A Survival Analysis of Revictimisation

Michael Townsley, Jill Dando Institute

ABSTRACT
This study addresses the weaknesses of measuring the time between victimisations, conventionally called the time course of victimisation. Survival analysis -- measuring the time until an event of interest -- is presented as an alternative method for examining temporal patterns of victimisation. However, adopting statistical techniques developed in other fields requires careful scrutiny of the assumptions inherent with the technique. Most applications of survival analysis assume that subjects will eventually experience the event of interest. This is not a valid assumption for modelling the time until revictimisation because the majority of victims do not experience a second victimisation. Finally, a survival analysis model that utilises information about non-repeats is applied to a five year burglary dataset incorporating correlates.

(Return to Program Resources)

Updated 05/20/2006