorder, and conduct disorders. Forensic psychologists integrate the BPAQ into violence risk appraisal batteries to gauge institutional misconduct risk among incarcerated offenders, evaluate post-release recidivism likelihood, and monitor therapeutic progress in anger management and cognitive behavioral therapy (CBT) rehabilitation programs.

5. Psychological Construct

The construct operationalized by the BPAQ is trait aggression, defined as the stable, enduring disposition to inflict psychological or physical harm, harbor resentful and malevolent beliefs toward others, or experience physiological and emotional states of rage and indignation. Buss and Perry (1992) posited that aggression is fundamentally tripartite, encompassing motor, affective, and cognitive modalities. Within the BPAQ, these three overarching modalities are divided into four precisely defined, interrelated subscales:

Physical Aggression (Motor Component)

Physical Aggression represents the behavioral, instrumental, and somatic manifestation of aggression. It involves the direct application of bodily force or mechanical weapons against other human beings or inanimate objects to cause damage, injury, coercion, or intimidation. Individuals who score high on this subscale exhibit a low behavioral threshold for kinetic retaliation, frequently escalating conflicts into overt physical altercations, assaultive acts, property destruction, or battery. This subscale comprises 9 items reflecting overt kinetic actions (e.g., hitting, striking, using physical force to defend rights, breaking objects, or getting pushed to blows). It measures actual motor behavior rather than latent ideation or mere subjective impulse.

Verbal Aggression (Motor Component)

Verbal Aggression constitutes the communicative and vocal branch of the motor component. It involves the deliberate use of spoken or written discourse to inflict interpersonal distress, dominate social encounters, harass, insult, or express aggressive opposition. Unlike physical violence, verbal aggression seeks to demean, degrade, or coerce through linguistic channels. The 5 items comprising this subscale capture tendencies toward excessive disputatiousness, argumentative confrontation, direct verbal insults, and uncontrollable outspokenness when irritated (e.g., engaging in arguments when disagreements arise, overtly asserting disagreement, telling people what one thinks of them when provoked). Verbal aggression represents an accessible and socially prevalent vehicle for instrumental and expressive hostility.

Anger (Affective Component)

Anger embodies the physiological, emotional, and autonomic aspect of aggression. It is characterized by visceral activation of the sympathetic nervous system, emotional agitation, temper dysregulation, and feelings of fury, rage, or explosive irritability. Buss and Perry conceptualized anger as the crucial psychological “bridge” connecting cognitive appraisals with motoric execution: while hostility prepares the cognitive framework and physical aggression executes the act, anger supplies the acute emotional fuel and autonomic drive. The 7 items of the Anger subscale evaluate subjective experiences of emotional volatility, difficulty with self-regulation, rapid affective escalation, feeling like a “powder keg,” and impulsive frustration responses.

Hostility (Cognitive Component)

Hostility represents the cognitive, attitudinal, and interpretive dimension of aggression. It consists of negative, suspicious, resentful, and cynical evaluations of human relationships and the social environment. Hostile individuals interpret benign or ambiguous social actions as intentionally malicious, maintain pervasive distrust of strangers and acquaintances alike, harbor jealousy toward others’ achievements, and experience chronic bitterness regarding their own life outcomes. The 8 items within this subscale gauge interpersonal paranoia, cynicism, jealousy, and social resentment (e.g., suspecting overly friendly strangers, feeling that people talk or laugh behind one’s back, and believing that one has received an unjust or “raw deal” from life). Hostility operates as a cognitive schema that perpetually justifies and primes defensive or preemptive aggressive action.

6. Theoretical Framework

The Buss-Perry Aggression Questionnaire is grounded in an integration of prominent behavioral, socio-cognitive, and affective theories of human aggression developed throughout the twentieth century. The theoretical architecture reflects the convergence of the classical Frustration-Aggression Hypothesis (Dollard et al., 1939), Albert Bandura’s Social Learning Theory (1973), Leonard Berkowitz’s Cognitive Neo-Associationistic Model (1989, 1990), and modern iterations of the General Aggression Model (GAM; Anderson & Bushman, 2002).

Dollard, Doob, Miller, Mowrer, and Sears (1939) initially argued that aggression is always a consequence of frustration, defined as the thwarting of goal-directed behavior. Although this deterministic premise was subsequently modified, the core proposition that external blockages elicit affective distress remains foundational to the BPAQ’s Anger subscale. Buss (1961) expanded upon this model by decoupling instrumental aggression—wherein aggressive acts are performed as calculated tools to secure extrinsic rewards or resources—from emotional or reactive aggression, which is driven by acute emotional agitation and visceral arousal. This distinction directly informed the structural bifurcation of the BPAQ into instrumental motor actions (Physical and Verbal Aggression) versus affective states (Anger).

Leonard Berkowitz’s cognitive neoassociation theory provided the theoretical scaffolding for integrating the cognitive (Hostility) and affective (Anger) components into a unified predictive network. Berkowitz posited that aversive events—such as physical pain, social insults, environmental stressors, or psychological frustration—generate unconditioned negative affect. Once negative affect is elicited, it activates an associative cognitive network composed of hostile cognitions, physiological fight-or-flight reflexes, motor impulses, and expressive somatic reactions. Within this theoretical network, hostile attributions (e.g., “people are laughing at me”) and angry feelings act as mutually reinforcing catalysts that dramatically elevate the statistical probability of overt physical or verbal behavioral output.

Bandura’s Social Learning Theory further accounts for the behavioral acquisition and maintenance of the motor components measured by the BPAQ. According to Bandura (1973), individuals acquire repertoire-specific aggressive motor skills through observational modeling, vicarious reinforcement, and direct social conditioning. While internal anger and hostile cognitive scripts create internal readiness, the selection of physical violence versus verbal rebuttal is dictated by learned outcome expectancies, perceived self-efficacy, and internal standards of moral conduct. The BPAQ operationalizes these diverse theoretical components into discrete, empirically verifiable factors, allowing researchers to explore how cognitive schemas (Hostility) activate physiological arousal (Anger) to produce overt behavioral consequences (Physical and Verbal Aggression).

7. Validity

The validity of the Buss-Perry Aggression Questionnaire has been extensively scrutinized across clinical, collegiate, adult community, and forensic populations, establishing exceptionally strong evidence for construct, convergent, discriminant, and predictive validity.

Construct and Factorial Validity

During the initial development and psychometric validation of the scale, Buss and Perry (1992) established construct validity by administering an initial pool of 52 candidate items to a foundational normative sample of 1,253 undergraduate students (612 men and 641 women). Exploratory and confirmatory factor analyses systematically culled ambiguous, weakly loading, and multi-loading items, resulting in a clean, four-factor solution comprising 29 items. Subsequent confirmatory factor analyses across international contexts (e.g., Harris, 1997; Archer et al., 2005) have supported the four-factor construct, demonstrating that the structural model fits observed data significantly better than unidimensional or two-factor models.

Convergent Validity

Convergent validity has been repeatedly corroborated through correlations with complementary personality and behavioral scales. Buss and Perry (1992) demonstrated substantial correlations between BPAQ subscales and measures of impulsivity, assertiveness, competitive disposition, and general emotionality. When evaluated against the Five-Factor Model of personality, the BPAQ exhibits robust, statistically significant negative correlations with Agreeableness (with correlation coefficients typically ranging between r = -.45 and r = -.60), reflecting the fundamental interpersonal antagonism shared by high BPAQ scorers. Furthermore, the Anger and Hostility subscales correlate positively and strongly with Neuroticism (r = .35 to .55), aligning with the emotional instability and threat-sensitivity central to that trait dimension. The BPAQ also correlates robustly with the State-Trait Anger Expression Inventory (STAXI; Spielberger, 1988), particularly between the BPAQ Anger subscale and STAXI Trait Anger (r > .70).

Discriminant Validity

Discriminant validity is evidenced by the distinct correlational patterns observed among the four subscales and external demographic and behavioral variables. In their original investigation, Buss and Perry (1992) observed marked sex differences across the subscales: men scored substantially higher than women on Physical Aggression (Cohen’s d ≈ 0.60 to 0.80), whereas sex differences for Verbal Aggression and Hostility were negligible or modest, and sex differences for Anger were minimal, with women occasionally reporting slightly higher rates of affective volatility. This divergence confirms that the subscales do not merely capture generic psychological distress, but instead isolate specific behavioral and affective dynamics.

Predictive and Criterion Validity

The BPAQ exhibits robust predictive and criterion-related validity in laboratory and ecological contexts. In experimental paradigms utilizing the Taylor Aggression Paradigm or the Hot Sauce Paradigm, individuals with high Physical Aggression baseline scores administered significantly higher shock voltages, longer noise blasts, or larger quantities of noxious stimulus to ostensible competitors. In peer-nomination studies, Buss and Perry (1992) collected external peer ratings of aggressive conduct from college roommates and close associates; these peer evaluations correlated substantially with self-reported Physical Aggression (r = .45) and Verbal Aggression (r = .38). In forensic settings, high scores on the Physical Aggression and Hostility subscales reliably predict institutional disciplinary infractions, violent criminal histories, and recidivism upon community release.

8. Reliability

The reliability of the Buss-Perry Aggression Questionnaire has been verified across multiple independent psychometric investigations, demonstrating high internal consistency and longitudinal temporal stability.

Internal Consistency

In the seminal standardization study conducted by Buss and Perry (1992) on 1,253 university participants, internal consistency was calculated using Cronbach’s alpha coefficients for the four subscales as well as the total questionnaire score:

• Physical Aggression (9 items): α = .85
• Verbal Aggression (5 items): α = .72
• Anger (7 items): α = .83
• Hostility (8 items): α = .77
• Total Aggression Scale (29 items): α = .89

These values demonstrate that all subscales possess acceptable to excellent internal consistency. The slightly lower alpha coefficient observed for Verbal Aggression (α = .72) is largely attributable to its smaller item pool (5 items) rather than poor inter-item covariance, as Cronbach’s alpha is inherently sensitive to scale length. Subsequent validation studies across diverse cultural, linguistic, and demographic samples have mirrored these findings: Physical Aggression consistently registers alphas between .80 and .88; Anger between .78 and .86; Hostility between .75 and .82; and Verbal Aggression between .70 and .78.

Test-Retest Reliability

To establish the temporal stability of the BPAQ, Buss and Perry (1992) re-administered the scale to a subsample of 372 participants after a nine-week temporal interval. The Pearson product-moment correlation coefficients across this nine-week window were:

• Physical Aggression: r = .80
• Verbal Aggression: r = .76
• Anger: r = .72
• Hostility: r = .72
• Total Aggression Scale: r = .80

These stability coefficients satisfy standard psychometric benchmarks for personality trait inventories, confirming that the BPAQ measures stable, enduring behavioral dispositions rather than fleeting state fluctuations or temporary situational reactions.

9. Factor Analysis

The structural composition of the BPAQ was initially established through exploratory factor analysis (EFA) and subsequently evaluated using confirmatory factor analysis (CFA) across numerous independent empirical investigations.

Original EFA and CFA by Buss and Perry (1992)

Buss and Perry began item reduction with an initial pool of 52 items derived from theoretical operationalizations and adapted BDHI content. Using principal components analysis followed by oblique rotation (promax/oblimin)—which appropriately allows psychological factors to intercorrelate—items with low primary loadings (< .35) or substantial cross-loadings were systematically eliminated. This iterative reduction produced a refined 29-item instrument comprising four cleanly differentiated factors accounting for substantial total variance.

Subscale intercorrelations reported in the original validation study demonstrated that while the factors represent distinct constructs, they share meaningful variance:

• Physical Aggression and Verbal Aggression: r = .45
• Physical Aggression and Anger: r = .43
• Physical Aggression and Hostility: r = .34
• Verbal Aggression and Anger: r = .48
• Verbal Aggression and Hostility: r = .34
• Anger and Hostility: r = .54

Notably, Anger and Hostility exhibited the strongest inter-factor correlation (r = .54), aligning with theoretical predictions that negative affective arousal and hostile cognitive attributions are tightly coupled.

Subsequent Factorial Replications and Structural Modifications

While the four-factor model is widely accepted, subsequent rigorous CFA studies utilizing modern fit metrics (such as the Root Mean Square Error of Approximation [RMSEA], Comparative Fit Index [CFI], and Tucker-Lewis Index [TLI]) have occasionally reported suboptimal fit for the original 29-item model when applied to non-college, clinical, or international populations. Researchers such as Harris (1997) and Archer, Kilpatrick, and Bramwell (2005) identified that certain reverse-scored items (specifically items 9 and 16) and items with complex semantics exhibited secondary cross-loadings or method effects associated with negative wording.

To address these structural issues, Bryant and Smith (2001) performed extensive exploratory and confirmatory factor analyses, identifying a refined, psychometrically superior 12-item short form (featuring 3 items per subscale) that yielded exceptional goodness-of-fit indices (CFI > .95, RMSEA < .05). Similarly, Webster et al. (2014) validated a Brief Aggression Questionnaire (BAQ) using item-response theory and CFA. Nonetheless, the full 29-item four-factor model remains the universally recognized standard in clinical and academic literature, with hierarchical and bifactor models demonstrating that a general “Aggression” overarching factor successfully captures shared variance, while the four distinct group factors retain robust construct-specific variance.

10. Instrument / Measurement Tool

• Test Name: Buss-Perry Aggression Questionnaire (BPAQ)
• Alternative Titles: The Aggression Questionnaire (AQ)
• Authors: Arnold H. Buss and Mark Perry (1992)
• Construct Measured: Multidimensional trait aggression (Physical Aggression, Verbal Aggression, Anger, Hostility)
• Test Type: Self-report psychometric rating scale
• Target Population: Adolescents (ages 12+) and adults; utilized in college, community, forensic, and psychiatric cohorts
• Administration Format: Individual or group administration; paper-and-pencil or computerized/digital formats
• Administration Time: Approximately 5 to 10 minutes
• Number of Items: 29 items
• Response Scale: Authentic 5-point Likert scale:
  • 1 = extremely uncharacteristic of me
  • 2 = somewhat uncharacteristic of me
  • 3 = neither characteristic nor uncharacteristic of me
  • 4 = somewhat characteristic of me
  • 5 = extremely characteristic of me
• Reverse-Scoring Rules: Items 9 and 16 are reverse-scored prior to aggregation (1 = 5, 2 = 4, 3 = 3, 4 = 2, 5 = 1).
• Subscale Item Composition:
  • Physical Aggression (9 items): Items 2, 5, 8, 11, 13, 16, 22, 25, 29
  • Verbal Aggression (5 items): Items 4, 6, 14, 21, 27
  • Anger (7 items): Items 1, 9, 12, 17, 19, 23, 28
  • Hostility (8 items): Items 3, 7, 10, 15, 18, 20, 24, 26
• Scoring and Interpretation:
  • Subscale scores are derived by calculating the direct sum of the designated items within each dimension.
  • The Total Aggression Score is calculated by summing all 29 items, yielding a possible score range of 29 to 145.
  • Higher total and subscale scores indicate higher levels of trait aggression. Normative comparisons frequently examine sex-stratified percentiles, as males typically exhibit higher mean scores on physical aggression.

11. Permissions & Fee and Test Year

The Buss-Perry Aggression Questionnaire was originally published in 1992 in the Journal of Personality and Social Psychology, a journal owned and published by the American Psychological Association (APA). The foundational bibliographic citation is: Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality and Social Psychology, 63(3), 452–459.

Under conventional academic fair use practices, the 29 items of the BPAQ are widely utilized for non-commercial, scholarly research and educational purposes without licensing fees, provided proper academic attribution is maintained. However, commercial utilization, integration into proprietary clinical assessment software, or inclusion in for-profit diagnostic services requires formal permission and licensing authorization from the copyright holder (the American Psychological Association). Researchers planning clinical trials, organizational testing, or commercial diagnostic implementations must secure official approval via the APA Permissions Office prior to deployment.

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

Archer, J., Kilpatrick, G., & Bramwell, R. (2005). Comparison between a modified and standard version of the Buss–Perry Aggression Questionnaire. Aggressive Behavior, 31(1), 68–73. https://doi.org/10.1002/ab.20038

Bandura, A. (1973). Aggression: A social learning analysis. Prentice-Hall.

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

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

Bryant, F. B., & Smith, B. D. (2001). Refining the architecture of aggressive forms: A confirmatory factor analysis of the Buss–Perry Aggression Questionnaire. Journal of Research in Personality, 35(2), 138–167. https://doi.org/10.1006/jrpe.2000.2302

Buss, A. H. (1961). The psychology of aggression. John Wiley & Sons.

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

Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. (1939). Frustration and aggression. Yale University Press. https://doi.org/10.1037/10022-000

Harris, J. A. (1997). A further evaluation of the Aggression Questionnaire: Issues of validity and scoring. Behaviour Research and Therapy, 35(11), 1047–1053. https://doi.org/10.1016/S0005-7967(97)00064-7

Spielberger, C. D. (1988). State-Trait Anger Expression Inventory: Professional manual. Psychological Assessment Resources.

Webster, G. D., DeWall, C. N., Pond, R. S., Deckman, T., Jonason, P. K., Le, B. M., Richard, A. T., Schember, T. O., Crysel, L. C., Crosby, F. J., & Bator, R. J. (2014). The Brief Aggression Questionnaire: Structure, validity, reliability, and generalizability. Journal of Personality Assessment, 96(2), 120–138. https://doi.org/10.1080/00223891.2013.838170

13. Items of the Scale