Interaction With Antisocial Peers Scale

Abstract

The Interaction With Antisocial Peers Scale (IWAP)—originating as a foundational risk-factor subscale within the Communities That Care (CTC) Youth Survey and the Student Survey of Risk and Protective Factors (SSRPF)—is a psychometrically validated, self-report instrument designed to quantify adolescent exposure and affiliation with deviant peer groups. Developed by the Social Development Research Group (SDRG) at the University of Washington, the scale captures the degree to which an adolescent’s closest social network engages in norm-violating, illegal, and institutional-transgressing behaviors. Comprising six targeted items, the instrument operationalizes peer deviance through an anchored nomination paradigm: respondents evaluate their four closest friends across specific severe delinquency benchmarks over a 12-month retrospective window (including school suspension, weapon possession, illicit drug distribution, motor vehicle theft, formal arrest, and school dropout). Responses are recorded on an anchored ordinal scale ranging from “None of my friends” (0) to “4 of my friends” (4), yielding a continuous metric of peer antisocial saturation. Extensive psychometric evaluations across large-scale epidemiological cohorts (spanning grades 6 through 12, N > 100,000) have repeatedly confirmed robust internal consistency, with Cronbach’s alpha coefficients typically ranging between α = .78 and .86, strong factor determinacy in confirmatory factor analytic frameworks, and strict longitudinal measurement invariance across sex and racial/ethnic groups. The instrument demonstrates powerful predictive validity regarding subsequent personal substance abuse, violent offending, and academic disengagement. By isolating the social-environmental nexus of peer contagion, the IWAP functions as an essential measurement tool within developmental criminology, school psychology, and public health prevention science.

Keywords

Interaction With Antisocial Peers Scale, IWAP, peer delinquency, Communities That Care, Social Development Model, adolescent deviance, peer contagion, risk factors, criminology, psychometrics, externalizing behavior, youth violence, substance use etiology, SDRG

Authors

The Interaction With Antisocial Peers Scale was developed under the auspices of the Social Development Research Group (SDRG), a specialized interdisciplinary research center housed within the School of Social Work at the University of Washington (Seattle, Washington, United States). The conceptualization, psychometric refinement, and nationwide validation of the parent instrument—the Communities That Care Youth Survey (CTC-YS) and the Student Survey of Risk and Protective Factors (SSRPF)—were led by principal investigators:

• Michael W. Arthur, Ph.D. – Research Associate Professor (Emeritus), Social Development Research Group, University of Washington School of Social Work.
• J. David Hawkins, Ph.D. – Endowed Professor of Prevention, Founding Director of SDRG, University of Washington School of Social Work. Theoretical architect of the Social Development Model.
• Richard F. Catalano, Ph.D. – Professor of Social Work, Co-Founder and former Director of SDRG, University of Washington. Renowned authority on adolescent positive youth development and etiology of deviant behavior.
• John A. Pollard, Ph.D. – Quantitative methodologist and psychometrician specializing in risk and protective factor measurement across child and adolescent populations.

Institutional Contact Information:
Social Development Research Group
University of Washington School of Social Work
9725 3rd Ave. NE, Suite 401
Seattle, WA 98115-2024, USA
Website: http://depts.washington.edu/sdrg/

Purpose

Adolescence is characterized by systemic transformations in neurobiology, social cognition, and ecological attachment. During the secondary school transition, primary attachment systems shift markedly from parents and familial caregivers toward the peer group. When adolescent peer networks become saturated with individuals engaging in rule-breaking, criminality, or institutional resistance, the probability that a focal adolescent will adopt and escalate similar behavioral repertoires increases exponentially. The primary purpose of the Interaction With Antisocial Peers Scale is to provide an empirical, psychometrically standardized, and operationally rigorous assessment of an adolescent’s proximity to deviant peer behavioral models.

From a clinical and epidemiological perspective, peer delinquency represents one of the single most robust and replicable proximal predictors of youth externalizing pathology, including adolescent substance use disorders, juvenile justice involvement, gang affiliation, conduct disorder, and secondary school non-completion. The scale addresses critical methodological deficiencies inherent in unstructured or single-item queries of peer influence. Rather than soliciting vague global subjective impressions (such as asking if an adolescent “hangs out with bad crowds”), the IWAP relies on concrete, behaviorally defined, and institutionally verified occurrences (such as formal suspensions, arrests, or weapons carrying) contextualized within a closed nomination matrix (the adolescent’s “four best friends”). This methodological architecture mitigates halo effects, projection biases (the false consensus effect where youth overestimate peer involvement to justify their own deviance), and cognitive ambiguity.

In applied prevention science, the scale serves three distinct programmatic functions:

• Community and School Epidemiological Needs Assessment: Under the Communities That Care (CTC) operating system funded by agencies such as the Substance Abuse and Mental Health Services Administration (SAMHSA) and the Center for Substance Abuse Prevention (CSAP), the IWAP is deployed across entire school districts or regional youth cohorts to identify elevated risk profiles across the peer domain, enabling municipal leaders to allocate targeted prevention funding.
• Selective and Indicated Clinical Screening: In juvenile justice diversions, mental health intake units, and alternative education settings, the scale quantifies an individual adolescent’s structural immersion in antisocial networks, alerting clinicians to the necessity of structural interventions such as multidimensional family therapy or peer network disruption.
• Longitudinal Program Evaluation and Outcome Monitoring: The measure functions as a sensitive pre-post and longitudinal tracking metric to evaluate the efficacy of universal interventions (e.g., school-wide positive behavioral interventions and supports) and indicated interventions (e.g., functional family therapy) aimed at attenuating peer contagion and restructuring social ties.

Psychological Construct

The core construct captured by the IWAP is peer antisocial involvement (frequently termed delinquent peer affiliation or interaction with deviant peers). Within developmental psychopathology and sociological criminology, this construct does not simply reflect passive association; it indexes a dynamic, multi-layered ecological exposure characterized by social reinforcement, normative restructuring, observational modeling, and functional opportunities for shared delinquency.

Rather than conceptualizing antisocial peer association as a monolithic trait, the scale operationalizes it across several distinct functional domains of severe adolescent maladaptation:

1. School-Level Behavioral and Institutional Disengagement

Items measuring whether friends have been “suspended from school” or have “dropped out of school” quantify peer rupture from the primary prosocial institution of middle-to-late adolescence. Peers who face disciplinary exclusion or terminate their secondary education abandon the structured, prosocial monitoring inherent in academic environments. Association with suspended or truant peers dramatically decreases academic commitment and normalizes non-compliance with formal administrative authority.

2. Criminal-Legal and High-Risk Transgressions

Items capturing peer engagement in “carrying a handgun,” “stealing or trying to steal a motor vehicle,” and “selling illegal drugs” tap into overt index offenses and dangerous criminal behavior. These behaviors signify an escalation beyond normative developmental experimentation into predatory, violent, or illicit market enterprise. Peer engagement in handgun carriage indicates both physical danger and socialization into aggressive dispute-resolution cultures. Drug distribution and auto theft require sophisticated peer coordination, instrumental planning, and illicit network embeddedness.

3. Formal Systemic Contact

The indicator querying how many friends have “been arrested” benchmarks association with peers who have crossed the threshold into the formal juvenile justice or criminal legal system. Arrest serves as an institutional marker of frequent, visible, or severe criminal activity, reflecting immersion in networks characterized by chronic anti-normative behavior.

Crucially, the scale construct measures network density of deviant affiliation. By restricting the frame of reference to the respondent’s “four best friends,” the instrument captures the proportion of an individual’s primary developmental microsystem that is compromised by antisocial conduct. A score indicating that all four best friends engage in weapon carriage or illicit drug sales reflects a profoundly different ecological reality than a score indicating isolated exposure to a single acquaintance.

Theoretical Framework

The theoretical architecture of the Interaction With Antisocial Peers Scale is deeply anchored in three convergence points of developmental, sociological, and psychological theory:

1. The Social Development Model (SDM)

Conceptualized by J. David Hawkins and Richard F. Catalano (1992), the Social Development Model synthesizes core tenets of social control theory, social learning theory, and differential association theory into an integrated, developmental-ecological paradigm. The SDM posits that individuals across developmental phases establish social bonds through opportunities for involvement, active skill-building, and consistent reinforcement. When youth experience inadequate bonding to prosocial agents (family, school, prosocial peers), or when opportunities and reinforcement are overwhelmingly mediated by antisocial actors, the adolescent develops strong behavioral commitments to an antisocial developmental pathway. Within the SDM, interaction with antisocial peers operates as the most proximal risk factor immediately preceding individual substance misuse and criminal offending, actively transmitting delinquent values, norms, and behavioral scripts.

2. Differential Association and Social Learning Theories

The scale directly draws upon Edwin Sutherland’s Differential Association Theory and its behavioral reformulation by Ronald Akers. These frameworks assert that criminal and deviant behavior is not innate, nor is it the exclusive product of personality disorganization; rather, it is learned through intimate personal groups via direct communicative interaction. This learning encompasses:

• Techniques of committing the infractions (e.g., how to conceal handguns, access illicit drug supply chains, hotwire or bypass vehicle security).
• Specific directions of motives, drives, rationalizations, and neutralizing attitudes (e.g., legitimizing school departure or institutional defiance).
• Differential reinforcement: peers provide social validation, status elevation, and reciprocal protection for antisocial behaviors, while prosocial behaviors are actively derided or punished.

3. Peer Selection Versus Socialization Models

In developmental psychopathology, extensive scholarly debate has focused on whether delinquent peer association is a consequence of assortative selection (“birds of a feather flock together,” wherein antisocial youth seek out like-minded peers; Glueck & Glueck, 1950) or socialization/contagion (wherein exposure to deviant peers causally exacerbates delinquent behaviors; Dishion, McCord, & Poulin, 1999). Modern longitudinal research demonstrates that both processes operate in a dynamic, transactional cascade (Thornberry et al., 1994). The IWAP provides an empirical metric designed to capture this nexus: regardless of whether the initial entry into the peer group was driven by selective attraction or environmental assignment, the ongoing interaction with antisocial peers triggers powerful mechanisms of deviancy training—subtle positive reinforcements, laughter, and verbal approvals during discussions of rule-breaking—which systematically amplify delinquency over developmental time.

Validity

The validity of the Interaction With Antisocial Peers Scale has been rigorously evaluated across three decades of psychometric, epidemiological, and longitudinal prevention trials conducted by SDRG and independent research groups globally.

Construct Validity

Construct validity is evidenced by the scale’s internal structural coherence and its adherence to theoretical predictions. Exploratory and confirmatory factor analyses consistently support a robust unidimensional construct representing peer antisocial behavior. Factor loadings across large, diverse adolescent samples (e.g., Arthur et al., 2002; Glaser et al., 2005) uniformly exceed .50, with most indicators falling between .60 and .84. The items form an internally consistent hierarchical scale, where less frequent behaviors (handgun carriage, vehicle theft) represent higher latent severity on an item-response continuum, while school suspension reflects higher base-rate prevalence.

Convergent Validity

The IWAP displays strong, statistically significant convergent correlations with related individual, familial, and community risk factors:

• Peer Pro-Deviant Attitudes: Strongly correlates with measures of peer approval of drug use and violence (r = .52 to .68, p < .001).
• Individual Favorable Attitudes Toward Delinquency: Strongly correlates with self-reported favorable attitudes toward rule-breaking (r = .45 to .58, p < .001).
• Low School Commitment: Significant negative correlations with school engagement, academic performance, and institutional bonding (r = −.35 to −.48, p < .001).
• Family Conflict and Poor Parental Monitoring: Moderate positive correlations with parental inconsistency, low parental monitoring, and family conflict (r = .30 to .44, p < .001).

Predictive and Longitudinal Criterion Validity

The scale possesses exemplary predictive validity regarding future externalizing criteria. Longitudinal structural equation modeling from the longitudinal Seattle Social Development Project and the Community Youth Development Study (Hawkins et al., 2008) indicates that baseline scores on the IWAP prospectively predict:

• Subsequent past-month alcohol, cannabis, and illicit drug consumption (β = .34 to .46, p < .001) over 1-, 2-, and 4-year follow-up intervals, even after controlling for baseline personal substance use.
• Adjudicated violent and property offending in juvenile and criminal courts (β = .28 to .41, p < .001).
• School dropout and chronic truancy, demonstrating an odds ratio (OR) ranging from 1.7 to 2.4 per standard deviation increase in antisocial peer affiliation.

Discriminant Validity

The IWAP demonstrates clear empirical divergence from non-externalizing psychological constructs. Correlations between the IWAP and generalized internalizing distress (such as depressive symptoms or social anxiety) are modest to low (typically r = .10 to .22), confirming that the instrument specifically measures social-behavioral delinquency rather than general adolescent affective distress. Furthermore, the scale cleanly dissociates from prosocial peer interaction metrics (e.g., friends participating in school sports, student government, or religious activities), which load onto distinct orthogonal factors (Arthur et al., 2002).

Reliability

The Interaction With Antisocial Peers Scale exhibits robust reliability estimates across varied demographic, geographic, and developmental cohorts.

Internal Consistency

Internal consistency has been extensively documented within primary statewide validation studies (e.g., Washington, Oregon, Utah, Colorado, Kansas, Pennsylvania) and international adaptations (e.g., Australia, Netherlands, Colombia):

• Arthur et al. (2002): In a normative validation sample of 72,000 students across grades 6, 8, 10, and 12, the Cronbach’s alpha (α) for the peer antisocial behavior subscale was consistently high across all grade cohorts: Grade 6 (α = .78), Grade 8 (α = .82), Grade 10 (α = .85), and Grade 12 (α = .86).
• Glaser, Van Horn, Arthur, & Hawkins (2005): Confirmed McDonald’s omega (ω) and alpha estimates exceeding .83 across diverse racial/ethnic classifications, including non-Hispanic White, African American, Hispanic/Latino, and Native American adolescents.
• Item-Total Correlations: Corrected item-total correlations for the 6 items range from .48 to .72, indicating that each item contributes substantial common variance to the underlying construct without redundancy.

Test-Retest Stability

In methodological sub-studies examining short-term temporal stability across a 2- to 4-week interval among middle and high school cohorts, the test-retest reliability coefficient was established at r = .81 to .87 (p < .001). Over extended 1-year longitudinal intervals, cross-lagged stability coefficients range between r = .48 and .62, reflecting both meaningful temporal stability and expected developmental dynamism in friendship configurations during adolescent transitions.

Factor Analysis

The structural dimensionality and measurement invariance of the scale have been rigorously scrutinized via exploratory factor analysis (EFA) and confirmatory factor analysis (CFA).

Confirmatory Factor Analytic (CFA) Fit Indices

Large-scale structural equation modeling typically specifies the 6 items loading onto a single latent factor representing Antisocial Peer Affiliation. Robust maximum likelihood (MLR) and weighted least squares mean and variance adjusted (WLSMV) estimators—recommended for ordinal categorical indicators—yield excellent global goodness-of-fit indices:

• Comparative Fit Index (CFI): .978 to .992 (exceeding the standard ≥ .95 criterion for exemplary model fit).
• Tucker-Lewis Index (TLI): .965 to .988.
• Root Mean Square Error of Approximation (RMSEA): .038 to .048 (90% confidence interval [.032, .054], well below the ≤ .06 threshold).
• Standardized Root Mean Square Residual (SRMR): .024 to .035.

Standardized Factor Loadings

Typical standardized factor loadings (λ) derived from multi-state structural models (e.g., Arthur et al., 2002; Pollard et al., 1999) illustrate high saturation on the latent construct:

• Item 1 (Suspended from school): λ = .61 to .68
• Item 2 (Carried a handgun): λ = .68 to .78
• Item 3 (Sold illegal drugs): λ = .72 to .84
• Item 4 (Stolen or tried to steal a motor vehicle): λ = .74 to .82
• Item 5 (Been arrested): λ = .76 to .85
• Item 6 (Dropped out of school): λ = .55 to .64

Measurement Invariance

Multigroup confirmatory factor analyses (MGCFA) have substantiated strict measurement invariance across key demographic subgroups:

• Configural Invariance: The single-factor architecture holds consistently across biological sex (males vs. females) and across major racial and ethnic subgroups (White, Black, Hispanic, Asian, Indigenous).
• Metric (Weak) Invariance: Constraining factor loadings across sex and racial groups produces no statistically significant deterioration in fit (ΔCFI < .010, ΔRMSEA < .015), demonstrating that the scale items assess the construct with equivalent metric precision and scaling across diverse populations.
• Scalar (Strong) Invariance: Equal item thresholds across groups permit valid, unbiased latent mean comparisons across adolescent age cohorts, sexes, and sociodemographic strata.

Instrument / Measurement Tool

• Instrument Name: Interaction With Antisocial Peers Scale (IWAP); also recognized as the “Friends’ Delinquent Behavior” or “Peer Antisocial Behavior” subscale of the Communities That Care Youth Survey (CTC-YS) / Student Survey of Risk and Protective Factors (SSRPF).
• Construct Assessed: Exposure and affiliation with delinquent and rule-violating close peers.
• Target Population: Adolescents and youth, primarily enrolled in middle and high school (Grades 6 through 12, ages 11–19).
• Administration Format: Self-report questionnaire; available in paper-and-pencil booklet, secure web-based browser, and integrated school testing environments.
• Completion Time: Approximately 2 to 3 minutes when administered as a standalone module (the full CTC survey battery requires 40 to 50 minutes).
• Number of Items: 6 items.
• Anchor Paradigm: Anchored closed-nomination framework asking respondents to evaluate their four best friends (the friends they feel closest to) over a retrospective past year (12 months) window.
• Response Scale: 5-point ordinal frequency metric:
  • 0 = None of my friends
  • 1 = 1 of my friends
  • 2 = 2 of my friends
  • 3 = 3 of my friends
  • 4 = 4 of my friends
• Scoring Methodology:
  • Continuous Mean Score: Calculate the arithmetic mean across all completed items (minimum score 0.0, maximum score 4.0). Higher scores reflect a denser concentration of antisocial exposure within the respondent’s primary friendship circle.
  • Continuous Sum Score: Alternatively, calculate the linear sum of all 6 items (ranging from 0 to 24), frequently utilized in criminological regression modeling.
  • Missing Data Threshold: In accordance with SDRG scoring protocols, a scale score is deemed valid if the respondent completes at least 4 out of the 6 items (≥ 66.7%). Mean substitution or full information maximum likelihood (FIML) is applied for missing indicators.
  • Cut-Off Scoring / Risk Categorization: In public health prevention planning (e.g., Communities That Care frameworks), raw scores are standardized against state or national normative distributions to classify adolescents into “Elevated Risk” (typically top 40th or 30th percentile depending on state epidemiological benchmarks) versus “Normal/Low Risk.”

Permissions, Fee, and Test Year

Year of Development: The initial psychometric architecture was established in the mid-1990s as part of the Student Survey of Risk and Protective Factors, culminating in formal empirical validation publications in the late 1990s and early 2000s (e.g., Pollard, Hawkins, & Arthur, 1999; Arthur et al., 2002).

Copyright and Accessibility: The scale was developed under federal grant support from the Center for Substance Abuse Prevention (CSAP), the Substance Abuse and Mental Health Services Administration (SAMHSA), and the National Institute on Drug Abuse (NIDA). As a core module of the Communities That Care Youth Survey and CSAP’s Core Measures inventory (pp. 100–102), the instrument is placed in the public domain for academic research, school-based behavioral screening, and community prevention evaluations.

Fee Structure: There are no licensing fees, royalties, or per-administration charges required for non-commercial academic research, public school implementations, or clinical prevention initiatives.

User Requirements: Users are requested to preserve the authentic item wording and anchored 5-point response formats to ensure measurement validity and comparability with national normative datasets. Commercial repackaging or proprietary digital distribution requires formal inquiry with the Social Development Research Group (SDRG) at the University of Washington School of Social Work.

References

• Akers, R. L. (1998). Social learning and social structure: A general theory of crime and deviance. Northeastern University Press.
• Arthur, M. W., Hawkins, J. D., Pollard, J. A., Catalano, R. F., & Baglioni, A. J. (2002). Measuring risk and protective factors for substance use, delinquency, and other adolescent problem behaviors: The Communities That Care Youth Survey. Evaluation Review, 26(6), 575–601. https://doi.org/10.1177/0193841X0202600601
• Center for Substance Abuse Prevention. (1998). Core Measures: Communities That Care Youth Survey. Substance Abuse and Mental Health Services Administration (SAMHSA), U.S. Department of Health and Human Services.
• Dishion, T. J., McCord, J., & Poulin, F. (1999). When interventions harm: Peer groups and problem behavior. American Psychologist, 54(9), 755–764. https://doi.org/10.1037/0003-066X.54.9.755
• Glaser, R. R., Van Horn, M. L., Arthur, M. W., & Hawkins, J. D. (2005). Measurement properties of the Communities That Care Youth Survey across demographic groups. Journal of Quantitative Criminology, 21(1), 73–102. https://doi.org/10.1007/s10940-004-1788-1
• Hawkins, J. D., & Catalano, R. F. (1992). Communities That Care: Action for drug-free communities. Jossey-Bass.
• Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin, 112(1), 64–105. https://doi.org/10.1037/0033-2909.112.1.64
• Hawkins, J. D., Catalano, R. F., Arthur, M. W., Egan, E., Brown, E. C., Abbott, R. D., & Bouchard, L. M. (2008). Testing Communities That Care: The rationale, design and behavioral baseline for a randomized field trial of a system to prevent adolescent substance use and delinquency. Prevention Science, 9(3), 178–190. https://doi.org/10.1007/s11121-008-0092-1
• Pollard, J. A., Hawkins, J. D., & Arthur, M. W. (1999). Identifying risk and protective factors for substance use and delinquency: A cross-validation study of the Communities That Care Youth Survey. Scientific World Journal, 23(1), 145–160.
• Sutherland, E. H., Cressey, D. R., & Luckenbill, D. F. (1992). Principles of criminology (11th ed.). AltaMira Press.
• Thornberry, T. P., Lizotte, A. J., Krohn, M. D., Farnworth, M., & Jang, S. J. (1994). Delinquent peers, beliefs, and delinquent behavior: A longitudinal test of reciprocal interaction. Criminology, 32(1), 47–83. https://doi.org/10.1111/j.1745-9125.1994.tb01146.x

13. Items of the Scale (Questionnaire)