2017 Census Test OMB Package Part B_final with updates

Department of Commerce

United States Census Bureau

OMB Information Collection Request

2017 Census Test

OMB Control Number 0607-XXXX

Part B – Collection of Information Employing Statistical Methods

1. Universe and Respondent Selection

A national sample of 80,000 households will be drawn for a self-response-only operation where a tribal enrollment question will be asked. It will be oversampled for areas with relatively higher concentrations of people estimated to identify as American Indian or Alaska Native. The sample households will receive the 2017 Census Test materials through the mail. The response rate is estimated to be 45 percent.

The universe for the 2017 Census Test will include all housing units in areas where Type of Enumeration is Self-Response. Housing units that were selected to participate in the 2015 National Content Test, 2016 Census Test, 2016 American Community Survey (ACS), 2017 ACS, and the recent Small Scale Mailout Testing will be removed from the universe to reduce respondent burden. The universe will be stratified at the tract level into the following five strata: households in Alaska, areas with relatively high populations of AIAN and low Internet response propensity, areas with relatively high populations of AIAN and high Internet response propensity, areas with relatively low populations of AIAN and low Internet response propensity, and areas with relatively low populations of AIAN and high response propensity. These strata were chosen because they carefully balanced sufficient sample sizes with a reasonable design effect. The national sample is intended to provide sufficient sample to provide statistical results on the feasibility of collecting information on tribal enrollment.

2. Procedures for Collecting Information

The mailing packages and other partnership and communication materials will provide the 2017 Census Test Uniform Resource Locator (URL) and the phone number for Census Questionnaire Assistance (CQA). People can respond by Internet, through a phone call to CQA, or by mailing back a paper questionnaire if one was received. This national sample is intended to provide sufficient sample to provide statistical results on the feasibility of collecting information on tribal enrollment. The self-response rate will also provide an additional data point for use in developing the set of contact strategies for the 2020 Census.

For this sample, there will be no follow-up to obtain response. However, there will be a subsample selected for reinterview on the telephone to check the quality of responses to the tribal enrollment question. The reinterview sample of 15,000 housing units will be selected from the response data from the initial survey. This sample selection will depend largely on the responses that we receive. There will be a heavy sample of housing units in Alaska, housing units with people who racially identify as AIAN, and housing units with people who indicate that they are enrolled in a tribe. The sampling will occur on a weekly basis. The analysis will include calculating unit and item nonresponse and consistency rates between the initial survey and reinterview responses. We expect to be able to estimate the consistency rate between the initial survey response and the reinterview response with a margin of error of three percent or less with an alpha level of 0.05.

Type of Respondent/Operation	Estimated Number of Responding Housing Units	Estimated Time per Response	Estimated Total Annual Burden Hours
Self-Response	36,000	10 minutes	6,000
Reinterview	7,500	7 minutes	875
Totals	43,500		6,875

3. Methods to Maximize Response

Response to the 2017 Census Test is mandatory.

The households in the sample will receive various mailings, incorporating the contact strategy intended to maximize self-response, as has been developed and refined in previous census tests this decade. There will not be an additional followup component intended for collecting responses. This contact strategy is shown in the table that follows.

2017 Census Test National Sample Contact Strategy

Panel	#1 (Monday, March 20)	#2 (Thursday, March 23)	#3* (Monday, April 3)	#4* (Monday, April 10)	#5* (Thursday, April 20)
1)Internet Push English Only	DG-16(L1) + DG-17(CQA) + DG-6B(IN)(E/S)	DG-16(L3) + DG-5(E/S)	DG-9(P)	DG-1D + DG-17(L1) + DG-6A(1)(IN)(E/S) + DG-8A(E/S)	“It’s Not Too Late” postcard DG-9
2)Internet Push English/Spanish	DG-16(L1)(E/S) + DG-17I(E/S) + DG-17(CQA) + DG-6B(IN)(E/S)	DG-16(L3)(E/S) + DG-5(E/S)	DG-9(P)(E/S)	DG-1D(E/S) + DG-17(L1)(E/S) + DG-17I(E/S) + DG-6A(1)(IN)(E/S) + DG-8A(E/S)	“It’s Not Too Late” postcard DG-9(E/S)
3) Internet Choice English only	DG-1D + DG-16(L2) + DG-17(CQA) + DG-6A(IN)(E/S) + DG-8A(E/S)	DG-16(L4) + DG-5(E/S)	DG-9(C)	DG-1D + DG-17(L1) + DG-6A(1)(IN)(E/S) + DG-8A(E/S)	“It’s Not Too Late” postcard DG-9
4) Internet Choice English/Spanish	DG-1D(E/S) + DG-16(L2)(E/S) + DG-17I(E/S) + DG-17(CQA) + DG-6A(IN)(E/S) + DG-8A(E/S)	DG-16(L4)(E/S) + DG-5(E/S)	DG-9(C)(E/S)	DG-1D(E/S) + DG-17(L1)(E/S) + DG-17I(E/S) + DG-6A(1)(IN)(E/S) + DG-8A(E/S)	“It’s Not Too Late” postcard DG-9(E/S)

4. Tests of Procedures or Methods

We are testing a tribal enrollment question. This is entirely new methodology, and as such, significant qualitative testing, which included focus groups and cognitive interviews, were conducted in advance of this field test. The national sample is intended to provide sufficient sample to provide statistical results on the accuracy and reliability of a question on tribal enrollment. The self-response rate will also provide an additional data point for use in developing the set of contact strategies for the 2020 Census.

The 2017 Census Test will continue to include the response categories recommended by the OMB Interagency Working Group (see Section 11 of this document – Justification for Sensitive Questions) for opposite-sex and same-sex husband/wife/spouse households and for the category for unmarried partner. There may be additional results from this test that could inform final decisions about the use of these categories. The Census Bureau is working to refine the operation of the question in electronic instruments, in terms of agreement between the relationship report for coupled households and the sex reports for the householder and their spouse or partner.

The 2017 Census Test will also help advance the body of knowledge about the combined race/ethnicity question. OMB and the Census Bureau have made no decision on which race/ethnicity question(s) format will be used for the 2020 Census. The final decision made by OMB and the Census Bureau will be based on empirical research, results from decennial census testing, and years of engagement with key stakeholders, the U.S. Congress, and the public.

More details about this research are provided in Supporting Statement A.

5. Contacts for Statistical Aspects and Data Collection

For more information, contact Robin A. Pennington (301) 763-8132, [email protected].

3