Memo to OMB on Nonsubstantive Change

U.S. Department of Justice

Office of Justice Programs

Bureau of Justice Statistics

Washington, D.C. 20531

MEMORANDUM

TO: Shelly Wilkie Martinez

Office of Statistical and Science Policy

Office of Management and Budget

THROUGH: Lynn Murray

Department Clearance Officer

Department of Justice

William J. Sabol, Ph.D.

Director

Bureau of Justice Statistics

FROM: Michael Planty, Ph.D.

Chief, Victimization Unit

Bureau of Justice Statistics

SUBJECT: Submission of non-substantive modification for the National Crime Victimization Survey (NCVS) (OMB# 1121-0111)

DATE: January 13, 2016

This memo concerns a non-substantive modification to the National Crime Victimization Survey (NCVS). Specifically, the Bureau of Justice Statistics (BJS) is proposing to add five socio-demographic questions to the crime screener instrument (NCVS-1) regarding respondent’s veteran status, citizenship, disability status, sexual orientation, and gender identity; and expand the household income response categories (see attachment 1).

The NCVS Instrument Redesign and Testing Project, which is designed to provide scientific and technical support for the redesign and testing of the NCVS roster control card, crime screener (NCVS-1), and crime incident (NCVS-2) instruments, is currently ongoing. However, this project work and recommended changes to the NCVS instruments will not be completed for at least 2 years. In the meantime, BJS feels that important near term changes can be made to improve the relevance and utility of the NCVS without affecting the victimization rates. Through the initial assessment work with this project, one area in need of modernization that was identified was the socio-demographic information collected about respondents. This information should be enhanced to better address policy-relevant questions about victimization. Specifically, new socio-demographic measures should be added including the above-mentioned items: veteran status, citizenship, disability status, sexual orientation, and gender identity. These measures have been identified in other research as subgroups of interest to key stakeholders and, correlates to victimization. For example, sexual orientation and gender identity are recognized in the 2013 reauthorized Violence Against Women Act (VAWA). Additionally, the inclusion of these items will allow researchers to better understand the relationships between these variables and experiences with criminal victimization.

These proposed socio-demographic measures will be added to the NCVS beginning July, 1 2016. The proposed items are in Attachment 1. This includes recommendations on which respondents should be asked the questions, how often the questions should be asked, and the question wording.

Items

The proposed socio-demographic questions on veteran status, citizenship status, disability, sexual orientation, and gender identify would be added to the end of the NCVS-1 (person interview), where other demographic items are currently collected (see attachment 2). The new questions would be added after the crime screening questions and before the current employment question (75a, see attachment 3).

The new items all come from existing data collection sources. Veteran status and citizenship status questions come from the American Community Survey (ACS). Sexual orientation questions are from the National Health Interview Survey (NHIS) and gender identity questions are modified from the California Health Interview Survey (CHIS). The disability status questions are currently on the NCVS-2, but they will be moved to this new section of the instrument with the other new demographic questions.

These new questions were cognitively tested on the upcoming Supplemental Victimization Survey on stalking (NCVS SCVS Gen IC #201312-0607-004). The Census Bureau conducted the 60 cognitive interviews with respondents age 16 and older. The existing measures on veteran status, citizenship, and disability were mostly tested for placement. There were no reported issues with these measures. The measures for sexual orientation and gender identity were tested further, with respondents probed on how they performed, in terms of ease of understanding, burden, and level of discomfort caused. These questions performed well, and no issues were found. Respondents understood the questions, and had no trouble answering them. In addition, BJS consulted with persons knowledgeable of the NHIS research, key external researchers from the UCLA Williams Institute and CHIS, and members of the OMB Working Group on measuring sexual orientation and gender identity.

In addition to these new socio-demographic questions, the household income response categories will be expanded. Currently, the household income categories are capped at $75,000 and over (see attachment 1). Expanding the household income categories to be more in line with the American Community Survey would allow for a greater level of detailed collected about victimization by household income. Further, BJS has implemented an imputation strategy for dealing with missing income responses (see here: https://www.ncjrs.gov/pdffiles1/bjs/grants/248563.pdf and here: http://www.bjs.gov/content/pub/pdf/hpnvv0812.pdf). This expansion of the response categories will provide more detail to inform this imputation strategy.

Depending on measures of precision, in part due to the victimization rate, BJS expects to analyze and disseminate the new items in these subgroups:

Veteran status:

Veteran and non-veteran

Veteran by period

Citizenship:

U.S. Citizen

Not a U.S. Citizen

Disability status:

Disabled

Not disabled

Disability type

Sexual orientation:

Gay/Lesbian

Straight

Bisexual

Gender identity:

Transgender

Nontransgender

Impact

These proposed modifications to the NCVS instrument are expected to have minimal impact on burden and on victimization estimates. They have proven to be relatively easy to understand and their placement after the victimization screener questions will have limited impact by design. The questions will be monitored while in the field, and necessary adjustments will be made if the questions are not performing well. The addition of these questions will allow for a better understanding of the relationship between victimization and these new demographic characteristics.

If there are any questions concerning this request, please contact Michael Planty, Victimization Unit Chief, Bureau of Justice Statistics, at (202) 514-1062 or by email at [email protected].

Attachment 1. NCVS New Demographic Items

Attachment 2. NCVS Core Changes Flowchart

Attachment 3. NCVS-1 instrument (OMB# 1121-0111)