OMB Memo

2023 Census Test

Submitted Under Generic Clearance for Census Bureau Field Tests and Evaluations

Request: The Census Bureau plans to conduct additional research under the generic clearance for Field Tests (OMB #0607-0971). We will be administering an online mandatory survey with a paper questionnaire sent in the last mailing to nonresponding households to start answering some research questions posed in two project plans in the 2030 Census Enhancement Area 1.1 on Self-Response.

Purpose: The 2023 Census Test is the first planned test under the renewed Small-Scale Response Testing (SmaRT) program. The purpose of the SmaRT program is to initiate and maintain an ongoing and iterative small-scale response testing program by using nationally representative samples or small, thoughtfully selected sites starting in 2023 and continuing with at least two tests per year until 2027. Small-scale tests would allow the Census Bureau to gather initial data about proposed census enhancements while plans for any larger census tests that may take place later in the decade are being developed.

The purpose of this first test is start answering some research questions posed in two project plans in the 2030 Census Enhancement Area 1.1 on Self-Response. The research questions for the 2023 Census Test were chosen to allow more time to test different iterations based on the early results (if needed), work on any necessary messaging for the 2030 Census, and inform, as well as develop, post-processing procedures.

The information collected in this survey will be used within the Census Bureau and incorporated into the efforts for the 2030 Census. The results may be reported in Census Bureau working papers or in peer-reviewed journal articles.

Population of Interest: Results will inform 2030 Census planning, so we are interested in data representing all mailable housing units in the United States (excluding Puerto Rico). Some of our research questions are for specific subpopulations such as unrelated people living in the same housing unit as reported in the 2020 Census, housing units with rural route addresses, housing units that were self-reported vacant in the 2020 Census, and housing units with good administrative records (AR). For the purposes of this test, a housing unit with good AR is one where everyone in the housing unit can be matched to AR using the protected identification key (PIK). We will use stratification and sorting in our sampling methods to be certain we have these subpopulations represented. The 2023 Census Test will be available in English and Spanish.

Timeline: The internet instrument will be available to collect responses for a little over six weeks starting from the day of the first mailing scheduled as February 17, 2023, through the end of the survey period scheduled as April 4, 2023.

Experimental Panel Design: The 2023 Census Test contains five experiments: no mailing dates on mail materials versus including mailing dates, using the presorted USPS First-Class Mail for the final mail package versus USPS Priority Mail, sending mail materials that do not mention the use of AR for nonrespondents versus mail materials that do in one mailing piece or in all mailing pieces, and testing a new address collection screen in the internet instrument. These experiments are not fully crossed.

To summarize all panels in the March 2023 SmaRT Test, see Table 1.

Table 1. Panels by Treatments for the 2023 March SmaRT Test

Panel Number	Panel Description	Mailing Date	Priority Mail	Administrative Records Language	Address Screen
1	Control for Address Screen Testing	No	No	No	2020 Census Address Screens
2	Address Screen Testing	No	No	No	Updated Address Screens
3	Control for all but Address Screen Testing	No	No	No	Randomly Assigned
4	Mailing Dates on Materials	Yes	No	No	Randomly Assigned
5	Priority Mail Only	No	Yes	No	Randomly Assigned
6	Light Administrative Records Language	No	No	Yes (first mailing)	Randomly Assigned
7	Heavy Administrative Records Language	No	No	Yes (all mailings)	Randomly Assigned

Sample: We will select a sample of 15,500 stateside mailable housing units addresses (Self-Response and Update Leave Type of Enumeration Area with good addresses as determined by the 2020 Census Address Composition Specification) from the 2020 Census Unedited File. The frame will exclude housing units selected for the American Community Survey and its supplements in 2022 and 2023 as well those selected for the 2020 Census User Experience Survey.

The overall and panel sample sizes were determined based on the following assumptions:

• An alpha of 0.10, a beta of 0.20, and a design effect of 1.5.

• A detectable difference of 5 percentage points.

• A login rate of 30 percent, based on login rates from the October 2015 small-scale test.

• The rate of addresses requiring clerical matching being 20 percent, based on 2020 Census data and subject matter expertise.

• A desire to analyze at least 20 responses from rural route addresses for both versions of the address screen.

• A desire to analyze at least 20 responses from self-reported vacants for the new vacant reporting path.

Prior to sampling, addresses will be placed in one of the following strata:

1. Unrelated households: housing units where the 2020 Census response included someone unrelated to the householder on the roster.

2. Rural route addresses: housing units not in (1) that are identified as rural route addresses in the Master Address File by the 2020 Census Address Composition Specification.

3. Vacant housing units: housing units not in (1) or (2) that were vacant in the 2020 Census.

4. All remaining housing units: housing units not in (1), (2), or (3).

Within each stratum, address will be sorted by geographic identifiers, a flag indicating that 2020 Census roster members were able to be assigned PIKs, and other demographic and operational data flags. A systematic random sample will be taken, and each housing unit selected will be assigned a random number. After sorting by the random number, the housing units will be sequentially assigned to the panels. Housing units that are not in the address screen testing panels will then again be sorted by random number and sequentially assigned to one of the two address screen treatments. All housing units in the address screen testing control panel are assigned the 2020 Census address screens while those in address screen testing panel are all assigned the new address screens.

Recruitment: We plan to conduct this test with a Census Day of March 1, 2023, and a maximum of four mailings to each housing unit. The first mailing is the initial letter. The initial letter content will differ in the first mailing for the treatment and control groups. The second mailing is a reminder letter that will be different for the treatment and control groups. The third mailing is a reminder postcard. Again, the content will differ. The fourth and final mailing will be a letter that contains a paper questionnaire and return envelope. The content will differ by treatment group. The survey will close out April 4, 2023, approximately six weeks after the first mailing. See Appendix 3 through 9 for mail materials.

Survey Administration: All sampled housing units will be mailed invitations to respond online, and a paper questionnaire will be sent in the fourth mailing to nonresponding households. The online questionnaire will be administered using the survey platform Qualtrics. Qualtrics has a FedRAMP Moderate approval and a Census Authority to Operate to collect T13 data. Respondents will receive the link to the online survey on all mailing materials.

Questionnaire: The online instrument specification and the paper questionnaires are attached (see Attachment 1: 2023 Census Test Online Questionnaire Specification, and Attachment 2: 2023 Census Test Paper Questionnaire) with branching identified in the comments for the online questionnaire. This questionnaire is an abbreviated version of the 2020 Census questionnaire where demographics are only asked for Person 1, which is ideally the respondent. The online instrument has an extra experimental question to determine if the respondent is counting everyone in the household.

Informed Consent: The Census Bureau will conduct the 2023 Census Test under the authority of Title 13 United States Code Sections 141 and 193. All respondents who participate in the 2023 Census Test will be informed that the information they provide is confidential under Section 9 that law. Additionally, respondents will be advised that Section 221 of this same law makes participation mandatory. All collected information that identifies individuals will be held in strict confidence according to the provisions of Title 13 United States Code, Section 9.

Incentive: There is no incentive in this survey.

Length of Interview: We plan on contacting a sample of 15,500 addresses with up to four mailings each. We estimate users will spend 10 minutes on average completing the survey and approximately 5 minutes reading mailing materials. The burden estimates in the table below are an upper bound and based on a theoretical 100 percent response, rather than our expected response rate. The total estimated respondent burden for this study is approximately 7,750 hours, assuming everyone reads the mailing material and answers the survey.

Table 2. Total Estimated Burden

	Max Number of Respondents	Max Number of Occurrence	Time for each Occurrence	Estimated Burden
Reading mail invitations	15,500	4	5 minutes	5,167 hours
Completing survey	15,500	1	10 minutes	2,583 hours
Maximum Total Estimate				7,750 hours

The following documents are included as attachments:

Attachment 1: 2023 Census Test Online Questionnaire Specification

Attachment 2: FAQ insert for Panels 1-5 Mailing 1

Attachment 3: FAQ insert for Panels 1-5 Mailing 4

Attachment 4: FAQ insert for Panels 6-7 Mailing 1

Attachment 5: FAQ insert for Panels 6-7 Mailing 4

Attachment 6: Invitation letter for Panel 1 (Mailing 1)

Attachment 7: Invitation letter for Panel 2 (Mailing 1)

Attachment 8: Invitation letter for Panel 3 (Mailing 1)

Attachment 9: Invitation letter for Panel 4 (Mailing 1)

Attachment 10: Invitation letter for Panel 5 (Mailing 1)

Attachment 11: Invitation letter for Panel 6 (Mailing 1)

Attachment 12: Invitation letter for Panel 7 (Mailing 1)

Attachment 13: Reminder letter for Panel 1 (Mailing 2)

Attachment 14: Reminder letter for Panel 2 (Mailing 2)

Attachment 15: Reminder letter for Panel 3 (Mailing 2)

Attachment 16: Reminder letter for Panel 4 (Mailing 2)

Attachment 17: Reminder letter for Panel 5 (Mailing 2)

Attachment 18: Reminder letter for Panel 6 (Mailing 2)

Attachment 19: Reminder letter for Panel 7 (Mailing 2)

Attachment 20: Reminder postcard for Panel 1 (Mailing 3)

Attachment 21: Reminder postcard for Panel 2 (Mailing 3)

Attachment 22: Reminder postcard for Panel 3 (Mailing 3)

Attachment 23: Reminder postcard for Panel 4 (Mailing 3)

Attachment 24: Reminder postcard for Panel 5 (Mailing 3)

Attachment 25: Reminder postcard for Panel 6 (Mailing 3)

Attachment 26: Reminder postcard for Panel 7 (Mailing 3)

Attachment 27: Final letter for Panel 1 (Mailing 4)

Attachment 28: Final letter for Panel 2 (Mailing 4)

Attachment 29: Final letter for Panel 3 (Mailing 4)

Attachment 30: Final letter for Panel 4 (Mailing 4)

Attachment 31: Final letter for Panel 5 (Mailing 4)

Attachment 32: Final letter for Panel 6 (Mailing 4)

Attachment 33: Final letter for Panel 7 (Mailing 4)

Attachment 34: Questionnaire for Panel 1 (Mailing 4)

Attachment 35: Questionnaire for Panel 2 (Mailing 4)

Attachment 36: Questionnaire for Panel 3 (Mailing 4)

Attachment 37: Questionnaire for Panel 4 (Mailing 4)

Attachment 38: Questionnaire for Panel 5 (Mailing 4)

Attachment 39: Questionnaire for Panel 6 (Mailing 4)

Attachment 40: Questionnaire for Panel 7 (Mailing 4)

Attachment 41: English phone script for Panels 1-5

Attachment 42: English phone script for Panels 6-7

Attachment 43: Spanish phone script for Panels 1-5

Attachment 44: Spanish phone script for Panels 6-7

The contact person for questions regarding data collection and the design of this research is listed below:

Kelly Mathews

Decennial Statistical Studies Division

U.S. Census Bureau

Room 4H472A

Washington, D.C. 20233

(301) 763-5639

kelly.m.mathews@census.gov

The U.S. Census Bureau reviewed this data product for unauthorized disclosure of confidential information and approved the disclosure avoidance practices applied to this release. CBDRB-FY23-DSSD007-0001

Page 6 of 6