Adolescent Defending Behaviors Questionnaire (ADBQ)

Abstract

The Adolescent Defending Behaviors Questionnaire (ADBQ) is an empirically validated psychometric instrument developed by Wang, Laninga-Wijnen, Garandeau, and Liu (2023) to measure the multidimensional spectrum of bystander intervention behaviors exhibited by youth in school bullying scenarios. Recognizing that peer defending is not a monolithic construct, the ADBQ differentiates between varied behavioral strategies that adolescents employ to alleviate distress, confront perpetrators, or de-escalate interpersonal violence. The instrument assesses five distinct subscales: Assertive Defending, Aggressive Defending, Comforting Victims, Reporting to Authority, and Tactical Defending. Administered via a 5-point Likert-type response format ranging from 1 (strongly disagree) to 5 (strongly agree), the finalized questionnaire comprises 24 items derived from cross-cultural item pool generation, qualitative semi-structured interviews, bilingual translation and back-translation procedures, and rigorous psychometric refinement.

Psychometric evaluation across early adolescent cohorts demonstrated robust construct validity and structural stability. Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) supported an optimal five-factor solution accounting for 56.1% of the cumulative variance, outperforming alternative structural configurations. Goodness-of-fit indices demonstrated strong structural alignment (χ²(166) = 817.21, p < .001; CFI = .961; TLI = .936; RMSEA = .060, 90% CI [.056, .064]). Internal consistency reliability was exemplary across factors, with Cronbach’s α coefficients ranging from .78 to .93. Longitudinal test–retest correlations affirmed adequate temporal stability across intervals (r = .25 to .40, p < .01). Furthermore, multigroup invariance testing verified that the ADBQ satisfies configural, metric, and scalar measurement invariance across biological sexes. By delineating constructive, aggressive, covert, and indirect forms of peer intervention, the ADBQ addresses critical methodological gaps in developmental psychometrics, peer-relations research, and anti-bullying program evaluation.

Keywords

Adolescent Defending Behaviors Questionnaire, ADBQ, School Bullying, Bystander Intervention, Peer Defending, Assertive Defending, Aggressive Defending, Tactical Defending, Victim Consolation, Measurement Invariance

Authors

The Adolescent Defending Behaviors Questionnaire was conceptualized, designed, and psychometrically validated by an international research team specializing in developmental psychopathology, school psychology, and peer dynamics:

• Zixuan Wang — School of Psychology and Cognitive Science, East China Normal University, Shanghai, China.
• Lydia Laninga-Wijnen — Department of Psychology and INVEST Flagship Research Center, University of Turku, Turku, Finland.
• Claire F. Garandeau — Department of Psychology and INVEST Flagship Research Center, University of Turku, Turku, Finland (ORCID: 0000-0002-0395-020X).
• Junsheng Liu — School of Psychology and Cognitive Science, Affiliated Mental Health Center (ECNU), Shanghai Key Laboratory of Mental Health and Psychological Crisis Intervention, East China Normal University, Shanghai, China (ORCID: 0000-0002-6823-0617; Correspondence Email: [email protected]).

Purpose

School bullying represents a widespread public health crisis characterized by repeated, intentional harm embedded within asymmetric power dynamics among children and adolescents. Historically, developmental research conceptualized bullying primarily through a dyadic lens focused on the bully and the victim. However, pioneering research by Christina Salmivalli demonstrated that bullying is an inherently group process wherein bystanders reinforce, assist, passively observe, or actively defend victimized peers. Early measurement traditions relied heavily on unidimensional or global conceptualizations of “defending,” treating any intervention on behalf of a victim as a homogeneous, prosocial construct. This broad framing masked substantial behavioral heterogeneity, failing to differentiate between constructive problem-solving, aggressive escalation, covert interventions, and adult-oriented reporting.

The primary purpose of the Adolescent Defending Behaviors Questionnaire (ADBQ) is to address these empirical and theoretical limitations by providing a nuanced, multidimensional operationalization of peer defending. Defending behaviors in peer contexts vary markedly in their behavioral execution, social consequences, and psychological underpinnings. While directly confronting a perpetrator may successfully halt an assault, it can escalate physical danger if delivered aggressively. Conversely, comforting a victim behind closed doors or reporting incidents to school authorities provides support without direct interpersonal conflict, while tactical distractions allow bystanders to intervene with minimal social risk. Without an assessment instrument capable of isolating these discrete dimensions, researchers and school psychologists could not evaluate which defending strategies reduce victim distress versus which inadvertently amplify peer victimization.

In clinical and educational settings, the ADBQ fulfills critical applied functions. First, it enables school psychologists, behavioral intervention teams, and educational researchers to conduct granular diagnostic screenings of peer dynamics within classrooms. By understanding which defending profiles predominate in a school ecology, educators can tailor anti-bullying interventions, such as the KiVa antibullying program or Positive Behavioral Interventions and Supports (PBIS), to cultivate constructive defending while actively discouraging counter-aggression. Second, the ADBQ serves as an evaluation metric in intervention studies, measuring longitudinal shifts from passive bystander disengagement toward prosocial mediation, victim solace, and strategic de-escalation. By isolating the distinct costs and benefits associated with different defending modalities, the ADBQ equips researchers with the empirical precision required to unpack the complex social ecology of adolescent peer relationships.

Psychological Construct

The psychological construct underlying the ADBQ is grounded in the operational definition of peer defending as behavioral actions performed by bystanders intended to protect, support, or intervene on behalf of peers targeted by bullying. Rather than treating defending as a singular prosocial trait, the ADBQ articulates a five-dimensional behavioral construct reflecting diverse cognitive, emotional, and tactical mechanisms:

1. Assertive Defending

Assertive Defending encompasses direct, proactive, non-violent, and normative verbal or physical actions targeted at halting the perpetrator. This dimension reflects high social agency, emotional assertiveness, and moral courage. Adolescents engaging in assertive defending directly communicate boundary limits to the bully (e.g., “I said to the bully that bullying is wrong” or “I told the bully to stop bullying by saying ‘stop it’”) or physically step into the situation to chase the perpetrator away without engaging in physical retribution. Assertive defending relies on robust socio-cognitive capabilities, interpersonal competence, and high perceived social status, enabling the defender to disrupt harassment without triggering escalating spirals of hostility.

2. Aggressive Defending

Aggressive Defending captures intervention behaviors that employ proactive or reactive aggressive mechanisms—spanning physical violence, verbal abuse, and relational sabotage—to stop the bully or retaliate on behalf of the victim. Items measuring this construct capture physical assault (e.g., pushing, kicking, or hitting the perpetrator), verbal counter-attacks (e.g., calling the bully offensive names), and relational manipulation (e.g., spreading rumors, gossiping, or socially excluding the bully from peer activities). Although motivated by the intent to protect a peer or punish a transgressor, aggressive defending operates through hostile attribution, deficient emotional regulation, and retaliatory aggression. Research demonstrates that while aggressive defending may terminate an immediate episode, it frequently reproduces systemic cycles of classroom violence, increases retaliatory victimization, and diminishes long-term peer preference.

3. Comforting Victims

Comforting Victims (also termed victim consolation or socio-emotional support) captures indirect, relationally oriented prosocial behaviors that occur alongside or in the aftermath of a bullying event. Rather than confronting the aggressor directly, the defender focuses cognitive and affective resources entirely on mitigating the victim’s emotional suffering. Specific behaviors include expressing positive affect, active listening, spending time together during breaks, forging intentional friendships, and attempting to cheer the victim up (e.g., “I stayed with the victim during the breaks”; “I was nice to the victim”). This dimension is profoundly tied to affective empathy, perspective-taking, and compassionate altruism. It provides critical psychological buffering against loneliness, depressive symptomology, and acute feelings of alienation.

4. Reporting to Authority

Reporting to Authority represents an institutionalized, help-seeking defending strategy wherein bystanders bridge peer interactions and adult intervention. By notifying responsible institutional figures—predominantly classroom teachers, school administrators, or parents—adolescents delegate enforcement to individuals possessing formal systemic authority (e.g., “I went to tell the teacher about the bullying”; “I told a parent about the bullying”). In early adolescence, reporting requires navigating complex peer group norms, such as overcoming pejorative “snitching” or “tattling” taboos. Consequently, this dimension correlates strongly with perceived adult support, institutional trust, and an adolescent’s adherence to conventional moral frameworks.

5. Tactical Defending

Tactical Defending refers to sophisticated, indirect, and strategically non-confrontational intervention maneuvers that mitigate bullying through cognitive redirection, humor, deception, or delayed problem-solving. Items reflect strategic distraction (e.g., diverting the bully’s focus to an unrelated task), mock alarm (e.g., pretending to shout “here comes the teacher!” to scare the aggressor), temporary appeasement (e.g., feigning compliance with the bully in the moment to resolve the conflict later), and peer mediation. Tactical defending represents a cognitively mature strategy employed by youth who recognize the acute risks of direct confrontation—such as targeted retaliation or bodily harm—and deliberately choose low-visibility, high-efficiency de-escalation tactics that diffuse power differentials while maintaining personal safety.

Theoretical Framework

The ADBQ is anchored in several prominent psychological paradigms that explain social interactions, prosocial intervention, and moral development during childhood and adolescence.

1. The Participant Role Approach

The foundational framework of the ADBQ is the Participant Role Approach pioneered by Christina Salmivalli and colleagues (1996). Salmivalli posited that school bullying cannot be understood solely as an isolated dyad between a perpetrator and a target; rather, it represents a structural group phenomenon maintained by the social architecture of the peer group. Within this ecology, peers adopt functional participant roles: assistants, reinforcers, passive outsiders, and defenders. While initial participant role inventories operationalized defending as a singular behavioural orientation, social evolutionary perspectives and subsequent adolescent research revealed that defending comprises an array of diverse behavioral responses characterized by distinct social goals, peer statuses, and risk profiles. The ADBQ extends the Participant Role Approach by decomposing the defender role into five specific action pathways, differentiating between direct confrontation, indirect consolation, institutional escalation, aggressive retaliation, and covert strategic disruption.

2. The Bystander Intervention Model

The cognitive-behavioral progression of defending is deeply aligned with the classic Bystander Intervention Model formulated by Bibb Latané and John Darley (1970). Applied to school harassment, bystander action requires navigating five sequential decision stages: (1) noticing the ambiguous or covert peer event, (2) interpreting the incident as unjust bullying rather than playful teasing, (3) assuming personal responsibility rather than succumbing to the diffusion of responsibility, (4) selecting an appropriate behavioral intervention, and (5) executing the chosen behavior. The ADBQ directly assesses the fourth and fifth stages of this cognitive decision-making sequence. When an adolescent assumes personal responsibility, their selection of a defending modality depends heavily on subjective cost-benefit appraisals, perceived peer norms, and their unique interpersonal skills.

3. Social Information Processing and Social Cognitive Theory

The behavioral divergence between assertive, aggressive, comforting, reporting, and tactical defending can also be understood through Social Information Processing (SIP) Theory (Crick & Dodge, 1994) and Social Cognitive Theory (Bandura, 1986). According to SIP theory, when adolescents witness peer abuse, their cognitive databases process internal social goals (e.g., maintaining relational intimacy, demonstrating dominance, ensuring physical safety, or upholding moral integrity). Defenders who prioritize dominance and retaliatory justice, or who exhibit hostile attribution biases, often select aggressive defending. In contrast, youth endowed with advanced Theory of Mind (ToM), social problem-solving capabilities, and emotional self-regulation select assertive or tactical defending, or direct their energy toward victim comfort. Furthermore, Bandura’s concepts of defending self-efficacy and moral disengagement dictate whether bystanders feel capable of confronting high-status bullies directly or whether they resort to lower-risk strategies such as reporting to adults or comforting the victim privately.

Validity

The psychometric validation of the ADBQ was established across two rigorous empirical studies involving Chinese early adolescents in Grades 4 to 6 (Wang et al., 2023). The validation protocol systematically evaluated construct validity, content validity, convergent validity, and discriminant validity against established sociometric and psychological markers.

Content and Face Validity

The item generation pipeline was grounded in comprehensive content validation. The authors synthesized existing Western and international defending inventories, translated the preliminary items into Chinese, conducted blind back-translations with bilingual developmental psychologists, and supplemented this theoretical pool with insights gathered from in-depth, semi-structured interviews with early adolescents. An expert multidisciplinary panel comprising developmental researchers, clinical psychometricians, and school educators scrutinized the combined pool of 34 items to confirm cultural relevance, developmental suitability, clarity of phrasing, and absence of semantic ambiguity. Psychometric item pruning resulted in the finalized 24-item, 5-factor instrument.

Construct, Convergent, and Discriminant Validity

Construct validity was demonstrated by examining differential associations between the five ADBQ dimensions and key psychological constructs:

• Cognitive and Affective Empathy: Comforting Victims exhibited the strongest positive correlation with both affective and cognitive empathy, confirming that socio-emotional consolation is heavily driven by empathic resonance and perspective-taking. Assertive Defending demonstrated moderate-to-strong positive associations with cognitive empathy, reflecting the role of social understanding in direct prosocial confrontation. Conversely, Aggressive Defending exhibited nonsignificant or negative associations with affective empathy, underscoring that retaliatory intervention is functionally detached from pure prosocial compassion.
• Aggression Subtypes: Aggressive Defending demonstrated pronounced, statistically significant positive correlations with both proactive aggression and reactive aggression. In contrast, Assertive Defending, Comforting Victims, and Reporting to Authority exhibited negligible or negative associations with aggressive metrics, confirming robust discriminant validity between prosocial defending and counter-aggressive intervention.
• Sociometric Peer Status (Peer Preference vs. Popularity): The five dimensions diverged markedly across sociometric indices. Assertive Defending and Comforting Victims were positively correlated with peer acceptance, sociability, and high peer preference (being well-liked by classmates). In contrast, Aggressive Defending was associated with mixed peer preference and correlated primarily with perceived popularity (social visibility and dominant social power), mirroring the complex peer mechanics of aggressive youth.
• Victimization Experiences: Longitudinal and concurrent analyses revealed that students engaging predominantly in Tactical Defending or Reporting to Authority frequently reported prior or concurrent victimization experiences, suggesting that youth who have experienced victimization personally often select lower-risk, indirect strategies to protect others without exposing themselves to renewed physical harm.

Reliability

The ADBQ demonstrates strong psychometric reliability across internal consistency metrics and temporal stability indices across multiple samples of early adolescents.

Internal Consistency Reliability

Internal consistency was evaluated using Cronbach’s alpha (α) across each of the five subscales. Across development and cross-validation cohorts, all five factors demonstrated high to exemplary internal consistency, exceeding accepted psychometric thresholds for scale reliability (α ≥ .70):

• Assertive Defending: Cronbach’s α values ranged from .85 to .91, reflecting high homogeneity in items measuring direct non-violent interventions.
• Aggressive Defending: Cronbach’s α values ranged from .82 to .88, confirming stable internal structure across physical, verbal, and relational aggression items.
• Comforting Victims: Cronbach’s α values ranged from .88 to .93, representing the highest internal consistency among the subscales.
• Reporting to Authority: Cronbach’s α values ranged from .78 to .84, demonstrating reliable measurement of adult-oriented reporting.
• Tactical Defending: Cronbach’s α values ranged from .78 to .85, reflecting solid internal coherence across diverse indirect and strategic maneuvers.

Test–Retest Reliability

Temporal stability was evaluated across two longitudinal measurement occasions (Time 1 and Time 2) spaced over several months. Pearson correlation coefficients (r) confirmed that defending behaviors represent moderately stable behavioral patterns across early adolescence, with all correlations achieving statistical significance at the p < .01 level:

• Assertive Defending: r = .32 (p < .01)
• Aggressive Defending: r = .25 (p < .01)
• Comforting Victims: r = .40 (p < .01)
• Reporting to Authority: r = .35 (p < .01)
• Tactical Defending: r = .30 (p < .01)

These moderate test–retest coefficients reflect both the behavioral stability of personal dispositions (e.g., empathy, assertiveness) and the dynamic, context-dependent nature of classroom bystander responses, which naturally fluctuate based on shifts in peer group hierarchy, seating arrangements, and teacher supervision.

Factor Analysis

The structural dimensionality of the ADBQ was analyzed through split-sample Exploratory Factor Analysis (EFA) followed by rigorous Confirmatory Factor Analysis (CFA) across two developmental cohorts of Chinese school children (Wang et al., 2023).

Exploratory Factor Analysis (EFA)

During initial structural testing, both a five-factor structure and a six-factor structure were examined using robust maximum likelihood estimation with geomin oblique rotation. Initial EFA results across the item pool indicated that both configurations fit the empirical data well:

• Five-Factor Model (34 candidate items): χ²(401) = 963.33, p < .001; CFI = .949; TLI = .928; RMSEA = .041 (90% CI [.038, .045]).
• Six-Factor Model: χ²(372) = 935.59, p < .001; CFI = .948; TLI = .925; RMSEA = .043 (90% CI [.039, .046]).

The six-factor model yielded negligible, statistically nonsignificant improvements over the five-factor solution, as indexed by a decrease in Comparative Fit Index (ΔCFI < .010), a decrease in Tucker-Lewis Index (ΔTLI < .010), and a slight increase in Root Mean Square Error of Approximation (ΔRMSEA < .015). Furthermore, the sixth factor comprised weak, cross-loading items lacking conceptual coherence. Following item reduction based on factor loadings (> .40), cross-loading thresholds (< .25 difference across factors), and conceptual clarity, a 24-item five-factor model was derived. This model accounted for 56.1% of the cumulative variance.

Confirmatory Factor Analysis (CFA)

The 24-item five-factor structure was confirmed in an independent adolescent validation sample. The model exhibited excellent goodness-of-fit indices:

• Chi-Square: χ²(166) = 817.21, p < .001
• Comparative Fit Index (CFI): .961
• Tucker-Lewis Index (TLI): .936
• Root Mean Square Error of Approximation (RMSEA): .060 (90% CI [.056, .064])
• Standardized Factor Loadings: All retained items exhibited strong, statistically significant standardized factor loadings on their designated latent factors, ranging from .58 to .89 (all p < .001).

Measurement Invariance

To verify that the ADBQ measures defending constructs equivalently across biological sexes, multigroup confirmatory factor analyses tested sequential tiers of measurement invariance:

• Configural Invariance: Unconstrained multi-group models demonstrated adequate fit, verifying that the five latent dimensions possess identical conceptual structures across boys and girls.
• Metric (Weak) Invariance: Constraining item factor loadings to equality across sexes produced nonsignificant changes in fit (ΔCFI ≤ .010, ΔRMSEA ≤ .015), establishing that respondents interpret item meanings and intervals equivalently across genders.
• Scalar (Strong) Invariance: Constraining item intercepts to equality across genders demonstrated sustained model fit (ΔCFI ≤ .010), confirming that observed mean differences between boys and girls reflect true differences in latent defending traits rather than measurement bias.

Instrument / Measurement Tool

• Test Type: Original psychometric self-report questionnaire
• Construct Measured: Multidimensional adolescent defending behaviors in school bullying situations
• Subscales / Dimensions (5 Factors):
  • Assertive Defending (direct verbal/physical non-violent confrontation)
  • Aggressive Defending (physical, verbal, and relational retaliatory aggression)
  • Comforting Victims (socio-emotional solace, companionship, and emotional support)
  • Reporting to Authority (informing teachers, school staff, or parents)
  • Tactical Defending (cognitive distraction, strategic de-escalation, mediation, and ruses)
• Target Population: Children and early adolescents; validated among students in Grades 4 to 6 (ages 9–13 years)
• Available Languages: Chinese (original validation) and English
• Item Count: 24 items in the validated full inventory (19 core published operational items available in primary public instruments)
• Response Scale: 5-point Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree)
• Scoring Procedures:
  • Each subscale score is computed by calculating the arithmetic mean of its corresponding items.
  • Higher subscale scores denote greater frequency or propensity to execute that specific defending strategy.
  • Because the five dimensions reflect conceptually divergent strategies, researchers are advised against aggregating items into a single composite score; subscale scores should be maintained independently or modeled as distinct latent variables in structural equation models.
• Administration Time: Approximately 10 to 15 minutes in group classroom settings or individual diagnostic evaluations.

Permissions & Fee and Test Year

• Test Publication Year: 2023
• Copyright & Ownership: © 2023 Zixuan Wang, Lydia Laninga-Wijnen, Claire F. Garandeau, and Junsheng Liu. Published by SAGE Publications on behalf of the American Psychological Association / Assessment division.
• Permissions & Usage Rights: The ADBQ is an open-access psychometric instrument permitted for academic research and non-profit educational/teaching purposes without royalty fees.
• Commercial Licensing: Use within proprietary commercial assessment platforms, private for-profit educational technologies, or commercial workplace assessments requires explicit written permission from the corresponding author and copyright holders.
• Contact for Permissions & Original Materials: Prof. Junsheng Liu, East China Normal University, School of Psychology and Cognitive Science, 3663 Zhongshan Road (North), Shanghai, China, 200062 (Email: [email protected]).

References

• Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-Hall.
• 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
• Latané, B., & Darley, J. M. (1970). The unresponsive bystander: Why doesn’t he help? Appleton-Century-Crofts.
• Salmivalli, C., Lagerspetz, K., Björkqvist, K., Österman, K., & Kaukiainen, A. (1996). Bullying as a group process: Participant roles and their relations to social status in the group. Aggressive Behavior, 22(1), 1–15. https://doi.org/10.1002/(SICI)1098-2337(1996)22:1<1::AID-AB1>3.0.CO;2-T
• Salmivalli, C., Voeten, M., & Poskiparta, E. (2011). Bystanders matter: Associations between reinforcing, defending, and the frequency of bullying behavior in classrooms. Journal of Clinical Child & Adolescent Psychology, 40(5), 668–676. https://doi.org/10.1080/15374416.2011.597090
• Wang, Z., Laninga-Wijnen, L., Garandeau, C. F., & Liu, J. (2023). Development and validation of the Adolescent Defending Behaviors Questionnaire among Chinese early adolescents. Assessment, 30(7), 2258–2275. https://doi.org/10.1177/10731911221149082

Items of the Scale