Response to the Commentary on “Questionable Mediated Effect of Peer Victimization on Prosocial Behavior in Adolescents: A Reanalysis of Guo et al. (2026)”

This response addresses the commentary proposing a reanalysis of our published study on peer victimization, empathy, psychological capital, and prosocial behavior in adolescents. We first clarify the scope of the original findings. Because the original study used cross-sectional observational data, its mediation results should be interpreted as evidence consistent with the proposed indirect associations rather than as definitive evidence of causal mechanisms. We then address the commentary's alternative model, which represents peer victimization, empathy, psychological capital, and prosocial behavior as reflective indicators of a single core self-evaluations (CSEs) factor. Our central concern is not whether this model can reproduce the observed covariance structure, but whether the proposed latent factor has sufficient theoretical and construct-validity support. Statistical fit does not by itself establish that heterogeneous variables represent manifestations of a common psychological construct. Accordingly, the alternative model requires additional theoretical justification and direct validation of its measurement assumptions. We view the CSE model as a possible alternative statistical representation of the observed associations, but not as an established alternative psychological explanation. Both the original mediation model and the proposed CSE model should therefore be evaluated through longitudinal, multi-method research and stronger construct validation. Residual covariance can improve statistical fit without resolving the underlying construct-validity question.