Supporting StatementBrevised

3

B. Description of Statistical Methodology

1. Respondent Universe

The sample universe is all children ages 12 through 18 who have attended school during the previous six months (Grade 12 or less). Approximately 9,445 children between the ages of 12 and 18 who live in NCVS households will be eligible for the survey. In the national sample interviewers are able to obtain interviews with approximately 70 percent of household members ages 12-18 in 90-91 percent of the occupied units in sample in any given month.

2. Statistical Methodology

In the national sample there are approximately 51,000 designated households, in clusters of four, in 694 primary sampling units located throughout the U.S. The sample consists of six parts, each of which is designated for an interview in a given month and again at six-month intervals; for example, July and January, August and February, September and March, and so forth. From 2005 through 2007 the NCVS sample is based on the 1990 and the 2000 decennial census as well as new construction universes. January 2005 marked the beginning of a two year phase-in of the 2000 sample design, which is based upon the 2000 decennial census. In January 2008 the 1990 design sample will have been completely phased-out and replaced by households selected from the 2000 sample design. Therefore, the 2009 SCS is based solely on 2000 decennial sample.

The NCVS uses a rotating sample. The sample consists of six groups for each month of enumeration. Each of these groups stays in sample for an initial interview and six subsequent interviews. The initial interview is essentially a bounding interview to establish a reference period to reduce telescoping, and it enables the interviewer to avoid recording duplicate reports on subsequent visits. During the course of a six-month period, we interview a full sample of six rotation groups — one-sixth each month. In addition, one rotation group enters the sample for its bounding interview each month.

Each interview period, the interviewer fills or updates the NCVS control card questions in the CAPI instrument for the household and fills the personal characteristics and crime screen question for each member 12 years old or older. The interviewer then fills a crime incident report for each reported crime incident. Each household member provides the information by self-response.

The first contact with a household is by personal visit. For the second through seventh visits, we use a mixture of personal visits and telephone interviews.

The full NCVS experienced methodological changes in 2006 including: a new sample, a change in the method of handling first-time interviews with households, and a change in the method of interviewing (see http://www.ojp.usdoj.gov/bjs/abstract/cv06.htm). The data from the 2007 SCS was just received at BJS at the end of August 2008, so there has not been sufficient time to determine whether the 2007 SCS was subject similar mode and instrumentation effects as the full NCVS. In fact, recent findings from an evaluation of the full NCVS data in 2007 indicate that we are observing a reversal in the same areas, enough to consider any remaining effects small enough to be confounded with sampling error.

3. Maximizing Response Rates

Census Bureau staff mails an introductory letter (NCVS-572(L), NCVS-573(L)) explaining the NCVS to the household before the interviewer's visit or call. The Census Bureau trains interviewers to obtain respondent cooperation and instructs them to make repeated attempts to contact respondents and complete all interviews. The interviewer obtains demographic characteristics of noninterview persons and the race of noninterview households for use in the adjustment for nonresponse. Household response rates are monitored on a daily basis within the automated survey control system. Person level response is measured weekly based on the receipt of weekly output data files. Household and person level response rates are compared to the previous months average to ensure their reasonableness. Regional office staff and field representatives are notified to take appropriate corrective action anytime household or person level nonresponse show a troubling trend.

The response rates for the last two (2005 and 2007) administrations of the School Crime Supplement were 79.9% and 59.6%, respectively. Two nonresponse bias analysis reports were produced for the 2005 SCS examining unit and item level nonresponse. These same reports will be prepared for the 2007 and 2009 supplements.

4. Test of Procedures

Any new or modified questions were cognitively tested and documented in “Pretesting of the 2009 School Crime Supplement: Final Results and Recommendations.” Most of the questions used in the previous five surveys have been incorporated in this current supplement.

5. Consultants on Statistical Aspects of the Design

The Census Bureau will collect all information. Mr. Stephen Ash and Ms. Barbara Blass of the Demographic Statistical Methods Division of the Census Bureau provided consultation on the statistical aspects of the supplement. Mr. Jeremy Shimer heads the Crime Surveys Branch of the Demographic Surveys Division, which manages and coordinates the NCVS and the SCS supplement.