ON THE RELATIONSHIP BETWEEN CRIMINAL VICTIMIZATION AND FEAR OF CRIME

Fast research has failed to detect consistent relationships between cr iminal victimization and fear of crime. For women and elderly people h igher levels of fear have been reported in spite of lower risks of vic timization when compared to other subgroups. These findings are suppos ed to be mainly artefacts, resulting from the violation of research re quirements: an adequate specification of fear indicators, the symmetry of victimization and fear indicators, the distinction of fear of actu al as opposed to fictitious victimization, the definition of non-victi ms, the correction for restriction of range of correlations, the contr ol of moderator effects, and the consideration of different variances within subgroups. The relevance of these requirements is demonstrated, using data from a representative German survey: First, correlational analyses of different fear indicators confirmed that fear of crime sho uld not be treated as a homogeneous phenomenon. Second, in applying di fferent assessment strategies, salience of family context and anonymit y proved crucial with respect to reporting violent incidents in close relationships. Finally, repeated analyses with additional methodologic al controls on every next step demonstrated that the fear-victimizatio n relationship depends considerably on the computational context.