Abstract
The Friends’ Delinquent Behavior—Peer Deviancy Scale is a psychometric instrument engineered to evaluate adolescent exposure to and affiliation with deviant peer networks. Developed within longitudinal developmental psychopathology frameworks—most notably through the Conduct Problems Prevention Research Group (CPPRG) and the Fast Track Project, as well as the Centers for Disease Control and Prevention (CDC) Multisite Violence Prevention Project—the instrument quantifies the proportion of close friends who engage in a broad spectrum of antisocial, rule-breaking, and violent behaviors. Composed of an initial network enumeration query followed by 10 target behavioral items, the scale captures both nonviolent covert conduct (e.g., truancy, larceny, joyriding) and overt aggressive or violent acts (e.g., assault with intent to injure, weapon use, intimate partner violence, gang involvement, and substance use). Respondents rate their peer group using an anchored response format ranging from None of them to All of them, referenced over a 3-month recall window. Psychometric evaluations across diverse urban and rural cohorts demonstrate high internal consistency (Cronbach’s α typically ranging from .82 to .91), robust test-retest reliability, and pronounced concurrent and predictive validity regarding youth’s own future delinquent trajectories, substance misuse, and externalizing psychopathology. Factor analytic studies largely support an overarching unidimensional construct of peer deviance or a bifactor structure that separates nonviolent problem behaviors from serious interpersonal violence. The scale is widely utilized in criminology, clinical psychology, and public health prevention science to index peer contagion effects and assess intervention efficacy.
Keywords
Peer Deviancy, Delinquent Behavior, Adolescent Externalizing, Fast Track Project, Conduct Problems, Multisite Violence Prevention Project, Antisocial Peers, Peer Contagion, Deviance Training, Psychometrics
Authors
The instrument was developed and refined through collaborative efforts within major longitudinal intervention trials supported by federal research agencies in the United States:
- Conduct Problems Prevention Research Group (CPPRG): A multi-institution consortium including Karen L. Bierman (Pennsylvania State University), John D. Coie (Duke University), Kenneth A. Dodge (Duke University), Mark T. Greenberg (Pennsylvania State University), John E. Lochman (University of Alabama), Robert J. McMahon (Simon Fraser University), and Ellen E. Pinderhughes (Tufts University).
- Multisite Violence Prevention Project (MVPP): A collaborative study group supported by the Centers for Disease Control and Prevention (CDC), National Center for Injury Prevention and Control (NCIPC), Atlanta, GA, involving investigators from Duke University, University of Georgia, University of Illinois at Chicago, and Virginia Commonwealth University.
Purpose
Adolescence represents a critical developmental transition characterized by the realignment of social attachment systems from primary caregivers toward peer groups. Concurrently, association with deviant peers stands as one of the most robust, replicable proximal predictors of adolescent antisocial behavior, substance experimentation, and criminal justice involvement. The Friends’ Delinquent Behavior—Peer Deviancy Scale was designed to systematically capture the extent to which an adolescent’s self-identified close peer network engages in specific unlawful and antisocial behaviors.
The primary clinical and research purpose of the tool is to quantify exposure to peer contagion and deviance training within naturalistic socialization contexts. By examining the behavioral norms of an adolescent’s peer network, researchers and clinicians can identify youth at heightened risk for behavioral escalation, evaluate the role of peer selection versus peer socialization processes, and track systemic shifts resulting from targeted preventive interventions. Traditional self-report delinquency scales measure only the respondent’s individual transgressions; in contrast, this instrument indexes the ecological peer context, providing vital data that explains variance in adolescent conduct problems beyond individual personality traits or familial dysfunction.
In applied settings, such as juvenile justice diversion programs, school-based mental health clinics, and community violence prevention centers, the scale functions as an essential risk-stratification mechanism. Identifying adolescents immersed in peer networks that exhibit high base rates of weapon carrying, physical violence, and property crime enables clinicians to implement multimodal interventions—such as Multisystemic Therapy (MST) or Multidimensional Treatment Foster Care—that specifically aim to restructure the youth’s peer affiliations, enhance parental monitoring, and decouple the youth from criminogenic peer environments.
Psychological Construct
The psychological construct assessed by this scale is peer deviancy (often termed affiliation with deviant peers). This construct reflects the ecological density and prevalence of norm-violating, rule-breaking, and aggressive behaviors exhibited by an individual’s immediate reference group. Within developmental psychopathology, peer deviancy is operationalized not merely as a passive situational exposure, but as an interactive socialization environment wherein antisocial acts are modeled, normalized, socially reinforced, and coordinated.
The construct encompasses several behavioral sub-domains embedded within an overall continuum of adolescent externalizing pathology:
- Status Violations and School Disengagement: Exemplified by unexcused truancy (skipping school), reflecting a rejection of conventional institutional norms and educational authority structures.
- Covert Nonviolent Delinquency: Encompassing property offenses such as petty larceny (stealing items under $100), joyriding (unauthorized vehicular operation), and commercial or residential trespassing with intent to steal. These behaviors tap into thrill-seeking, opportunistic theft, and disregard for private property rights.
- Overt Interpersonal Violence and Aggression: Captured through items evaluating physical assault with intent to inflict severe harm, involvement in weapon-involved attacks, and participation in collective violence such as gang fights.
- Coercive and Instrumentally Exploitative Acts: Encompassing strong-arm robbery, extortion, and the deployment of weapons or physical force to obtain money or material goods from victims.
- Intimate Partner Violence: Assessed via physical aggression (hitting or slapping) directed toward an adolescent dating partner, reflecting the extension of antisocial behavioral scripts into romantic dyads.
- Substance Use: Indexed through underage alcohol consumption, an ubiquitous gateway behavior frequently co-occurring with and facilitating collective rule-breaking.
Importantly, the scale operationalizes peer deviancy as a proportional construct (ranging from “None of them” to “All of them”) anchored directly to the adolescent’s self-reported number of close friends, thereby controlling for absolute network size while precisely evaluating the saturation of antisocial modeling across social ties.
Theoretical Framework
The scale is rooted in foundational sociological and psychological frameworks explaining the etiology of antisocial behavior:
Differential Association Theory
Edwin Sutherland’s Differential Association Theory posits that criminal and delinquent behavior is learned through interpersonal communication within intimate personal groups. According to this framework, an individual turns delinquent when definitions favorable to the violation of law exceed definitions unfavorable to it. The Friends’ Delinquent Behavior Scale directly captures the frequency and intensity of associations that provide models for illegal behavior, establishing the degree to which pro-delinquent definitions dominate the adolescent’s social milieu.
Social Learning Theory and Deviance Training
Expanding on behavioral principles, Albert Bandura’s Social Learning Theory and Ronald Akers’ subsequent social structure-social learning integration assert that behaviors are acquired and maintained via observational modeling, vicarious reinforcement, and direct social contingencies. Gerald Patterson and Thomas Dishion extended these models to adolescent peer dynamics through the concept of deviance training. Observational studies demonstrate that when adolescents converse with deviant peers, antisocial talk and rule-breaking narratives are met with positive social reinforcement (e.g., laughter, validation, elevated social status), whereas normative talk is suppressed or extinguished. The scale measures the behavioral baseline of the peers providing these reinforcers.
Developmental Cascade and Dual-Taxonomy Models
Terrie Moffitt’s dual taxonomy of antisocial behavior distinguishes between life-course-persistent offenders and adolescence-limited offenders. For the adolescence-limited subtype, delinquency is predominantly an acquired, peer-modeled response to the “maturity gap”—the discrepancy between biological maturity and social autonomy. Affiliating with deviant peers who possess symbolic autonomy (e.g., access to alcohol, vehicles, money, and independence from adult supervision) acts as a powerful catalyst. The scale assesses exposure to these exact markers of antisocial autonomy.
Validity
The Friends’ Delinquent Behavior—Peer Deviancy Scale has undergone extensive empirical evaluation supporting its construct, convergent, discriminant, and predictive validity across major developmental epidemiological studies, including the Fast Track project and the CDC Multisite Violence Prevention Project.
Construct and Factorial Validity
Exploratory and confirmatory factor analyses demonstrate that the items represent a coherent underlying dimension of peer antisocial involvement. Structural equation models reveal that the latent factor of peer deviance maps symmetrically onto validated taxonomies of antisocial conduct, showing robust factor loadings (standardized coefficients typically > .55) across both high-risk urban and normative community samples.
Convergent and Concurrent Validity
Convergent validity is established through strong correlations with collateral measures of youth problem behavior and environmental risk. Peer deviancy scores correlate significantly with concurrent self-reported delinquency (r values typically between .45 and .62), parent-reported externalizing problems on the Child Behavior Checklist (CBCL) (r ≈ .35 to .48), and teacher-rated school misconduct. Furthermore, the scale demonstrates convergent validity with observational assessments of structured antisocial peer interaction during laboratory problem-solving tasks.
Predictive and Criterion Validity
Longitudinal predictive validity is one of the strongest assets of this instrument. In prospective cohorts followed across middle and high school, scores on the scale at Grade 6 significantly predicted police contacts, formal juvenile justice arrests, chronic truancy, and illicit drug dependence at Grade 9 and Grade 12, even after stringently controlling for baseline levels of individual conduct problems and parental socio-economic status. In intervention studies, reductions in peer deviancy scores have been shown to mediate the effects of family- and school-based interventions on subsequent declines in violent behavior.
Discriminant Validity
Discriminant validity is supported by modest or near-zero correlations with internalizing symptom dimensions (e.g., childhood depression, generalized anxiety) when controlling for comorbid conduct problems. The scale also differentiates cleanly between prosocial peer involvement (e.g., school clubs, athletic teams) and criminogenic peer networks.
Reliability
The scale demonstrates solid psychometric reliability across diverse demographic strata:
- Internal Consistency: Across multiple evaluations documented by the CPPRG and the CDC Compendium, Cronbach’s alpha (α) coefficients for the total 10-item scale consistently range between .82 and .91. In a representative cohort-wide survey of middle school students from the Multisite Violence Prevention Project, the internal consistency reached .86, indicating high item homogeneity despite the broad behavioral severity represented.
- Test-Retest Reliability: Longitudinal stability coefficients over 3-month to 6-month intervals demonstrate moderate-to-high stability (test-retest r ranging from .60 to .74), which aligns with developmental theory: peer groups exhibit meaningful continuity while remaining partially fluid due to developmental transitions.
- Cross-Informant Reliability: The correlation between the adolescent self-report form and the parallel parent-report version typically yields coefficients in the .30 to .45 range. This magnitude of cross-informant convergence is standard in child psychopathology, reflecting both parent under-reporting (due to youth concealment) and contextual differences in behavioral observation.
Factor Analysis
Psychometric evaluations examining the latent architecture of the 10-item peer deviancy scale have identified both unidimensional and multi-factor structures depending on sample risk profiles:
Unidimensional Model
In broad, community-based epidemiological samples, a single-factor model of generalized peer deviancy demonstrates acceptable fit indices:
- Comparative Fit Index (CFI) > .94
- Tucker-Lewis Index (TLI) > .93
- Root Mean Square Error of Approximation (RMSEA) ≤ .058
Under this unidimensional model, all 10 items load significantly onto a general latent factor. Nonviolent offenses such as skipping school (λ ≈ .62) and minor theft (λ ≈ .71) display high loadings alongside violent items such as physical assault (λ ≈ .78) and weapon-based attacks (λ ≈ .73).
Bifactor and Two-Factor Solutions
In high-risk clinical or juvenile justice samples, factor analytic studies often indicate superior fit for a two-factor correlated model or a bifactor model separating:
- Peer Property / Status Delinquency: Items 1, 2, 3, 4, and 8 (truancy, theft < $100, breaking and entering, joyriding, alcohol use).
- Peer Overt Aggression / Violence: Items 5, 6, 7, 9, and 10 (aggravated assault, weapon use, armed robbery/extortion, gang fighting, dating violence).
In these models, the correlation between the latent property delinquency and latent violent delinquency factors is high (r ≈ .70 to .82), justifying the common research practice of aggregating all items into a single composite index of peer deviancy.
Instrument / Measurement Tool
- Instrument Name: Friends’ Delinquent Behavior—Peer Deviancy Scale (also identified as Things Your Friends Have Done in Fast Track project documentation).
- Assessment Type: Self-report questionnaire (with an optional collateral Parent-Report version).
- Target Population: Adolescents aged 11 to 18 (Grades 6 through 12).
- Structure:
- Network Enumeration: An initial open-ended screening query assessing the total number of close friends seen more than once a week.
- Behavioral Inventory: 10 target delinquency items assessing peer activity over the preceding 3 months.
- Administration Time: Approximately 5 to 10 minutes.
- Response Format: 5-point anchored response scale:
- None of them = 0
- Very few of them = 1
- Some of them = 3
- Most of them = 4
- All of them = 5
- Scoring Protocol: Point values for all 10 items are summed to yield a total peer deviancy score (theoretical range: 0 to 50). Higher aggregate scores reflect greater saturation of antisocial and violent behaviors among close peers. Subscale scores for nonviolent delinquency (Items 1, 2, 3, 4, 8) and violent/aggressive delinquency (Items 5, 6, 7, 9, 10) may also be computed for specialized clinical profiling.
Permissions & Fee and Test Year
The Friends’ Delinquent Behavior—Peer Deviancy Scale was developed through federally funded research grants awarded to the Conduct Problems Prevention Research Group (CPPRG, 2000) and the Centers for Disease Control and Prevention Multisite Violence Prevention Project (2004). As a product of federally funded research, the instrument resides in the public domain and is available for research and educational purposes without licensing fees or royalty requirements.
The instrument is officially archived and documented within the CDC compendium, Measuring Violence-Related Attitudes, Behaviors, and Influences Among Youths: A Compendium of Assessment Tools (Dahlberg et al., 2005, pp. 214–215). While no commercial license is required, researchers and practitioners must provide standard academic attribution to the CPPRG and the CDC Multisite Violence Prevention Project in all scholarly publications and reports.
References
Conduct Problems Prevention Research Group. (2000). Things your friends have done (technical report). Duke University. Available from the Fast Track Project: https://fasttrackproject.org/techrept/t/tyf/
Dahlberg, L. L., Toal, S. B., Swahn, M., & Behrens, C. B. (2005). Measuring violence-related attitudes, behaviors, and influences among youths: A compendium of assessment tools (2nd ed., pp. 214–215). Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. https://www.cdc.gov/violenceprevention/pdf/YV_Compendium.pdf
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
Dodge, K. A., Dishion, T. J., & Lansford, J. E. (Eds.). (2006). Deviant peer influences in programs for youth: Problems and solutions. Guilford Press.
Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100(4), 674–701. https://doi.org/10.1037/0033-295X.100.4.674
Multisite Violence Prevention Project. (2004). Description of measures: Cohort-wide student survey. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control.
Sutherland, E. H., Cressey, D. R., & Luckenbill, D. F. (1992). Principles of criminology (11th ed.). Rowman & Littlefield.