Strengthening anti-bullying research: An investigation into the misuse of dichotomous variables

Year: 2005

Author: Finger, Linda, Marsh, Herb, Craven, Rhonda, Parada, Robert

Type of paper: Refereed paper

Effective strategies for addressing school bullying are underpinned by the findings of anti-bullying research. However the latter is plagued with methodological flaws including the use of: (a) dichotomous variables with quantitative and continuous data, (b) uni-dimensional approaches when assessing multi-dimensional concepts; and (c) unsound measurement instruments whereby the psychometric properties of the measures employed have not been demonstrated all of which contribute to misleading conclusions. The purpose of the present investigation was to address these issues and use longitudinal causal models, to examine the relations of secondary school students' (N = 3445, Males = 1780) self-concepts, bullying, victimisation, and depression scores. Results suggest bullying and victimisation are mutually reinforcing constructs and that the use of dichotomous variables to explore relations systematically underestimates the size of the relation between the constructs considered in this investigation.