Daily Rules, Proposed Rules, and Notices of the Federal Government
During the years preceding the 2020 Census, the Census Bureau will continue to pursue its commitment to reducing the costs of conducting a decennial census, while maintaining the level of quality achieved for previous ones. A primary decennial census cost driver is the employment of a large temporary staff to collect data from members of the public from which the Census Bureau received no reply via initially offered response options. Increasing the number of people who take advantage of self-response options (such as completing a paper questionnaire and mailing it back to the Census Bureau) can contribute to a less costly census with high-quality results.
The 2013 Census Test will give the Census Bureau an opportunity to investigate a variety of different strategies and methods aimed at increasing the use of self-response options in a decennial census. An overall objective of the Census Bureau is to increase participation by making it easier for respondents to know about and to respond to the decennial census. As part of this data collection, the Census Bureau will test different strategies for contacting the public to notify and to remind them about the decennial census. In addition, the Census Bureau will offer multiple modes to self-respond.
The 2013 Census Test will also encompass research in additional key areas. One area pertains to testing different field data collection procedures for obtaining data from those who do not self-respond to the decennial census. Another area involves collecting data on attitudes regarding contact strategies, response modes, and other proposed methods of data collection. The Census Bureau will recontact a sample of those who self-responded, those who responded to a Census Bureau employee, and those who did not respond at all. The data collection to obtain respondent attitudes will be conducted by telephone.
The results from the 2013 Census Test will inform Census Bureau planners who are guiding the design of additional 2020 Decennial Census research on the topics summarized briefly above and discussed in more detail below.
For this test, the Census Bureau is intending to use multiple contact modes to notify respondents to participate in the census, to provide them instructions for completing a census questionnaire, and to remind them to respond. In addition to mail, the Census Bureau is considering contacting respondents by email and text messages using contact information purchased from commercial data vendors. In advance of this test, the Census Bureau will address any policy issues surrounding the use of email and text messages.
The email and text messages will contain an interactive link to a Census Bureau Internet site that respondents can click on to respond to the census. The Census Bureau plans to embed into the link an identifier that is unique to the respondent and their notification mode (for example, the same respondent with both an email and text account may have a unique identifier for each one). This identifier will allow the Census Bureau to measure the effectiveness of each mode of notification and to determine any response differences by demographic group or geographic area (such as urban, suburban, and rural).
In addition to altering the mode of contact, the Census Bureau will vary both the content of messages sent and the timing of when respondents will receive them. This testing will help the Census Bureau to develop effectively worded messages and delivery schedules that are optimized for each mode of contact. The Census Bureau will measure the effectiveness of differing mode, message content, and time of delivery on the response rate. This analysis will include breakdowns by various demographic populations.
The current NRFU procedure is to initially make a personal visit to an address to conduct an interview. If no one comes to the door during this visit, NRFU interviewers leave a notice informing the resident(s) of the interview attempt. Interviewers can leave a telephone number on the notice of visit encouraging the resident to call them back. If the resident calls an interviewer or an interviewer obtains a telephone number for a household (from a neighbor, for example), the interviewer can conduct the interview over the telephone instead of making another personal visit.
As part of the overall effort to reduce the operational cost of NRFU, this test will explore alternatives to the current NRFU procedures for contacting households. For example, the Census Bureau will experiment with the number of attempts to contact each household before allowing the field staff to obtain information from proxy respondents. Doing so will supplement research already done on this topic and will help planners to determine the optimal number of in-person visits and telephone contacts to make during this operation. The Census Bureau also wishes to learn whether altering the way interviewers contact a household for the first time (that is, by telephone instead of by personal visit) results in a more efficient way to conduct NRFU (telephone contact is considerably less expensive than personal visit). To explore this alternative, the Census Bureau plans to provide field staff with telephone numbers from commercial data vendors for addresses in their workload, so they can first contact respondents using the telephone rather than making the first contact a personal visit.
The Census Bureau will also test different notification strategies and messages that the census staff can leave at the household. One strategy is to leave instructions for the household on how to use the Internet to submit responses to the census. Obtaining an Internet response in this type of a scenario can save the expense associated with census staff making a return visit (or a telephone contact).
The Census Bureau will use the newly devised mobile computing devices to conduct interviews and will enter the responses into the device rather than recording them on paper questionnaires. The mobile computing device can automate manual tasks such as managing the field staff work assignment. The Census Bureau expects to use the mobile computing device to collect a more complete and accurate recording of attempts to complete an interview than has been possible in the past with only paper questionnaires. Obtaining better data on the actual number of contact attempts will help planners to develop future contact strategies. In addition, the Census Bureau may allow respondents to self-respond directly on the mobile computing device during a personal visit.
The Census Bureau will select a sample of up to 160,000 housing addresses for the 2013 Census Test. The majority of the addresses (approximately 60 percent) will be located in several different geographic test sites (locations to be determined). The Census Bureau will attempt to select geographic test sites that comprise urban, suburban, and rural areas as well as contain a diversity of socio-economic populations. However, budget limitations may affect the final selection. Approximately 40 percent of these addresses will be randomly selected from a national sample. Current plans do not target Tribal sites and group quarters addresses for this test due to limitations of sample size and budget. The Census Bureau estimates a 45 percent self-response rate overall. The Nonresponse Followup workload will be no more than 40,000 household addresses in the geographic test sites due to budget limitations. Of the 40,000 addresses, the Census Bureau will recontact approximately five percent for purposes of quality assurance.
The sample size for the attitudinal survey will be no more than 50,000 households comprised of 25,000 respondents and 25,000 nonrespondents. The sample will be drawn from a cross section of the national sample and the different geographic test sites. In order to reduce costs, the quality assurance sample will be drawn from the same nonrespondents sampled for the attitudinal survey. During the attitudinal survey, we will ask this quality assurance sample if the NRFU interviewer visited the household. The Census Bureau estimates that it will take 10 minutes to complete this additional survey for respondents and 12 minutes to complete the survey for nonrespondents.
In general, the 2010 Census Questionnaire will be the basis of the questions asked during this test. While the question topics will remain the same, the Census Bureau may revise question wording based upon further research and testing, such as results from the 2012 National Census Test (OMB 0607-0970). The Census Bureau estimates that it will take the average household 10 minutes to complete the questionnaire. It includes probes (such as, are there any college students listed) that attempt to find people not initially listed on the housing unit roster and people that have multiple residences. The Census Bureau plans to identify housing units that are more likely to answer positively to each probe and to change the order of the probes in the non-paper questionnaires, so the most relevant probe is asked first for that housing unit. In addition, the questionnaire may need modification based on mode (Internet, or Computer Assisted Telephone Interviewing (CATI)) to improve the flow of the wording. The Census Bureau will design the form for viewing on different Internet, CATI, and mobile computing device platforms used by the field staff (sometimes referred to as Computer Assisted Personal Interviewing, or CAPI).
All households in the sample will initially receive a notification to participate in the 2013 Census Test by email or letter sent in the mail. Reminder notifications will be sent by several methods: email, text message, postcard in the mail, or letter in the mail. The exact content and timing of the notifications are still under consideration. Respondents will have
For a sample of households that do not self-respond, the Census Bureau will conduct a Nonresponse Followup (NRFU) operation. The Census Bureau will use mobile computing devices to conduct interviews and will enter responses into the device. The Census Bureau will also recontact some respondents as part of quality assurance activities of its field staff.
Finally, the Census Bureau will conduct a follow-up telephone survey for a sample of respondents and nonrespondents to explore attitudes regarding contact modes, response modes, use of administrative records to collect data, and other proposed methods of data collection.
Comments are invited on: (a) Whether the proposed collection of information is necessary for the proper performance of the functions of the agency, including whether the information shall have practical utility; (b) the accuracy of the agency's estimate of the burden (including hours and cost) of the proposed collection of information; (c) ways to enhance the quality, utility, and clarity of the information to be collected; and (d) ways to minimize the burden of the collection of information on respondents, including through the use of automated collection techniques or other forms of information technology.
Comments submitted in response to this notice will be summarized and/or included in the request for OMB approval of this information collection; they also will become a matter of public record.