Justification for nonsubstantive change

Request for Nonsubstantive Change to 0607-0110

The Census Bureau requests a temporary change to the approval for the Survey of Construction (SOC) data collection. The SOC is used to collect information on newly construction residential buildings in the United States. The SOC data collection has been approved by the Office of Management and Budget (OMB) through June 30, 2009.

SOC data are collected by Census Bureau field representatives (FRs) using Computer Assisted Data Collection (CAPI). We are in the process of rewriting the SOC data collection instrument. This provides an opportunity to add new data items to the survey to provide new data products requested by key data users. We have identified several possible new data items requested by the Department of Housing and Urban Development (HUD), which provides partial funding for the survey, and by other key data users such as the National Association of Home Builders (NAHB). We need approval to test the new questions to determine whether SOC respondents are able to provide the information and to identify any improvements needed to the question wording.

The test will be conducted by FRs as part of one normal monthly interview with SOC respondents, using cases already in the SOC sample. To reduce computer programming costs for the test, the FRs will record the responses on a paper worksheet. This paper form will not be provided to respondents to complete; the test interview will always be conducted by the FR. One questionnaire will be administered for single-family buildings and another for multifamily buildings.

To determine whether respondents can provide the information, the test will be administered to cases that were sampled during the test month and to sampled buildings whose construction was reported as completed during the test month. The two test groups are needed to determine whether the information is available at the initial interview and also whether it is available when the unit is completed. Most characteristics of sampled SOC cases are collected during the initial interview. Availability at completion is also relevant because any of the test items that are added to the new questionnaire would eventually be published as characteristics of new housing units completed. Approximately 1,400 single-family cases and 600 multifamily cases from each of these two groups will be interviewed, so the total number of interviews for the test will be about 4,000. We have determined that this number of interviews will guarantee at least a 5% level of accuracy for our test data, which will be reviewed by region as well as at the U.S. level.

We are requesting approval to conduct the test in conjunction with the SOC data collection during either October 2007 or November 2007. The test month may need to be adjusted to accommodate other previously scheduled tasks in the project to redesign the data collection instruments. After the testing is completed, we will determine which new questions will be added to the monthly electronic questionnaire, and the entire new questionnaire will be submitted to the OMB for approval. We plan to begin using the new questionnaire in production in May 2008.

The testing will be conducted for one month only. We do not expect the testing to raise the overall annual reporting burden calculation.