Revisiting Respondent Fatigue Bias in the National Crime Victimization Survey

Timothy C. Hart, Bureau of Justice Statistics
Callie Rennison, Bureau of Justice Statistics

The National Crime Victimization Survey (NCVS) offers researchers and policy makers valid and reliable estimates of the nature and extent of criminal victimization in the United States. Since its inception, the NCVS has bendfitted from careful attention from researchers. For example, prior research identifies a number of types of nonsampling errors associated with the survey. However findings from past methodological studies remain largely unchallenged, though many were conducted prior to the NCVS redesign and the advent of other technologies.

The current research attempts to revisit an important type of nonsampling error commonly associated with the NCVS. Data from the NCVS longitudinal file were used to explore the relationship between a positive response burden and onresponse rates. Preliminary results do not support the hypothesis that exposure to longer interviews, administered as a result of answering positively to certain screening questions, result in higher refusal rates among victims than among non-victims during subsequent interviews. In short, findings from this study fail to support the widely held notion that "respondent fatigue bias" exists as a source of nonsampling error in the NCVS.

(Return to Program Resources)

Updated 05/20/2006