2008 DR supporting statement_Part B with resp rates

Part B. Collections of Information Employing Statistical Methods

1. Universe and Respondent Selection

NRFU Reinterview (RI) Workload

For the 2008 Census Dress Rehearsal, the total Nonresponse Followup (NRFU) workload is estimated at 285,000 cases. The response rate we expect to obtain during the mailout/mailback operation is 50% in Stockton, CA and 55% in Fayetteville, NC, while a response rate of 46% is expected during the update/leave operation that is only conducted in our Fayetteville office. The NRFU Reinterview (RI) is the only activity employing statistical methodology for the 2008 Census Dress Rehearsal. Approximately five percent of households in the NRFU universe will be reinterviewed as part of our procedures for quality control. An enumerator who did not conduct the original interview will visit each household selected for reinterview and ask the questions again in order to verify the information that was collected previously. The reinterviews, which will be conducted using HHC, are expected to take 10 minutes each -- the same length of time as the original interview.

NRFU RI Data Collection Procedures

The local census office staff will attempt to contact all cases for which a telephone number is available. If after three telephone attempts the RI clerk has been unable to speak with a respondent, the case will be sent to the field to be enumerated. A separate RI field staff will conduct personal visits.

Those households that did not provide a telephone number will be assigned to a RI enumerator to be completed by a personal visit. If a household respondent cannot be located, then the RI enumerator may, in some cases, visit a proxy respondent to collect household data.

2. Procedures for Collecting Information

During the first two weeks of April 2008, we will identify the addresses for which we have not received a census response. Enumerators will visit each address to determine the Census Day housing unit status and complete a questionnaire using their HHC. Enumerators also will complete a census questionnaire for addresses that they find that are not shown on their assignment list.

The NRFU workload will continually be updated to remove addresses for which a census response has been received (i.e. Late Mail Returns) and add addresses for mail returns that do not constitute an acceptable census return (i.e. Blank Mail Returns).

During the course of NRFU, we will sample a specified number of cases for each enumerator to visit during the NRFU Reinterview (NRFI RI) operation. The NRFU RI operation is the quality control component for NRFU. It is designed to ensure that the production enumerator followed field procedures and to identify enumerators who intentionally or unintentionally produce data errors. The NRFU RI operation is conducted concurrently with production NRFU, beginning one week after the start of NRFU and ending about one week after the finish of NRFU.

Cases selected for NRFU RI will either go to telephone reinterview, if a telephone number is available, or to the field for personal visit. Cases that go to telephone reinterview for which we are unable to contact a respondent (after three telephone attempts) will be reassigned to the field for personal visit follow up. Completed RI cases will be matched to the original NRFU case via the Matching Review and Coding System (MaRCS). Cases will first go through an automated match. Cases that do not pass the automated match will go to the National Processing Center (NPC) to undergo clerical matching. Cases that do not pass clerical matching will be deferred to the LCO for final adjudication.

3. Methods to Maximize Response

The 2008 Census Dress Rehearsal will use a combination of mailing strategies that historically have been shown to improve response rates, (e.g., multiple first-class mailing of the questionnaires, replacement questionnaires and user-friendly design features such as easier-to-read fonts). We plan to use a multiple mailing strategy similar to the one used in the mailout/mailback areas during the 2006 Census Test. It will consist of advance letter, questionnaire, and reminder (postcard and Bilingual letter). Please reference attachments A through D for copies of all letters, questionnaires, envelopes and other public use forms, respectively, that will be implemented throughout the various dress rehearsal operations. All forms and letters include a statement informing respondents that completing the questionnaire is required by law. Table 1 provides an index of all forms.

• Advance Letter [DX-5(L), English and DX-5(L)(E/S), bilingual] – In early March, the USPS will deliver the advance letter for areas where the USPS delivers the Census questionnaire. In late February, the advance letter will be distributed in areas where census workers deliver the Census questionnaire (mailout/mailback) [DX-5(L)(UL)]. The letter alerts the respondent to look for their questionnaire and explains why it is important for them to participate. Additionally, the letter will explain that their response is mandatory and that their answers will be kept confidential. The advance letters will be mailed in the same envelope [DX-5]. Prior to the For the Group Quarters Advance Visit (GQAV) operation, the Census Bureau will distribute the GQAV advance letter to group quarters facility manager [DX-30(L)FM], health care facilities [DX-30(L)HC], and dormitories [DX-30(L)SH].

• Initial Questionnaire package – Approximately a week later, the USPS will deliver an initial questionnaire package to city-style addresses. The package will consist of a cover letter [DX-16(L)], the initial questionnaire [DX-1 or DX-1(E/S)], and a first class postage-paid return envelope [DX-8A]. In all other areas, census workers will deliver the initial questionnaire package consisting of a cover letter [DX-16(L)(UL), the initial questionnaire [DX-1(UL)], and a first class postage-paid return envelope [DX-6(UL)]. The cover letter will explain the purpose for the dress rehearsal. It also will inform residents that responding to the questionnaire is mandatory and that their responses are confidential. The bilingual questionnaire and letters are all in a “swimlane” format, in which the English version is on the left side of the page, and the Spanish is on the right. The initial questionnaire mailing package also contains an outgoing envelope [DX-6A], or [DX-6(UL)] for non-city style addresses.

• Thank You/Reminder Postcards, and Bilingual Reminder Letter – In an effort to encourage participants to respond by Census Day (April 1, 2008), a blanket reminder postcard [English, DX-9; Bilingual DX-9(L)(E/S); Update Leave areas [DX-9(UL); outgoing envelope [DX-6C]; Update Leave outgoing envelope [DX-8C] will be delivered via USPS approximately a week after the initial questionnaire packages are delivered. This postcard will serve as a thank you for respondents who have mailed back their questionnaire, or as a reminder for those who have not mailed one.

• Replacement Questionnaire Package – Since mailing replacement questionnaires increases mailback response rates (and therefore reduces costs), this procedure is now part of our assumed (baselined) plan for the 2010 Census. Households will receive an English-only replacement questionnaire if we have not received a completed initial questionnaire by a pre-determined data. In addition to the questionnaire, the replacement questionnaire package contains a cover letter [DX-17(L), an outgoing envelope [DX-6C] and a return envelope [DX-8C]. All replacement questionnaires will be in English.

4. Testing of Procedures

The procedures for the 2008 Census Dress Rehearsal were tested in the previously described series of iterative tests conducted during this decennial cycle, (e.g. two national mail-out/mail-back tests (2003 National Census Test and 2005 National Census Test), and two site tests (2004 Census Test and the 2006 Census Test).

5. Contacts for Statistical Aspects and Data Collection

Frank Vitrano, Chief, Decennial Management Division, (301) 763-3961 or via Internet at [email protected].

Table 1. Index of 2008 Census Dress Rehearsal Forms

Description	Form #
Letters - (Contained in Attachment A)
Advance Letter (English)	DX-5(L)
Advance Letter Bilingual—English/Spanish	DX-5(L)(E/S)
Advance Letter Update Leave (UL) Areas	DX-5(L)(UL)
Reminder Postcard (English)	DX-9
Reminder Letter Bilingual—English/Spanish	DX-9(L)(E/S)
Reminder Postcard Update Leave (UL) Areas	DX-9(UL)
Cover Letter—Be Counted—English	DX-10(L)
Cover Letter—Be Counted - Spanish	DX-10(L)(S)
Cover Letter—Be Counted - Chinese	DX-10(L)(C)
Cover Letter—Initial Mailing	DX-16(L)
Cover Letter – Chinese Fulfillment	DX-16(L)(C)
Cover Letter--Update Leave and U/L ADDs	DX-16(L)(UL)
Cover Letter—Bilingual Mailing	DX-16(L)(E/S)
Cover Letter—Replacement Mailing	DX-17(L)
Group Quarters Advance Visit Letter – Facility Manager	DX-30(L)FM
Group Quarters Advance Visit Letter – Health Care Facilities	DX-30(L)HC
Group Quarters Advance Visit Letter – Student Housing	DX-30(L)ST
Questionnaire – (Contained in Attachment B)
Mailback—English (Initial, U/L, Replacement and Fulfillment)	DX-1
Update Leave (ADD)	DX-1(UL)
Bilingual—English/Spanish (Initial and Fulfillment)	DX-1(E/S)
Chinese Fulfillment (Using latest version we have)	DX-1(C)
Be Counted—English	DX-10
Be Counted—Spanish	DX-10(S)
Be Counted—Chinese	DX-10(C)
Enumeration at Transitory Locations	DX-15
Individual Census Report—English	DX-20
Individual Census Report—Spanish	DX-20(S)
Military Census Report—English	DX-21
Envelopes – (Contained in Attachment C)
Outgoing Envelope for Advance Letters and Reminder Letter	DX-5
Outgoing for Initial Mailing	DX-6A
Outgoing for Update Leave (U/L)	DX-6(UL)
Outgoing for English Fulfillment	DX-6B
Outgoing for Bilingual Fulfillment	DX-6(E/S)
Outgoing for Replacement Mailing and English Fulfillment	DX-6C
Outgoing for Chinese Fulfillment	DX-6(Chinese)
Return for Initial Mailing (2nd window for postal tracking barcode)	DX-8A
Return for Initial Bilingual (2nd window for postal tracking barcode)	DX-8B
Return for U/L, U/L ADDs, Replacement, and Fulfillment (English and Chinese)	DX-8C
Return for Bilingual U/L, U/L ADDs, and Fulfillment	DX-8D
Outgoing for Be Counted—English	DX-12
Outgoing for Be Counted—Spanish	DX-12(S)
Outgoing for Be Counted—Chinese	DX-12(C)
Return for Be Counted - English	DX-14
Return for Be Counted - Spanish	DX-14(S)
Return for Be Counted - Chinese	DX-14(C)
Outgoing/Return for Individual Census Report--English	DX-40
Outgoing/Return for Individual Census Report—Spanish	DX-40(S)
Other Public Use Forms – (Contained in Attachment D)
Flashcard Booklet – English and	DX-1(F)
Notice of Visit	DX-26
Privacy Act Notice	DX-31P
Language Assistance Guide	DX-60
Language ID Flashcard	DX-3309

-2-