2008 YVCES OMB Statement Part B 8-27-08

4

SUPPORTING STATEMENT

U.S. Department of Commerce

U.S. Census Bureau

National Youth Volunteering and Civic Engagement Survey

OMB Control No. 0607-0913

B. COLLECTION OF INFORMATION EMPLOYING STATISTICAL METHODS

1. Description of Universe

The sample for the National Youth Volunteering and Civic Engagement Survey will be drawn from about 8,000 households selected from expired Current Population Survey (CPS) housing units. We will randomly select one person aged 12 to 18 in each household to be interviewed. We estimate a 50% response rate at minimum. This survey will be conducted in October through December 2008.

The CPS samples were selected from 2000 decennial census files with coverage in all 50 states and the District of Columbia.

The interviews will be attempted by telephone by interviewers working out of one or more of the Census Bureau’s three centralized telephone facilities. Census Bureau interviewers will ask questions about the type and frequency of participation in volunteer activities.

The survey will ask about volunteer activities with a reference period from September 2007 through September 2008.

2. Sampling Methodology and Estimation Procedures

The estimation procedure will, for the most part, follow the usual statistical principles used in other surveys. The final weight for each case will be the product of the inverse of the selection probability (accounting for selection to the CPS and selection into the volunteer survey), a weight adjustment to account for noninterviews, a first-stage weighting factor to reduce the variance due to the selection on nonself-representing primary sampling units (PSUs), and a second-stage weighting factor to bring sample estimates into agreement with independent population controls by age, race, and sex.

The noninterview adjustment will include an adjustment to take into account the nontrivial possibility of a differential between respondents’ and nonrespondents’ volunteerism rates, first documented by Abraham et al (2007). This adjustment will be based on a comparison between 2008 Youth Volunteer and Civic Engagement Survey respondents’ and nonrespondents’ responses to the 2006 and 2007 CPS Volunteer Supplements. All housing units from which the 2008 Youth Volunteer and Civic Engagement Survey sample is being drawn were eligible for one or both of those supplements. A memorandum describing the adjustment procedure more fully is attached (Attachment B).

The 2005 Youth Volunteering Survey had 3,178 respondents, including 1,773 who were volunteers. This included 632 respondents from disadvantaged circumstances, of whom 285 were volunteers, and 347 who weren’t volunteers. The Corporation for National and Community Service drew comparisons between the latter two groups in its third brief in its Youth Helping America series, “Leveling the Path to Participation: Volunteering and Civic Engagement Among Youth From Disadvantaged Circumstances,” and would like to draw similar comparisons using data collected in the 2008 survey.

Based on an estimated design effect of 1.28, for comparisons of proportions between cells of that size to be statistically significant at the  = .10 level which is the Census Bureau standard, the differences would need to be up to 7.4 percent. If testing were done at the  = .05 level which the Corporation for National and Community Service used for its briefs based on data from the 2005 survey, differences of up to 8.9% would be necessary to achieve statistical significance. While there is no hard and fast minimum cell size requirement, substantially smaller cells would likely result in too few statistically valid comparisons to be of much use.

No specialized sampling was done to minimize the overlap of the NYVCES and the ATUS because it would have resulted in an inadequate sample (See Attachment D).

3. Efforts to Maximize Response

The Census Bureau telephone center staff will perform standard procedures to keep the noninterview rate as low as possible. If necessary, at least ten attempts will be made to get a completed interview. Calls will take place after 3 p.m. local time with at least two call attempts on weekends. We also plan to allow parents to serve as proxies if we are unable to contact the teen or if the parent is uncomfortable with Census interviewers talking directly to the teen. Interviews conducted using parents as proxies will be collected within the survey instrument.

To maximize response, we will also include a toll-free number on the advance letter sent to eligible respondents. The respondent can use the number to call in and set an appointment or to complete the interview. In addition, interviewers have job aids – answers to frequently asked questions (FAQs)—designed to help answer questions about the survey and to assist them in gaining respondent’s cooperation to participate.

We also are offering a $20 incentive to sample persons upon completion of the survey. Refer to A.9 and Attachment D for further details.

The estimation procedure includes a noninterview adjustment to adjust for those cases that the field staff cannot interview. We expect to achieve a response rate of at least 50 percent.

4. Tests of Procedures or Methods

New questions introduced to the survey have either been in use in the Volunteer Supplement which has been conducted yearly since 2002 or were field tested as part of the Corporation’s pretest of the Civic Engagement questionnaire which will be administered in November 2008.

5. Contacts for Statistical Aspects and Data Collection

You may consult the following Census Bureau individuals for information regarding sample design and data collection:

Sample Design:

Kimball Jonas

301-763-2483

Data Collection:

Richard A. Schwartz

301-763-7491

List of Attachments

Attachment A Advance Letters

Attachment B Nonresponse Bias Adjustment

Attachment C Questionnaire

Attachment D Coordination of the NYVCES and the ATUS