A Potential Validity Threat in Victim Surveys? Assessing the Increase in Sensitivity to Violence and Its Implications for Measuring Victimization Trends

Victimization surveys have become a standard means of measuring crime trends, with the aim of capturing unrecorded crime in addition to crimes known to the police. "Reporting propensity" is often considered to be a major threat to the validity of police-recorded crimes as a measure of crime levels, but analogous processes can also impact the validity of victim surveys. Changes in sensitivity to perceiving incidents as "violence" is a case in point. In particular, the present study provides an empirical assessment of the extent to which changes in sensitivity to violence may pose a threat to the validity of survey-based research on violence. We drew on three sweeps of the Finnish National Crime Victim Survey (2013, <i>N</i> = 6,561; 2017, <i>N</i> = 5,615; 2021, <i>N</i> = 4,309) to examine changes in sensitivity and the association between sensitivity and reported victimization trends. An increase in sensitivity to violence during the research period was observed. Results from our mediation analysis suggest that adjusting for sensitivity significantly changes the association between survey year and the propensity to report multiple types of violent victimization. In particular, an increase in sensitivity may suppress an actual decrease in reporting victimization. Implications for measuring victimization trends in surveys and developing sensitivity measures are discussed.