Abstract
The present article discusses the data source (self- as opposed to peer assessments) as one of the factors producing substantial variation in estimations of prevalence rates of peer victimization. It is argued that self-report data produce higher variation in prevalence estimations than peer reports. It is suggested that peer data may be preferable for research purposes, whereas self-reports may be more useful for intervention efforts. Second, the need for differentiating “discipline problems” and “genuine peer victimization” on a conceptual level is stressed. Third, it is outlined that this differentiation will lead to different intervention approaches (i.e., direct ones for discipline problems and more indirect ones for genuine peer victimization). Finally, more direct approaches to theory testing are called for. In doing this, the present article comments on data by Atria et al. (this issue) as well as questions raised by the analysis of Mahdavi and Smith (this issue).