Peer Deviancy Scale

Abstract

The Peer Deviancy Scale (PDS), also known within longitudinal prevention science as an adaptation of the “Things That My Friends Have Done” inventory, is an established psychometric instrument engineered to quantify an adolescent’s exposure to and affiliation with deviant, antisocial, and rule-breaking peer networks. Originating in foundational developmental delinquency research at the University of Michigan and subsequently refined within the multi-site Fast Track Project by the Conduct Problems Prevention Research Group (CPPRG), the PDS captures a broad spectrum of antisocial peer behaviors. The inventory comprises 17 items that assess covert delinquent acts (e.g., property theft, shoplifting, trespassing), overt interpersonal aggression and violence (e.g., severe physical assault, weapon carrying, weapon use, intimate partner battery, gang fighting), substance use (alcohol and tobacco consumption), and school-based academic noncompliance (e.g., truancy, academic dishonesty, disruptive behavior). Responses are recorded on a 5-point ordinal frequency metric ranging from 0 (“None of them”) to 4 (“All of them”). The tool features both youth self-report and parent-informant parallel forms. Psychometric evaluations consistently demonstrate robust internal consistency, yielding Cronbach’s alpha coefficients spanning α = .84 to .88 across normative community, high-risk, and parental samples (Miller-Johnson et al., 2004). Confirmatory and exploratory factor analyses support both an overarching unidimensional peer deviancy composite and differentiated multi-factor architectures delineating serious delinquency, substance engagement, and academic infractions. The scale exhibits robust concurrent, predictive, and discriminant validity, serving as an indispensable operationalization of peer contagion, deviancy training, and differential association in developmental psychopathology, criminology, and clinical prevention trials.

Keywords

Peer Deviancy Scale, adolescent delinquency, peer contagion, deviancy training, antisocial behavior, Conduct Problems Prevention Research Group, Fast Track Project, psychometrics, externalizing disorders, differential association theory

Authors

The historical architecture of the Peer Deviancy Scale traces its roots to developmental sociology and juvenile delinquency research pioneered at the Institute for Social Research at the University of Michigan during the 1970s. The standardized operationalization and modern psychometric validation of the 17-item measure were executed by the Conduct Problems Prevention Research Group (CPPRG) as part of the landmark Fast Track longitudinal intervention study. The core CPPRG consortium comprised principal investigators:

• Karen L. Bierman, Ph.D. – Department of Psychology, The Pennsylvania State University.
• John D. Coie, Ph.D. (deceased) – Department of Psychology and Neuroscience, Duke University.
• Kenneth A. Dodge, Ph.D. – Sanford School of Public Policy, Duke University.
• Mark T. Greenberg, Ph.D. – Department of Human Development and Family Studies, The Pennsylvania State University.
• John E. Lochman, Ph.D., ABPP – Department of Psychology, The University of Alabama.
• Robert J. McMahon, Ph.D. – Department of Psychology, Simon Fraser University and BC Children’s Hospital Research Institute.
• Ellen E. Pinderhughes, Ph.D. – Eliot-Pearson Department of Child Study and Human Development, Tufts University.

Crucial psychometric norming, adaptation for multi-site community violence prevention, and dual-informant validity modeling were further established by Shari Miller-Johnson, Ph.D., Terri N. Sullivan, Ph.D., and Thomas R. Simon, Ph.D., within the Multisite Violence Prevention Study coordinated through the Centers for Disease Control and Prevention (CDC).

Purpose

Affiliation with antisocial peers represents one of the single most robust, reliable, and proximal predictors of delinquent behavior, substance misuse, school failure, and escalating conduct problems across middle childhood and adolescence. The primary objective of the Peer Deviancy Scale is to provide an empirically validated, brief, and ecologically sound measurement of the density of deviancy within an adolescent’s immediate social ecology. Rather than inquiring exclusively about an adolescent’s personal infractions—which frequently invokes substantial social desirability bias and underreporting—the PDS systematically surveys the behaviors perpetrated by the respondent’s closest social circle (“How many of your closest friends have…”). This methodological approach provides a dual window: it measures the peer group norms to which the adolescent is directly exposed and serves as an indirect proxy for the individual’s own covert externalizing involvement.

In clinical, educational, and juvenile justice settings, the PDS serves multiple critical functions:

• Risk Stratification and Triage: The tool efficiently identifies youth transitioning from childhood to adolescence who are becoming embedded in peer networks characterized by severe aggression, weapons carriage, substance use, or chronic truancy.
• Mechanism of Change Evaluation: In randomized controlled trials of family therapy (e.g., Multisystemic Therapy, Functional Family Therapy) or school-based universal and targeted interventions (e.g., the Fast Track PATHS curriculum), reducing peer deviancy is frequently a primary mediator of long-term delinquency prevention. The PDS captures treatment-induced shifts in friendship selection.
• Informant Discrepancy Analysis: Because the PDS possesses parallel forms for youth self-report and parent proxy-report, clinicians can assess parental knowledge and monitoring accuracy. Major discrepancies between parent perceptions of peer delinquency and youth self-reports highlight deficits in parental monitoring or active concealment by the youth.

Psychological Construct

The overarching psychological construct measured by the Peer Deviancy Scale is Antisocial Peer Affiliation, conceptualized as the degree to which an individual’s primary friendship network models, reinforces, and engages in norms that violate societal laws, institutional rules, and interpersonal rights. Psychometrically and theoretically, this construct is multidimensional, encapsulating distinct topographies of rule-breaking behavior:

1. Covert Delinquency and Property Violations

Covert delinquency encompasses clandestine antisocial actions aimed at property acquisition or destruction without direct, face-to-face confrontation with victims. Items assess property theft from private domains (lockers, desks, purses, homes), retail larceny (shoplifting), motor vehicle theft or joyriding, unlawful entry (breaking into buildings, cars, or houses to steal), and intentional vandalism or property destruction across school, commercial, and residential environments. These behaviors indicate an established departure from societal legal codes and reflect shared covert group activities that reinforce instrumental rule violation.

2. Overt Interpersonal Aggression, Violence, and Weapon Involvement

Overt aggression involves direct physical harm, threat of force, and lethal risk behaviors within interpersonal encounters. The PDS captures severe physical assaults (“Hit someone really badly”), intimate partner violence (“Hit or slapped a boyfriend/girlfriend really badly”), aggravated assault with weapons or physical force, predatory armed extortion (“Used a weapon or force to get money or things from another person”), carrying concealed weapons (knives, guns) onto academic campuses, and organized collective violence (gang fighting). Endorsement of these items signifies embeddedness within high-risk, violent subcultures where physical dominance and coercive intimidation operate as normalized behavioral scripts.

3. Substance Engagement

This dimension measures engagement in developmentally proscribed licit and illicit substance consumption, specifically alcohol consumption and tobacco/cigarette smoking. Peer substance use during early and middle adolescence is rarely an isolated phenomenon; it predominantly functions as a shared social ritual that facilitates group cohesion, lowers behavioral inhibition, and accelerates cross-tolerance for further deviant experimentation.

4. Academic Noncompliance and Institutional Disengagement

The academic domain captures defiance toward formal educational structures. Items measure unexcused school absenteeism (truancy), institutional deception (cheating on examinations or homework, lying to teachers to evade disciplinary consequences), and active behavioral disruption within the classroom environment. This facet reflects an alienation from prosocial institutional expectations and often constitutes the earliest observable stage of the developmental cascade toward chronic delinquency.

Theoretical Framework

The architecture of the Peer Deviancy Scale is grounded in prominent psychosocial theories of crime, developmental psychopathology, and social ecology:

Differential Association Theory

Formulated by Edwin Sutherland (1947), Differential Association Theory posits that criminal and deviant behaviors are learned through intimate personal group interactions. An individual turns toward delinquency when exposed to an excess of definitions favorable to violation of law over definitions unfavorable to violation of law. The PDS directly operationalizes Sutherland’s core tenets: by measuring the proportion of an individual’s closest friends exhibiting deviancy, the scale quantifies the frequency, priority, duration, and intensity of deviant normative models within the respondent’s developmental niche.

Social Learning Theory and Deviancy Training

Albert Bandura’s (1977) Social Learning Theory, expanded through the coercion and peer interaction models of Gerald Patterson and Thomas Dishion (Dishion et al., 1996), explains the micro-mechanisms captured by the PDS. Dishion’s construct of deviancy training describes contingent positive social reactions (e.g., laughter, verbal encouragement, peer status elevation) provided by antisocial peers during rule-breaking discourse. Over time, adolescents immersed in such dyads or cliques undergo intensive behavioral shaping that normalizes covert theft, violent confrontation, and institutional defiance. The PDS serves as a macroscopic survey of environments rich in deviancy training opportunities.

Interactional Theory

Terence Thornberry’s (1987) Interactional Theory merges social control theory and social learning theory into a reciprocal, life-course developmental model. Thornberry argues that the relationship between peer deviancy and individual conduct problems is bidirectional: while weak attachment to prosocial institutions (schools, family) facilitates selection into deviant peer groups, the deviant peer network actively amplifies and solidifies individual delinquent engagement over time. The PDS captures this dynamic nexus across the childhood-to-adolescent developmental transition.

Validity

The Peer Deviancy Scale has undergone extensive empirical evaluation demonstrating exemplary psychometric validity across community, clinic-referred, and high-risk urban and rural cohorts.

Construct and Convergent Validity

Construct validity is substantiated by robust correlations with established measures of adolescent externalizing pathology. Scores on the PDS correlate significantly and positively with the Externalizing Broadband Scale, Rule-Breaking Behavior Subscale, and Aggressive Behavior Subscale of the Achenbach Child Behavior Checklist (CBCL) and Youth Self-Report (YSR), with Pearson product-moment correlations typically ranging between r = .45 and r = .62 (p < .001). Furthermore, the PDS correlates strongly with measures of low self-control, sensation seeking, and impulsivity (e.g., Barratt Impulsiveness Scale), while demonstrating moderate inverse associations with measures of prosocial academic attachment and parental monitoring.

Predictive and Criterion Validity

Longitudinal investigations conducted via the Fast Track Project and the Multisite Violence Prevention Study (Miller-Johnson et al., 2004) establish the extraordinary predictive utility of the PDS. Baseline scores on the 10-item core index and 17-item full inventory longitudinally predict future court adjudications, formal juvenile arrest records, violent community altercations, illicit drug consumption, and high school dropout rates across multi-year follow-up intervals, even after statistically controlling for prior individual aggression and socioeconomic disadvantage. Longitudinal structural equation modeling indicates that elevations in peer deviancy mediated the intervention treatment effects observed in the Fast Track randomized trial.

Discriminant Validity

The scale demonstrates adequate discriminant separation from internalizing symptoms (e.g., generalized anxiety, major depressive symptomatology, somatic complaints). Correlations between the PDS and the CBCL Internalizing Broadband scale remain modest (typically r < .15), affirming that the inventory specifically indexes behavioral noncompliance and delinquent ecology rather than generalized psychological distress.

Reliability

The psychometric stability and precision of the Peer Deviancy Scale have been replicated across diverse ethnic, geographic, and sociodemographic strata.

Internal Consistency

The internal consistency of the PDS items is consistently high across independent validation cohorts. In the comprehensive psychometric evaluation conducted by Miller-Johnson et al. (2004) within the Multisite Violence Prevention Study:

• Random/Normative Sample: Cronbach’s alpha was α = .85 for youth self-report scores.
• High-Risk Targeted Sample: Cronbach’s alpha reached α = .88, indicating outstanding reliability in populations with elevated base rates of aggressive and delinquent behavior.
• Parent Informant Sample: Cronbach’s alpha was α = .84, confirming that parent-proxy evaluations of their child’s peer group maintain strong internal consistency.

Test-Retest Stability

In prospective assessments over 6-month and 12-month retest intervals during middle school, the PDS exhibits moderate-to-high rank-order stability coefficients ranging from r = .60 to r = .72. Because friendship configurations naturally shift during adolescent developmental transitions, these coefficients reflect both the expected stability of chronic antisocial trajectory involvement and the dynamic, reconfigurable nature of adolescent friendship cliques.

Inter-Rater Reliability

Concordance between youth self-report and parent proxy-report is typically moderate, with Pearson correlation coefficients spanning r = .30 to r = .46. This level of correspondence is consistent with cross-informant meta-analytic trends in adolescent assessment, where youth possess direct, unobserved knowledge of peer actions that are systematically masked from parental oversight.

Factor Analysis

Structural evaluations of the Peer Deviancy Scale using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) reveal an overarching general deviancy factor alongside well-defined thematic sub-dimensions.

Exploratory Factor Structure

Principal Axis Factoring and Maximum Likelihood exploratory models utilizing oblique rotation (Promax or Oblimin) consistently extract a dominant first factor accounting for upwards of 40% to 50% of the total variance. Factor loadings for items measuring overt physical aggression, weapon usage, and serious property theft typically exceed .60, establishing a potent core delinquent construct. Secondary factor extractions isolate:

1. Severe Delinquency and Violent Conduct (Items 1–10, 13)
2. Substance Experimentation (Items 11–12)
3. Academic Noncompliance and School Misbehavior (Items 14–17)

Confirmatory Factor Analysis and Model Fit

In rigorous CFA testing across multi-site adolescent cohorts, both a higher-order hierarchical model (a general Peer Deviancy second-order factor overarching three first-order factors: Delinquency/Violence, Substance Use, Academic Noncompliance) and a bi-factor structure demonstrate superior fit to empirical data compared to a strictly unidimensional 1-factor model. Standard model fit indices routinely meet or exceed conventional psychometric benchmarks:

• Comparative Fit Index (CFI): ≥ .94 to .96
• Tucker-Lewis Index (TLI): ≥ .93 to .95
• Root Mean Square Error of Approximation (RMSEA): ≤ .045 to .058 (90% Confidence Interval: [.038, .064])
• Standardized Root Mean Square Residual (SRMR): ≤ .042

Standardized item-factor loadings (λ) across target subscales are robust, with covert and overt delinquent items demonstrating loadings between .55 and .82, and academic noncompliance items loading between .58 and .76.

Instrument / Measurement Tool

The technical parameters and administrative architecture of the Peer Deviancy Scale are structured as follows:

• Instrument Type: Standardized behavioral rating scale; dual-informant questionnaire (Youth Self-Report and Parent Proxy-Report forms).
• Target Population: Late childhood, early adolescence, and mid-to-late adolescence (approximate age range: 10 to 18 years).
• Administration Format: Paper-and-pencil questionnaire, clinician- or researcher-administered structured interview, or secure computer-assisted self-interviewing (CASI).
• Total Item Count: 17 items total. The primary inventory features 13 items covering community, property, violent, and substance behaviors, supplemented by 4 items explicitly assessing academic-related infractions.
• Prompt Question:
  • Youth Version: “How many of your closest friends have:”
  • Parent Version: “How many of your son’s/daughter’s best friends have:”
• Response Format: 5-point ordinal Likert scale scored from 0 to 4:
  • 0 = None of them
  • 1 = Very few of them
  • 2 = Some of them
  • 3 = Most of them
  • 4 = All of them
• Scoring Protocols:
  • 10-Item Core Scale: In several major epidemiological and intervention trials (including specific Fast Track Project assessment modules), a primary composite index is computed by calculating the arithmetic mean of items 2 through 11 (spanning retail theft, joyriding, breaking and entering, vandalism, assault, domestic battery, weapon force, extortion, weapon possession at school, and alcohol consumption).
  • Full 17-Item Total Deviancy Score: Calculated as either the arithmetic mean or the summative total score across all 17 items (range: 0 to 68). Higher scores reflect greater saturation of antisocial behaviors within the friendship network.
  • Subscale Scores: Derived by averaging or summing items within specific theoretical clusters: Property Delinquency (Items 1–5), Violent Aggression/Weapons (Items 6–10, 13), Substance Use (Items 11–12), and Academic-Related Behaviors (Items 14–17).
• Completion Time: Approximately 5 to 8 minutes.

Permissions & Fee and Test Year

The foundational items of the Peer Deviancy Scale were operationalized in developmental sociological studies at the University of Michigan during the 1970s and subsequently adapted into the “Things That My Friends Have Done” instrument by the Conduct Problems Prevention Research Group (CPPRG) in 1998 for the Fast Track Project. Methodological validation for multi-site violence prevention initiatives was formally published by Miller-Johnson, Sullivan, and Simon in 2004.

Licensing and Fee Structure: The instrument is maintained within the public academic domain for non-commercial research, educational, and clinical evaluation purposes. No user licensing fees or per-administration royalties are required. Researchers and practitioners may access the measure through published compendia, including technical documentation from the Fast Track Project Assessment Battery and the Pan American Health Organization (PAHO) compendium, Latino Families and Youth: A Compendium of Assessment Tools (pages 94–96). When utilizing the scale, researchers are obligated to maintain standard academic attribution by citing the original instrument developers (CPPRG, 1998) and relevant psychometric validation literature (Miller-Johnson et al., 2004).

References

• Bandura, A. (1977). Social learning theory. Prentice-Hall.
• Chadwick, B. A., Top, B. L., & McClendon, R. J. (2010). Shield of faith: The power of religion in the lives of LDS youth and young adults. BYU Religious Studies Center. https://rsc.byu.edu/archived/shield-faith-power-religion-lives-lds-youth-and-young-adults
• Conduct Problems Prevention Research Group. (1998). Technical reports for the Fast Track assessment battery (Unpublished technical report). Conduct Problems Prevention Research Group. https://fasttrackproject.org/
• Dishion, T. J., Spracklen, K. M., Andrews, D. W., & Patterson, G. R. (1996). Deviancy training in male adolescent friendships. Behavior Therapy, 27(3), 371–383. https://doi.org/10.1016/S0005-7894(96)80023-2
• Miller-Johnson, S., Sullivan, T. N., & Simon, T. R. (2004). Evaluating the impact of interventions in the Multisite Violence Prevention Study: Samples, procedures, and measures. American Journal of Preventive Medicine, 26(1, Suppl.), 48–61. https://doi.org/10.1016/j.amepre.2003.09.020
• Sutherland, E. H. (1947). Principles of criminology (4th ed.). J. B. Lippincott.
• Thornberry, T. P. (1987). Toward an interactional theory of delinquency. Criminology, 25(4), 863–892. https://doi.org/10.1111/j.1745-9125.1987.tb00823.x

13. Items of the Scale (Questionnaire)