Peer victimization (PV) is widespread among children and adolescents in China. Extant research indicated that children of adverse backgrounds, such as rural-to-urban migrant children and left-behind children, are more vulnerable to PV. Relatively few longitudinal studies, however, have examined how PV changes over time and whether children exposed to adversity are consistently victimized over time. Furthermore, PV can lead to serious psychological problems. Consistent with interpersonal risk and social information processing models, the deleterious effects of PV on externalizing problems and depressive symptoms are well-documented in both cross-sectional and longitudinal studies. However, previous studies mainly examined the separate effects of the initial level and change rates of PV on externalizing problems and depressive symptoms and ignored possible synergistic effects of the initial level and change rates of PV. Three competing models—the additive, stress-amplification, and stress-sensitization models—can explain how the synergistic effects of the initial level and change rates contribute to externalizing problems and depressive symptoms. Thus, the present study used a longitudinal design to explore changing trajectories of PV and investigated whether the PV trajectories differed across migrant status. This study further examined how the initial level and change rates of PV synergistically predicted externalizing problems and depressive symptoms among children and adolescents at two years later.
This study employed a 4-wave longitudinal design spanning two years. Participants included 1, 580 students from three provinces in China who were recruited through multiple schools including elementary (43.2%), middle (33.1%) and high (23.7%) schools. The mean age of participants at the baseline was 12.37 years (SD = 2.52, range = 9 to 19 years old; 54.9% boys). All measures in this study were based on participants’ self-report. At baseline, participants completed demographic information, including gender, grade, migrant status and subjective socioeconomic status. Multidimensional PV Scale was used to measure PV at waves 1 to 3. At baseline and wave 4, Child Behavior Checklist and Center for Epidemiological Studies Depression Scale were used to measure externalizing problems and depressive symptoms, respectively. All analyses were conducted using Mplus 8.0. The analyses were conducted in three steps. First, the unconditional latent growth model was used to estimate individual change in PV over time. Second, multi-group latent growth models were used to examine whether the initial level, change rates and the correlation between initial level and change rates of PV differed across migrant status. Finally, conditional latent growth models with latent variable interactions (between initial level and change rates) were used to examine how the initial level and change rates across wave 1 to wave 3 jointly predict externalizing problems and depressive symptoms at wave 4.
First, unconditional latent growth modeling revealed a significant linear decline in PV over the three time points, and that individuals with higher initial levels of PV showed a faster decline; Second, multi-group latent growth model revealed significant differences in the trajectory of PV across migrant status. Specifically, compared to urban children, rural-to-urban migrant children, left-behind children and rural children had significantly higher initial levels of PV. Moreover, rural-to-urban migrant children and left-behind children exhibited a steep decline in PV; Third, conditional latent growth model with latent variable interactions showed that the interaction between the initial level and change rates of PV significantly predicted depressive symptoms and externalizing problems two years later, after controlling for demographics and depressive symptoms and externalizing problems at baseline. Supported by stress-sensitization model, individuals who were exposed to higher initial levels showed more depressive symptoms even when they exhibited a steeper decline in PV. But individuals exposed to lower initial levels would require more severe stress to trigger stronger stress reactivity (in our case, reflected as more depressive symptoms) at the slower rate of PV. Supported by stress-amplification model, individuals with higher initial levels of PV would have greater externalizing problems when experienced a slower change rate of PV. In contrast, individuals with lower initial levels would have fewer externalizing problems even when experienced a slower rate of PV.
The present study provides methodological and conceptual contribution: (1) we examined three key components— the initial level, change rates and the correlation between initial level and change rates of PV —to elaborate changing trends in PV over time and distinct victimization trajectories across children of different migrant status. These findings contribute to more comprehensive understanding of PV from developmental perspectives and precise detection and intervention aimed at supporting high-risk children and adolescents; (2) The analysis examining the interaction between initial level and change rates of PV adds novelty to extant literature and provides nuanced insights into synergistic effect that the initial level and change rates of PV affect problem behavior in an synergistic (rather than isolated) manner; (3) This study revealed that the initial level and change rates of PV play a unique role in depressive symptoms (risk sensitivity) and externalizing problems (risk enhancement). This highlights that the synergistic effect of initial level and change rates of PV may have significant implications for developmental science on adolescence health.
