Abstract
While the life-course perspective dominates the study of offending, its adoption in the study of victimization has barely begun, leading to calls in the literature for the development of a life-course victimology. However, longitudinal self-report data present methodological challenges. This article investigates these methodological challenges with the ‘accelerated longitudinal design’ of the US National Crime Victimization Survey (NCVS). It tests whether multiple age cohorts in the NCVS show one overall underlying age−victimization curve, or whether (and why) they show systematically different age−victimization curves. The analyses show that, for every age cohort studied, self-reported victimization decreases over consecutive measurements. Explanations offered for this decrease include selective panel attrition, the American crime drop, and panel-respondent fatigue. The separate age cohorts do not share the same underlying age−victimization curve. Implications for research are discussed.
Get full access to this article
View all access options for this article.
