Bias-Based Harassment among Adolescents: Differential Associations with Functioning Based on Involvement Type

• In Event: Adverse Experiences at School: Bias-based Bullying by Peers and Unfair Treatment by Teachers

Abstract

Background: Nearly 1 in 5 high school students have experienced bias-based harassment on school property in the past year (Mulvey et al., 2018). Compared to general bullying there are particularly pronounced effects on well-being for those involved in bias-based harassment (Russell et al., 2012). While research on general bullying has identified bully-victims as youth at particularly high risk for mental health, substance use, and school challenges (Sangalang et al., 2016; Lee et al. 2018; Hysing et al., 2021), much less is known about whether a similar pattern exists for bias-based harassment. This study extends past research through examining the degree to which mental health, substance use, and school belonging indicators vary among adolescents classified as (a) no perpetration/victimization (a) victimization only, (b) perpetration only, and (c) both victimization and perpetration of bias-based harassment.

Methods: This study was conducted using NORC’s AmeriSpeak panel, a U.S. household panel that uses a national sample frame to provide a representative sample of U.S. adolescents. The current sample consists of 639 adolescents, ages 13-17, who completed online surveys in 2021 (see Table 1). California Healthy Kids Survey (Austin et al., 2016) items were used to assess past 30-day bias-based harassment (perpetration and victimization) related to race or ethnicity, national origin, religion, sexual minority identity, immigration status, and gender identity. Items were used to create a mutually exclusive categorical variable that included no perpetration or victimization (NOPV), victimization only (VON), perpetration only (PON), both perpetration and victimization (BOPV). Four dependent variables were: anxiety (GAD-7), depression (PHQ-9), substance use (Problem Behavior Frequency Scale), and school support (Student Support Scale).
Latent variable models were fit in Mplus 8.4 using the Weighted Least Squares Means and Variance adjusted estimator to account for the ordered categorical nature of items using weighted data. Beginning with a measurement model that included the four dependent variables we established acceptable fit (CFI = .98, TLI = .978, RMSEA = .028. SRMR = .063) prior to examining the hypothesized structural model that included several demographic controls (e.g., grade, ethnicity, race, gender) and the bias-based victimization and perpetration variables.

Results: Findings from the multivariate structural model with all four dependent variables and demographic controls (see Figure 1) indicated that compared to the NOPV group, the BOPV group reported significantly higher levels of depression, anxiety, and substance use and lower levels of school support; the VON group reported significantly higher levels of depression and anxiety and lower levels of school support; and the PON group reported significantly higher levels of depression, anxiety, and substance use. We also examined the remaining comparisons and found that the VON group was associated with significantly lower levels of substance use compared to the PON and BOPV groups.

Discussion: Findings highlight that involvement in bias-based harassment in any capacity is associated with deleterious functioning, with youth involved as both perpetrators and victims reporting adverse effects across the most domains. Future longitudinal research could attend to how shifts in bias-based harassment over time are linked to changes in functioning.

Authors

• Melissa Holt, Boston University
  Presenting Author

• Gabriel J Merrin, Syracuse University

• Nicolina Fusco, Boston University

• Jennifer Greif Green