OMB Memo

Re-identification Question Pretesting Survey

Submitted Under Generic Clearance for Internet Nonprobability Panel Pretesting

Request: The Census Bureau plans to conduct additional research under the generic clearance for Internet Nonprobability Panel Pretesting (OMB #0607-0978). We will be conducting additional online pretesting of select questions that did not perform well in previous testing. The questions are part of a larger survey designed to measure respondent privacy and confidentiality concerns for the 2020 Census. We are targeting specific questions that performed poorly in two rounds of cognitive interviews previously conducted under the generic clearance for questionnaire pretesting (OMB #0607-0725). Further evidence for the necessity of revisions came from an online pilot test conducted in April under this generic clearance.

Purpose: The purpose of conducting this survey is to measure individuals’ privacy risk tolerance with the goal of informing the privacy loss budgets allowed in mathematical privacy models for decennial data releases. Responses to the questionnaire items provide information about respondent concerns surrounding each of the different pieces of individual and household-level information collected on the 2020 Census and included in statistical summaries released by the Census Bureau. To achieve its research objectives, the survey will need to be administered to a large number of respondents drawn from a nationally-representative sample of households. The information collected in the larger, representative survey will be used by experts within the Census Bureau to help make policy decisions about the differential privacy protections associated with 2020 Census data releases.

This collection is additional pretesting for a set of questions that performed poorly in previous testing. This series of questions asks respondents about the unfamiliar concept of re-identification. In previous testing, we found that respondents had trouble understanding the term re-identification. However, evidence also suggests that respondent comprehension increases when behavioral explanations and accessible examples are part of the question. We plan to conduct a split-panel experiment to test three alternative wordings for this set of questions (See Enclosure I: Alternate Wordings for Re-identification Questions). We used lessons learned from previous testing to inform the design of the alternative questions.

Population of Interest: Residents of the United States and its territories.

Timeline: We intend to conduct this survey with a non-probability sample drawn from the Census Bureau’s Affinity Panel in September 2019. The survey will be open for a period of two weeks after the first email invitation is sent.

Sample: In September 2019, Census Bureau staff will sample 5,000 email addresses from the Census Bureau’s Affinity Panel. The Affinity Panel includes email addresses from individuals who signed up to participate in Census Bureau research through a link on the census.gov website. Based on past use of this frame, we expect a 2% response rate. Our minimum goal is 100 complete responses so that we can compare at least 30 responses for each of the alternative question wordings. We will cap responses at 150 completes.

Recruitment: Respondents will be invited to respond to the online survey by means of a series of emails containing a link to the survey (see Enclosure III: Respondent Privacy and Confidentiality Concerns Email Invitations). For this study, each email address in the sample will receive a maximum of three notification emails:

1. An initial email on a Monday,

2. A reminder email on the following Thursday (if they have not yet clicked on the link to the survey), and

3. A final reminder email on the following Monday with the survey closing the following Friday.

Survey Administration: The questionnaire will be administered online using the survey platform Qualtrics. Respondents will receive an invitation with a link to the survey, which will then take them to the Qualtrics instrument.

Questionnaire: Respondents will be randomly assigned to one of the three variations of the re-identification questions (see Enclosure II: Re-Identification Question Testing Survey). Each respondent will be asked two or three probes to gauge their comprehension of the re-identification questions. The next section of the questionnaire asks questions about respondent privacy concerns. Also embedded in the questionnaire is a usability experiment designed to test different formats for how to list state names and abbreviations in a drop down box. Respondents will be randomly assigned to one of three treatments.

Informed Consent: In the survey invitation materials, we will inform participants that their response is voluntary and that the information they provide is confidential and will be accessed only by employees involved in the research project. Additional required notices about confidentiality and privacy are included on the first page of the survey.

Incentive: Participants will not receive any payment for their participation in the survey.

Length of Interview: We estimate that the survey will take an average of 7 minutes for each complete response. We will allow up to 150 complete responses for a total of about 17.5 hours. Each email will take about 1 minute to read. Each person could get up to three emails and we will send emails to 5,000 email addresses for a total burden of 250 hours. The total estimated burden of this research is 267.5 hours.

Table 1. Total Estimated Burden

Category of Respondent	No. of Respondents	Participation Time	Burden
Reading email invitations	5,000	3 minutes	250 hours
Survey	150	7 minutes	17.5 hours
Totals			267.5 hours

The following documents are included as attachments:

Enclosure I: Alternate Wordings for Re-identification Questions

Enclosure II: Re-Identification Question Testing Survey

Enclosure III: Respondent Privacy and Confidentiality Concerns Email Invitations

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

Aleia Clark Fobia

Center for Behavioral Science Methods

U.S. Census Bureau

Room 5K024E

Washington, D.C. 20233

(301) 763-4075

aleia.yvonne.clark.fobia@census.gov

Page 3 of 3