Nonsubstantive Change (Audiotaping & Debriefings)

Audio Taping and Debriefing Questions for a Sample of 2010 Census Interviews

Non-Substantive Change Request

OMB Approval Number (0607-0919)

June 9, 2010

The U.S. Census Bureau requests a non-substantive change to the 2010 Census Information Collection (OMB approval number 0607-0919). This request consists of two brief follow-on activities to the currently approved information collection that will have a minimal affect on the public (the total number of additional burden hours is about 318).

The first activity pertains to the audio taping of a sample of 2010 Census data collection interviews for research purposes. The 2010 Census clearance (and clearances closely related to the 2010 Census, such as those for 2010 Census Coverage Measurement) accounts for the burden hours associated with the data collection interviews that the Census Bureau plans to audio tape. This request documents additional burden incurred while obtaining consent from respondents to audio tape their interviews.

The second activity pertains to asking debriefing questions at the end of some audio taped data collection interviews. This request also documents additional burden incurred while asking debriefing questions at the end of some taped interviews, and later by phone, if any information critical to the evaluation was not captured during the interview or the debriefing.

The observation and audio taping of data collection interviews coupled with the option of asking debriefing questions at their conclusion yield a rich supplement to the final 2010 Census data. The Census Bureau’s researchers and planners will use this information to gain insight into data collected during the 2010 Census and to improve current data collection methods for the 2020 Census. For example, the audio taping of data collection interviews enables the Census Bureau to conduct behavior coding studies. Behavior coding is a method in which researchers examine exchanges between interviewers and respondents by assigning codes to verbal behaviors captured on audio recordings. The results of these types of studies can lead to a better understanding of patterns observed in interviewer collected census data plus recommendations for how to improve question wording, interviewer methodology, and interviewer training. Moreover, audio recordings of data collection interviews allow observers to focus on interaction between the interviewer and the respondent rather than on writing down their observations on paper.

Posing debriefing questions at the completion of a data collection interview allows researchers to gather information on specific topics. The debriefing questions associated with this request revolve around two broad areas of interest to the Census Bureau: (1) sources of population coverage error, and (2) barriers to collecting census data from non-English speaking respondents. While observers will have specific topics to ask about that fall under each of these two areas, the wording of the debriefing questions will vary. On one hand, some observers will tailor debriefing questions based on answers given by a respondent during an interview. On the other, observers will tailor debriefing questions based upon the language predominantly spoken by the respondent. Some observers will ask debriefing questions at the end of every observed interview, while other observers may ask debriefing questions at the end of only some of their observed interviews.

The Census Bureau plans to observe a wide range of respondents as they provide their data to interviewers. Most interviews observed will be in English or Spanish; however, the Census Bureau aims to observe approximately 60 interviews conducted in each of the following languages: Arabic, Chinese, Korean, Portuguese, Russian, and Vietnamese. Moreover, the Census Bureau plans to select observation sites containing population centers for the following race/ethnic groups: American Indian; Alaskan Native; Native Hawaiian; Hispanic; Asian (Chinese); Non-Hispanic White; Black; and White (Middle Eastern) and one generalized site. Some of the interviews in the Chinese, Hispanic, and Middle Eastern sites may also be conducted in those non-English languages.

Procedures for Audio Taping Interviews

For audio taping of data collection interviews, researchers will accompany interviewers, observe actual production interviews, and -- when given permission by respondents -- audiotape the interviews. These observations will occur during the collection of data for the 2010 Census.

Training developed by Census Bureau staff outlines the following standard procedures for researchers to follow when accompanying interviewers and asking respondents for permission to audio tape their interviews: After the production interviewer reads the introduction and the confidentiality statement, he or she will introduce the researcher and say that he or she has a question for the respondent.  The researcher will then ask the respondent whether he or she consents to the audio taping of the interview.  If the respondent says, “yes,” then the researcher will start taping the interview and will repeat the request to tape the interview, capturing on tape the respondent’s permission to record the interview.

The following script exemplifies how the Census Bureau envisions the exchange among the interviewer, researcher, and respondent in regard to obtaining the respondent’s permission to audio tape the interview (please note that the interviewer incorporates the introduction of the observer and the request to audio tape the interview into the first couple of interview questions):

Interviewer: Hello, I’m (interviewer’s name) from the U.S. Census Bureau. This is (name of observer). (Show IDs) Is this (address)?

Interviewer: I’m here to complete a Census questionnaire for this address. It should take about 10 minutes. (Hand respondent Information Sheet.) The first part explains that your answers are confidential. I’ll refer to the other parts later.

Observer: Is it OK with you if I tape record this interview? It is confidential.

Observer may alter approach as appropriate: Your answers and opinions will help us make future censuses better. Because it would be difficult for me to remember all of the things you say, I’ll be tape recording this interview. If you’re ready, I’ll let (interviewer’s name) get started with the first part of the interview now.

Note: If respondent objects to the interview being tape recorded and refuses to complete the interview, observer can proceed to next line as she deems appropriate to do so. Observers are to respect the wishes of the respondent and should not push the respondent to do something he does not want to do.

Observer: I understand your concerns about being taped, but I’d like to let you know that your answers to the questionnaire and your personal information are confidential. Your cooperation would really help us, so may I please record this interview?

NOTE: If respondent objects and will not complete interview, observer will not turn on tape, but will take notes.

Observer: Thank you. I will start the tape now and then ask for your consent again. (Start the tape) Is it okay with you if I tape record this interview? (Respondent gives permission to audio tape interview)

Interviewer: Did you or anyone in this household live or stay here on April 1, 2010?

NOTE: Interviewer continues as instructed during interviewer training.

 

Procedures for Asking Debriefing Questions

Observers may pose debriefing questions to respondents on one of two broad topics after an interviewer completes the data collection interview: (1) sources of population coverage error, and (2) barriers to collecting census data from non-English speaking respondents.

Population Coverage Error

To identify sources of coverage error, the Census Bureau plans to have researchers conduct an immediate followup interview consisting of debriefing questions based on the researcher’s observations during the data collection interview, if there were cues during the interview of a potential coverage problem or complex household structure that would be masked by the answers to the relationship questions. Researchers will develop adhoc debriefing questions based on ambiguities in a respondent’s answers pertaining to coverage, residence rules, and housing unit structure. The goal of this research is to identify types and sources of coverage error, the characteristics of households and persons affected by coverage error, and whether this appears to vary by race/ethnicity. Observers will be watching for a wide range of sources of error including, but not limited to question wording, respondent-interviewer interaction, accuracy of residence coding, cultural issues, race, literacy, language, race/ethnic factors, and so on. If a later review of the interview/debriefing transcript indicates that some critical coverage or household structure issues were not fully resolved during the interview or the debriefing, the observer may conduct a brief telephone followup with the respondent to ask for the critical information. The Census Bureau anticipates that this telephone followup will occur rarely.

Examples of unscripted debriefing questions that an observer may pose to a respondent based upon answers given by the respondent during a data collection interview are as follows:

1. You hesitated when telling us how many people live or stay here most of the time. Will you tell me more about what you were thinking when you were answering that question?

2. When reporting how many people live or stay here most of the time, you originally said X number of people, but then you changed your answer to Y. Why did you change your answer?

3. You mentioned that <NAME> sometimes stays somewhere else. Can you tell me more about the other place?

4. I’m confused about where <NAME> stays most of the time. You said that <NAME> stays in <OBSERVER REPEATS INFORMATION HEARD DURING INTERVIEW>. Is that correct?

5. You said that <NAME> was away at <LOCATION>. How often does <NAME> stay at <LOCATION>?

Barriers to Collecting Census Data from Non-English Speaking Respondents

To learn more about difficulties associated with collecting census data from non-English speaking households, the Census Bureau plans to have researchers conduct an immediate followup interview consisting of debriefing questions that are tailored to respondents who speak one of the following seven languages: Arabic, Chinese, Korean, Portuguese, Russian, Spanish, and Vietnamese. For purposes of comparison, the same questions will be asked of respondents who speak English.

Researchers will pose debriefing questions to respondents on the following topics:

• Experience with past United States (U.S.) censuses;

• Views of government-sponsored activities;

• Experience with other survey interviews;

• Issues encountered with the census process; and

• Factors impacting communication between the interviewer and the respondent.

Examples of debriefing questions that the Census Bureau plans to pose to respondents are below. These questions will be posed in the language of the respondent.

Experience with past U.S. censuses

1. Have you responded to the U.S. Census before?

2. Was there a difference between last time and this time?

3. Did the use of written materials in <RESPONDENT’S LANGUAGE> make any difference in your comprehension of the process and questions?

Views of government-sponsored activities

1. Did you know that the census is a government related activity?

2. Who do you think uses this information and for what purpose(s)?

3. How does your opinion(s) (in terms of who uses the information and for what purposes) influence your participation in the census?

Experience with other survey interviews

1. Have you participated in a census in any other country?

2. Did any previous experience with the census in <respondent’s country> affect your experience with the interview today?

3. Have you ever been surveyed in the United States for other reasons?

Issues encountered with the census process

1. Do you have any difficulties reading/understanding the mail that you receive? Please explain them to me.

2. If you see mail in English or another language that you do not understand, what do you do?

3. How often do you read/go through your mail? Who usually gets the mail in your household?

4. Do you remember receiving the census form in the mail several months ago? If yes, what was your reason for not filling it out? What made you cooperate with the interviewer today?

5. Have you heard about the census this year? If yes, how did you hear about the census?

Factors impacting communication between the interviewer and the respondent

1. At any time, did you feel that the questions were unclear? What was the cause of the confusion?

2. Did you feel comfortable with the interpreter? Was it helpful to have an interpreter? In what ways?

3. Was there anything about this interview process that you found unusual or surprising?

Estimate of Burden Hours

Collection	Total # of Respondents	Estimated Response Time	Estimated Burden Hours
Audio Taping and Debriefing Questions	1270	15 minutes	318 hours