In a newly reported study of secondary school classrooms, researchers found that higher levels of teacher-reported teacher–student co-positive experience were statistically associated with a weaker indirect link between adolescent peer victimization and symptoms of problematic internet use operating through loneliness. Published in Frontiers in Psychiatry, the investigation evaluated whether relational climates within school classrooms are associated with variation in psychological vulnerability among victimized youth. The statistical modeling revealed that the strength of the indirect association between experiencing peer victimization and exhibiting problematic internet behaviors differed substantially depending on the quality of the relational climate established across classroom settings.
Problematic internet use represents a pressing mental health concern for adolescents, with peer victimization recognized as a prominent correlate of heightened internet symptoms. Prior research had documented this baseline association, yet the underlying psychological mechanisms accounting for the connection and the classroom contextual factors that might temper or amplify it remained insufficiently understood. To address these gaps, the study specifically investigated whether student loneliness statistically accounts for the relationship between peer victimization and problematic internet use, while assessing whether teacher-reported teacher–student co-positive experience, conceptualized as a class-level environmental indicator, explains variations in that indirect statistical pathway.
The study sample comprised 1,854 Chinese adolescents with a mean age of 13.60 years and a standard deviation of 1.53 years, of whom 46.7 percent were girls. These adolescent participants were nested across 40 distinct classrooms located within three schools. In addition to the student sample, the research incorporated independent reports provided by the classroom teachers responsible for each of the 40 classes, establishing a multilevel data structure designed to separate individual adolescent perceptions from shared, teacher-evaluated classroom relational conditions across the educational cohorts.
Data collection utilized distinct informant sources to measure individual vulnerabilities and classroom-level relational dynamics. Peer victimization, subjective loneliness, and problematic internet use symptoms were gathered directly via student self-report instruments. In contrast, teacher–student co-positive experience was evaluated through teacher reports and treated as a class-level variable. By assessing teacher–student co-positive experience at the classroom level rather than relying solely on individual adolescent assessments, the investigators sought to capture an independent measure of the shared relational climate prevailing within each classroom environment.
To analyze the nested hierarchical data, the investigators estimated a multilevel moderated-mediation model using cluster-robust regression equipped with finite-sample corrections. The analytic plan evaluated conditional indirect associations and computed the formal index of moderated mediation utilizing Monte Carlo confidence intervals generated from 50,000 draws. Small-cluster inference was corroborated using a wild cluster bootstrap.
The analytical results demonstrated that peer victimization was positively associated with symptoms of problematic internet use, with a statistically significant within-class indirect association through student loneliness estimated at 1.516 with a 95 percent confidence interval ranging from 1.125 to 1.942. Higher teacher-reported teacher–student co-positive experience was linked to a significant attenuation of this first-stage pathway. Specifically, the conditional indirect association was 1.890 in low-TSCPE classes compared to 1.154 in high-TSCPE classes, reflecting a 39 percent difference. The index of moderated mediation was −0.209 with a 95 percent confidence interval of [−0.398, −0.034], and the wild cluster bootstrap interval similarly excluded zero at [−0.423, −0.030].
To test the robustness of the teacher-reported climate findings, the authors conducted a sensitivity analysis substituting a parallel student-reported measure evaluated at both the individual level and the class-mean level. This student-reported measure failed to replicate the buffering pattern identified through teacher reports. Instead, when evaluated as a class-mean aggregate, the student-reported measure demonstrated a statistically significant interaction that exhibited an opposite sign, underscoring substantial divergence between teacher-evaluated classroom climates and aggregated student perceptions of teacher–student interactions.
The investigators emphasized important methodological boundaries, explicitly noting that because the study employed a cross-sectional research design, the documented findings capture statistical associations rather than proven causal relationships. While teacher–student co-positive experience may indicate an environmental resource associated with reduced vulnerability among victimized students, the cross-sectional data cannot determine temporal precedence or developmental progression. The authors concluded that future longitudinal tracking and formal classroom intervention studies remain essential to establish directional relationships among peer victimization, loneliness, classroom climate, and problematic internet use.
Subject of Research: Relational classroom climate, peer victimization, loneliness, and problematic internet use in adolescents
Article Title: Teacher–student co-positive experience and the peer victimization–loneliness–problematic internet use association in Chinese adolescents: a multilevel moderated mediation study
Article References: Zhang, Z. (2026). Teacher–Student Co-Positive Experience and the Peer Victimization–Loneliness–Problematic Internet Use Association in Chinese Adolescents: A Multilevel Moderated Mediation Study. Frontiers in Psychiatry, 17, 1989218. https://doi.org/10.3389/fpsyt.2026.1989218
Image Credits: AI Generated
DOI: 10.3389/fpsyt.2026.1989218
Keywords: problematic internet use, peer victimization, loneliness, teacher–student co-positive experience, relational classroom climate, adolescent mental health, multilevel moderated mediation, Chinese adolescents, cluster-robust regression, finite-sample corrections, wild cluster bootstrap, Monte Carlo confidence intervals, conditional indirect associations, cross-sectional study, class-level climate, school psychology, Frontiers in Psychiatry
News Source: Glenn Wilkins. (November 18, 2026). Positive Teacher Relationships Associated With Weaker Victimization Link to Problematic Internet Use. Scienmag.



