Improving the measurement of attitudinal responses: An example from a crime survey

Abstract

The questions used in crime surveys (such as the British Crime Survey and Scottish Crime Survey) have been criticized for a number of different reasons, all of which call into question the validity of the questions employed to measure 'fear of crime'. Using data generated by research aimed at improving the conceptualization and measurement of the fear of crime, the current article outlines how assessments regarding the validity of such questions can be built into crime surveys. The procedures are eminently transferable to other topics as required.