Staff undertook work to conduct cognitive research on newly proposed questions for the NCVS screener questionnaire on the topic of identity theft. We conducted the research in February and March 2004 with respondents who had been the victim of identity theft. Some were victims within the six-month reference period of the survey, and others were victims less recently. We conducted three iterations of interviews, making changes to the questionnaire as problems became apparent. The main problems that surfaced were that: 1) respondents were not able to easily divide their victimization experiences into types of identity thefts, and questions for which reporting was requested for the most recent type of identity theft were difficult for them; 2) respondents misinterpreted the concept of the person responsible for the identity theft, thinking they would have to be able to identify them by name or face rather than by their type of interaction (e.g., the person who sold them a cell phone); 3) respondents were not able to provide dollar amounts for time lost from work to resolve identity theft issues for two reasons: the concept of "time lost from work" was vague (did it include vacation time taken, work time spent doing non-work-related activities, or unpaid time?), and they tended to think in terms of hours spent on these activities, not the amount of money involved.