ACS Questionnaire Design Test Plan

The Census Bureau plans to conduct additional research under the generic clearance for questionnaire pretesting research (OMB number 0607-0725). The objective of this research is to conduct cognitive testing of an alternative version of the American Community Survey (ACS).

This ACS questionnaire is part of the 2013 ACS Questionnaire Design Test (QDT). Alternate ACS designs are being developed to accommodate additional content that may be introduced on ACS mail questionnaires in the future. The purpose of the QDT is to determine how alternative questionnaire designs affect response rates, data quality, and cost. Cognitive test results of this ACS questionnaire will be used to inform quantitative test results from the QDT.

One ACS questionnaire will be cognitively tested. A draft copy of this form is attached. A description of the form is as follow:

1. 28-page paper ACS for OCR (Optical Character Recognition)

   1. 8.5” x 11” page size

   2. Three-column layout for housing and detailed person questions

   3. Green borders around pages, between columns, and around answer choices

   4. Separated “halos” between multiple-choice answers. Halos are the white spaces in that envelop the green answer check boxes.

   5. OCR "dentils" for numeric write-in fields, which allow a computer to read answers without the need for manual keying staff. OCR dentils appear as white square-shaped partitions within the write-in box space.

Between November and December 2012, staff from the Center for Survey Measurement (CSM) will conduct a maximum of 15 interviews among respondents with low internet experience (that is, people who use the internet less than 2-3 times a week for activities such as searching for information, shopping online, or completing web surveys). These respondents will be recruited because we want to test the questionnaire with the types of people who are likely to complete the mail questionnaire rather than an Internet form. To the maximum extent possible, within these specific criteria we will attempt to interview respondents of varying ages, genders and educational levels. We will use a multi-pronged strategy to recruit the respondents needed for this study, such as posting ads in free local newspapers and contacting local associations. Respondents will be screened to ensure that they have low internet experience.

Interviews will be conducted using interviewer observations and the retrospective probing method. Each respondent will complete one version of the ACS, and will be shown a standard 2013 English ACS mailout questionnaire (also attached) for comparison at the end of the interview. During the cognitive test, each respondent will complete a printed paper-and-pencil version of the experimental ACS form. A copy of the protocol used in the retrospective debriefing is attached.

The interviews will be conducted in the Washington DC metropolitan area. The interviews will be conducted at the Census Bureau’s cognitive laboratory, offices of community associations, and locations convenient to respondents.

Cognitive interviews will be tape-recorded, with the participants' permission, to facilitate a summary of the results. All participants will be informed that their response is voluntary and that the information they provide is confidential. Respondents will receive $40 for their participation.

The estimated time for completion of each of the cognitive interviews is 90 minutes. Thus, the estimated burden for this research is 22.5 hours.

The contact person for questions regarding data collection and study design is:

Rodney L. Terry

Center for Survey Measurement

U.S. Census Bureau

Room 5K022E

Washington, D.C. 20233

(301) 763-5475

Rodney.Terry@census.gov

2