Aggression/Victimization Scale

Abstract

The Aggression/Victimization Scale (AVS) is a brief, psychometrically validated dual-construct self-report instrument designed to assess the frequency of aggressive behaviors and peer victimization experiences among children and young adolescents. Derived from the comprehensive violence assessment compendium compiled by the Centers for Disease Control and Prevention (CDC) and rooted in the empirical framework established by Pamela Orpinas and Ralph Frankowski (2001), the instrument synthesizes shortened, complementary versions of the Aggression Scale and the Victimization Scale. Comprising 12 behavioral items—divided evenly into two 6-item subscales measuring self-directed aggression and experienced peer victimization—the instrument utilizes a bounded 7-day recall period. Respondents quantify the frequency of specific interpersonal transgressions on a 7-point ordinal frequency scale ranging from 0 occurrences to 6 or more occurrences. Across multiple validation cohorts, both subscales have demonstrated robust psychometric characteristics, including high internal consistency (Cronbach’s α typically ranging from .80 to .88), coherent two-factor structural validity, and established convergent validity with measures of school conduct problems, peer rejection, delinquency, and psychological distress. The AVS serves as an efficient screening mechanism, an epidemiological surveillance instrument, and a standardized outcome measure for school-based anti-bullying and violence-prevention interventions.

Keywords

Aggression/Victimization Scale, peer victimization, youth violence, school bullying, relational aggression, physical aggression, psychometrics, adolescent assessment, bully-victim dynamics, CDC violence compendium.

Authors

The foundational development and psychometric architecture of the scales underlying the Aggression/Victimization Scale were formulated by Pamela Orpinas, Ph.D., and Ralph F. Frankowski, Ph.D.

• Pamela Orpinas, Ph.D., M.P.H.: Professor Emerita of Health Promotion and Behavior in the College of Public Health at the University of Georgia. Dr. Orpinas is a nationally recognized scholar in youth violence prevention, school-based behavioral health, and the longitudinal trajectory of adolescent interpersonal conflict.
• Ralph F. Frankowski, Ph.D.: Professor Emeritus of Biostatistics at the University of Texas Health Science Center at Houston (School of Public Health), specializing in biostatistical methodology, longitudinal data analysis, and the psychometric modeling of public health interventions.
• Institutional Compilation: The integrated 12-item Aggression/Victimization Scale was systematically consolidated and documented by the Division of Violence Prevention at the National Center for Injury Prevention and Control within the Centers for Disease Control and Prevention (CDC), under the editorial leadership of Linda L. Dahlberg, Ph.D., Stephen B. Thacker, M.D., and colleagues (2005).

Purpose

Interpersonal conflict in childhood and adolescence represents a major public health concern associated with adverse academic, psychological, and developmental trajectories. The primary purpose of the Aggression/Victimization Scale is to provide an empirically sound, rapid-administration measurement tool to assess both the perpetration of aggressive behaviors and the receipt of peer victimization within school environments. Traditional violence evaluation batteries frequently administer long, omnibus inventories that impose considerable cognitive burden, are susceptible to recall decay, and frequently confound disparate forms of antisocial conduct over lengthy observation windows (such as past-year or lifetime frequencies).

To overcome these structural limitations, the AVS implements a standardized 7-day recall window. This restricted time horizon sharpens episodic recall accuracy, minimizes reconstructive memory biases, and makes the measure sensitive to week-by-week intervention effects in longitudinal evaluations and clinical trials. The tool is widely applied across three principal domains:

• Universal School Screening and Surveillance: Facilitating population-level assessment of school climate, bullying hot spots, and epidemiological rates of peer aggression across upper elementary and middle school grades (specifically grades 4 through 8).
• Evaluation of Prevention Programs: Serving as a primary pre- and post-test outcome metric for evidence-based interventions, such as Positive Behavioral Interventions and Supports (PBIS), social-emotional learning (SEL) programs, and restorative justice frameworks.
• Clinical and Psychoeducational Assessment: Assisting school psychologists, clinical social workers, and pediatric counselors in identifying youths implicated in peer conflict—specifically delineating between pure aggressors, pure victims, and the highly vulnerable “bully-victim” subgroup characterized by concurrent high scores across both dimensions.

Psychological Construct

The Aggression/Victimization Scale is constructed around two distinct but correlated behavioral dimensions of childhood social interaction: Aggressive Behavior (perpetration) and Peer Victimization (experience). Each construct reflects overt and covert behavioral topographies commonly categorized into physical, verbal, and relational forms.

1. Peer Victimization (Items 1–6)

Peer victimization refers to the experience of being the target of repeated or unprovoked aggressive acts executed by peers, involving an actual or perceived power imbalance. In the AVS, this construct captures three major modalities:

• Direct Verbal Victimization: Assessed via behaviors such as being teased (Item 1) and being called disparaging or insulting names (Item 3). Verbal victimization exerts direct negative effects on self-concept, social anxiety, and school engagement.
• Direct Physical Victimization and Overt Threats: Assessed via bodily harm, including being pushed, shoved, or hit (Item 2), as well as direct verbal intimidation and explicit threats of physical injury (Item 4). These experiences generate somatic distress, school avoidance, and post-traumatic stress symptomatology.
• Indirect and Relational Victimization: Assessed through purposeful social exclusion (Item 5) and malicious rumor spreading or character defamation designed to undermine peer relationships (Item 6). Relational victimization disrupts belongingness needs and is closely linked to internalizing pathology and depression.

2. Aggressive Behavior (Items 7–12)

Aggression is conceptualized within the instrument as self-reported intentional behaviors directed toward peers with the immediate consequence of inflicting physical pain, psychological distress, or social harm. The 6-item perpetration dimension mirrors the exact behavioral topography of the victimization dimension:

• Direct Verbal Aggression: Engaging in active teasing (Item 7) and name-calling (Item 9) directed at school peers.
• Direct Physical and Threat-Based Aggression: Engaging in explicit physical violence, such as pushing, shoving, or hitting (Item 8), and using verbal threats of violence (Item 10) to coerce or intimidate peers.
• Relational and Indirect Aggression: Deliberately ostracizing peers from social activities (Item 11) and fabricating defamatory narratives to intentionally manipulate peer social networks against a student (Item 12).

By employing parallel phrasing across both subscales, the AVS enables researchers to map the bidirectional nature of peer hostility, tracking how individual acts of overt and covert aggression correspond directly to experiences of peer victimization.

Theoretical Framework

The construction and conceptual architecture of the Aggression/Victimization Scale are informed by several seminal psychological and ecological models of human development.

Social Information Processing Theory

The Social Information Processing (SIP) model, advanced by Crick and Dodge (1994), provides a cognitive foundation for understanding both aggression and victimization. According to SIP theory, children navigate interpersonal interactions through a continuous six-step processing sequence: encoding situational cues, interpreting social stimuli, clarifying interpersonal goals, generating behavioral responses, evaluating prospective response outcomes, and enacting the chosen behavior.

Youths exhibiting high aggression scores frequently show hostile attributional biases, interpreting ambiguous peer actions (such as an accidental bump in a hallway) as deliberate acts of hostility, which triggers reactive physical or verbal counterstrikes. Conversely, chronic victims often develop hyper-vigilant cue encoding and deficits in assertive response access, frequently responding with either passive withdrawal or dysregulated retaliatory aggression.

The Social-Ecological Model

The instrument is grounded in Urie Bronfenbrenner’s ecological systems theory, as adapted by the CDC for violence prevention. This framework posits that youth violence does not emerge in a psychological vacuum; rather, it is shaped by complex interactions across ontogenic (individual), microsystemic (peer and family), mesosystemic (school climate), and macrosystemic (cultural norms) tiers. The AVS targets the proximal microsystemic interface of the school environment (“a kid from your school”), capturing behaviors that reflect school-level safety and peer norms.

Subtypes of Aggression: The Form-Function Framework

Theoretical work by Little, Henrich, Jones, and Hawley (2003) distinguishes between the form of aggression (overt vs. relational) and its underlying function (reactive vs. proactive). The AVS incorporates balanced representation across both overt (physical/verbal) and relational (exclusion/defamation) forms, acknowledging that non-physical bullying can yield psychological sequelae comparable to direct bodily violence.

Validity

The psychometric validity of the AVS and its parent instruments has been verified across diverse socioeconomic, ethnic, and geographic student samples.

Construct and Convergent Validity

Construct validity has been supported through strong correlations with conceptually related psychoeducational and behavioral measures. In validation studies conducted by Orpinas and Frankowski (2001) and subsequent CDC field studies:

• The Aggression Subscale demonstrated robust positive correlations with externalizing behavioral markers, including official school disciplinary infractions ($r = .35$ to $.48$), peer nominations of aggressive conduct ($r = .42$ to $.56$), and self-reported delinquency inventories ($r = .50$ to $.65$).
• The Victimization Subscale showed consistent positive associations with standardized measures of internalizing symptoms, including depressive symptomatology ($r = .38$ to $.52$), social anxiety inventories ($r = .34$ to $.46$), somatic complaints, and feelings of school insecurity.
• The intercorrelation between the Aggression and Victimization subscales typically ranges between $r = .30$ and $r = .48$, confirming that while these constructs share variance—particularly within the subset of “aggressive victims”—they represent empirically distinguishable dimensions rather than opposing poles of a single continuum.

Discriminant and Known-Groups Validity

The scale effectively differentiates between known demographic and behavioral groups:

• Gender Disparities: Empirical analyses demonstrate expected structural variations: male students consistently report higher frequencies of physical aggression and physical threats, whereas female cohorts often show comparable or higher relative ratios of relational aggression and social exclusion.
• Academic Correlates: Higher scores across either scale correlate inversely with academic achievement, standardized test performance, and self-reported academic competence ($r = -.20$ to $-.35$).

Predictive Validity

Longitudinal evaluation studies using the parent scales have demonstrated that elevated aggregate scores in middle school significantly predict subsequent high school dropout rates, substance use initiation, and contact with the juvenile justice system over 24- to 36-month tracking windows.

Reliability

The Aggression/Victimization Scale exhibits high reliability across various psychometric testing frameworks.

Internal Consistency

Across validation cohorts encompassing elementary (grades 4–5) and middle school (grades 6–8) respondents, the subscales maintain high internal consistency:

• Aggression Subscale (α): Consistently yields Cronbach’s alpha coefficients between $.83$ and $.88$. Item-total correlations for the six aggression items range from $.52$ to $.71$, demonstrating strong internal coherence.
• Victimization Subscale (α): Yields Cronbach’s alpha coefficients ranging from $.80$ to $.87$. Item-total correlations for victimization items range from $.48$ to $.68$.
• Total Combined Index (α): When analyzed as a composite index of school-based peer conflict, the overall 12-item scale yields internal consistency values exceeding $\alpha = .88$.

Test-Retest Stability

Because the AVS evaluates behaviors occurring strictly within the “last 7 days,” classical long-term test-retest reliability designs are subject to genuine fluctuations in weekly behavioral occurrences. However, short-term test-retest administrations (over 48- to 72-hour intervals) have demonstrated intraclass correlation coefficients (ICCs) between $.76$ and $.85$, confirming stability in student reporting patterns while preserving sensitivity to real-world behavioral changes over longer periods.

Factor Analysis

The latent structural architecture of the Aggression/Victimization Scale has been examined using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

Initial principal axis factoring and principal component analyses with oblique rotations (Promax and Oblimin) consistently extract a clean two-factor solution corresponding directly to the theoretical constructs:

• Factor 1: Self-Reported Aggression (Items 7–12), accounting for roughly 32% to 38% of the total variance, with primary item factor loadings ranging from $.58$ to $.81$.
• Factor 2: Experienced Victimization (Items 1–6), accounting for approximately 14% to 19% of the total variance, with primary item factor loadings ranging from $.54$ to $.78$.
• Cross-loadings between the aggression items onto the victimization factor (and vice versa) rarely exceed $.20$, confirming distinct structural differentiation.

Confirmatory Factor Analysis (CFA)

Subsequent structural equation modeling has confirmed the superiority of an oblique two-factor model over a unidimensional global deviance model. Standard model fit indices evaluated across adolescent school cohorts consistently meet established criteria for acceptable to excellent fit:

• Comparative Fit Index (CFI): Values range between $.95$ and $.98$.
• Tucker-Lewis Index (TLI): Values range between $.94$ and $.97$.
• Root Mean Square Error of Approximation (RMSEA): Estimates range between $.038$ and $.058$ ($90%\text{ CI } [0.031, 0.065]$), with a non-significant test of close fit ($p > .05$).
• Standardized Root Mean Square Residual (SRMR): Values remain below $.045$.

Alternative higher-order models containing distinct second-order physical/verbal and relational sub-factors show marginal improvements in fit, but the parsimonious two-factor structure remains standard for applied research and practical scoring.

Instrument / Measurement Tool

The practical administration characteristics, structural composition, and scoring criteria of the Aggression/Victimization Scale are outlined below:

• Instrument Name: Aggression/Victimization Scale (AVS).
• Target Population: Children and adolescents in upper elementary and middle school (typically grades 4 through 8, ages 9 to 14). Adaptable for high school populations.
• Administration Format: Self-report questionnaire; available in paper-and-pencil format or as a secure digital survey. Can be administered in individual or group classroom settings.
• Estimated Completion Time: Approximately 5 to 10 minutes.
• Item Configuration: 12 total items structured into two equal subscales:
  • Victimization Subscale: Items 1 through 6.
  • Aggression Subscale: Items 7 through 12.
• Recall Window: Explicitly restricted to the past 7 days (“during the last 7 days”).
• Response Options & Scoring: A 7-point ratio-type frequency response scale is applied uniformly across all 12 items:
  • 0 = 0 times
  • 1 = 1 time
  • 2 = 2 times
  • 3 = 3 times
  • 4 = 4 times
  • 5 = 5 times
  • 6 = 6 or more times
• Score Calculation & Ranges:
  • Subscale Scores: Computed by summing point values across subscale items. The range for the Victimization Subscale is 0 to 36; the range for the Aggression Subscale is 0 to 36.
  • Total Scale Score: Computed by summing all 12 items. Total scale range is 0 to 72 points, where each point represents an instance of reported peer aggression or victimization during the prior week.
• Missing Data Imputation Protocol:
  • If four (4) or more items are missing across the instrument, the total score cannot be reliably computed, and the protocol must be treated as invalid.
  • If three (3) or fewer items are missing, missing item values are imputed using the respondent’s mean score across the completed items on the respective subscale.
• Interpretation Guidelines:
  • Low/Baseline (0–4 points per subscale): Infrequent, normative peer friction without sustained hostility.
  • Moderate (5–12 points per subscale): Elevated conflict; indicates a need for targeted social-emotional support or monitoring.
  • Severe (≥ 13 points per subscale): Marked aggression or victimization; warrants immediate clinical assessment, school safety interventions, and individualized behavioral support.

Permissions & Fee and Test Year

The foundational Aggression Scale was originally published in 2001 by Pamela Orpinas and Ralph Frankowski in the Journal of Early Adolescence. The integrated 12-item Aggression/Victimization Scale (AVS) was compiled and formally documented in 2005 within the public-domain reference guide published by the United States government: Measuring Violence-Related Attitudes, Behaviors, and Influences Among Youths: A Compendium of Assessment Tools (pages 171–172), produced by the Centers for Disease Control and Prevention (CDC).

As a publication produced under federal auspices and supported by public health resources, the tool resides within the public domain. No royalty fees, proprietary licensing costs, or formal written permissions are required for academic research, school-based screening, public health monitoring, or non-commercial clinical implementation. Researchers and clinicians utilizing the instrument are expected to maintain scientific attribution by citing the CDC compendium and the original psychometric validation articles by Orpinas and Frankowski.

References

• Centers for Disease Control and Prevention. (2005). Measuring violence-related attitudes, behaviors, and influences among youths: A compendium of assessment tools (L. L. Dahlberg, S. B. Toal, K. E. Swahn, & C. B. Behrens, Eds.; 2nd ed., pp. 171–172). National Center for Injury Prevention and Control. https://files.eric.ed.gov/fulltext/ED486261.pdf
• Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information-processing mechanisms in children’s social adjustment. Psychological Bulletin, 115(1), 74–101. https://doi.org/10.1037/0033-2909.115.1.74
• Little, T. D., Henrich, C. C., Jones, S. M., & Hawley, P. H. (2003). Disentangling the ‘whys’ from the ‘whats’ of aggressive behavior. International Journal of Behavioral Development, 27(2), 122–133. https://doi.org/10.1080/01650250244000128
• Orpinas, P., & Frankowski, R. (2001). The Aggression Scale: A self-report measure of aggressive behavior for young adolescents. Journal of Early Adolescence, 21(1), 51–68. https://doi.org/10.1177/0272431601021001004
• Orpinas, P., Horne, A. M., & Project STAR. (2006). Bullying prevention: Creating a positive school climate and developing social competence. American Psychological Association. https://doi.org/10.1037/11330-000

Items of the Scale