NCVS Questions and Answers

National Crime Victimization Survey

Questions and Answers

In SS A6, it says that “If a longer reference period is used, the survey will be impact in the following ways: 1) loss of annual change estimates,…. Please clarify this statement, as our understanding is that the redesign is proposing a 12 month reference period while continuing to produce annual change estimates.

Previous research has demonstrated that longer reference periods are associated with lower respondent recall and increases in error associated with respondent recall as well as increases in respondent burden. One of the redesign research projects now underway is exploring whether it is possible to reduce the negative impacts of using a 12 month rather than a 6 month reference period. BJS has not yet made the decision whether the longer reference period will be incorporated into the survey. If it is used, it will be implemented as part of the overall redesigned survey implementation because moving from a 6 to a 12 month reference period would inevitably cause a break in series. The statement in SS A6 refers to the effects of such a break.

1. In SS B2, item on “collecting information on race of offender,” please clarify to whom BJS discussed the issues. What are “the issues?” What literature or expert advice from other domains where racial data is collected by observation was obtained? Please provide the full report on the findings of that effort.

BJS is planning a research and field testing program to make a transition. As a first step, BJS will conduct a literature review to assess the extent of understanding of how well people can provide information about offender race and ethnicity. Building on this review, the next step is to research both how best to ask questions on this topic and also to evaluate the usefulness of building qualitative measures about how respondents determine the race/ethnicity of the offender. For example, a question may assess how sure a respondent is about the information being provided about the offender, or how good a look the victim got of the offender. Such information may be useful in mitigating the misuse of data about victim perceptions of offender characteristics.

In summary, BJS plans to develop and test a new set of questions in 2010 and potentially to put these new questions in place on the NCVS in 2011 to run parallel to the existing question through a split sample approach and then to transition completely to the new questions in 2012 (placing these questions on any new NCVS designs being tested during that year).