Part_B_-_SUPPORTING_STATEMENT_ 7-16-13 CENSUS AND OMB comments

SUPPORTING STATEMENT

U.S. Department of Commerce

U.S. Census Bureau

National Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR)

Pre-Screener Test

OMB Control Number 0607-<XXXX>

Part B – Collections of Information Employing Statistical Methods

Question 1. Universe and Respondent Selection

The National Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR) is an address based sample. For the pre-screener test, a sample of 15,000 cases will be selected from the Master Address File (MAF) which is maintained by the U.S. Census Bureau. The frame has no available telephone numbers. The purpose of this test is to determine if the new methodologies introduced in the test will help us obtain telephone numbers that reach our sample addresses.

Three panels of 5,000 pre-screener cases will be selected from New Jersey and South Dakota by working closely with the sponsor, the U.S. Fish and Wildlife Service. We will calculate the expected sample sizes for the 2016 survey and then split the 15,000 test cases between the two states based on the percentage of 2016 sample cases in each state. The base weights will be calculated at the FIPS county level. The base weight derives from the county’s sampling interval. We will calculate response rates based on the number of questionnaires returned with phone numbers reaching our sample addresses. These sample sizes within the two states and the three different treatment groups will allow us to detect changes in the overall response rates between the groups as low as 1.6% at the 90% confidence level with at least 85% power. We expect a fifty percent response rate for the pre-screener operation. Because we will not release any survey estimates, a fifty percent response rate is sufficient for the purposes of this test.

Question 2. Procedures for Collecting Information

The Census Bureau will conduct the FHWAR Pre-Screener test in two parts. The first part consists of mail operation that requires a household member to complete either a mail or Internet questionnaire. The mail operation will include three panels and up to three mailing per panel to improve nonresponse rates. See the table below.

Pre-Screener Panels	Mode	Sample Size
1	Letter and Paper Pre-screener	5,000
2	Letter and Internet Push	5,000
3	Letter including Internet Invite and Paper Pre-screener	5,000

The first panel will receive a letter and a self-administered paper pre-screener questionnaire. The letter will ask a household member to complete the paper questionnaire and to return it by mail to the Census Bureau.

The second panel will receive a letter with an Internet invitation for a household member to complete the pre-screener on the Internet. We refer to this panel as the Internet Push Panel. If the household does not respond after the second (follow-up) mailing, the third (final) mailing will include a paper questionnaire which will allow the respondent to complete the survey via Internet or paper questionnaire.

The third panel will receive a letter, paper questionnaire, and information on how to complete an interview by Internet. In this panel, the household member is given a choice of completing the pre-screener paper questionnaire or responding via the Internet.

The second part of the test includes delivering the completed pre-screener data to the Census Bureau’s Jeffersonville Contact Center who will conduct a telephone operation using a paper questionnaire to verify that the phone numbers collected from the mail and Internet pre-screener either reach, or do not reach, the sample addresses.

Interviewers will complete a short self-study that explains the pre-screener telephone operation. The self-study will include a laminated job aid (FH-JA1- See Attachment 12) that contains the standard script used during the telephone call, and instructions for the number of contact attempts and how to record the outcome of the cases.

The questionnaires will be separated by state because New Jersey and South Dakota fall into two different time zones. Calls will be made every three days and at different times of the day.

Interviewers will record up to three contacts per case for outcomes such as:

• Made contact with a potential respondent but interview not completed

• Left a message on an answering machine

• Potential respondent refused

• Determined language problem

• The phone number is out-of-service

Interviewers will not record outcomes such as:

• Ring no answers

• Busy signals

• Fax machines

• Hang up before making contact

until the last week of interviewing. Outcomes will be recorded on the back of the questionnaire in the “For Office Use Only” box at the bottom of the questionnaire. Final outcome will be recorded once the interview is completed or up to three contact attempts have been made

Of the 7,500 paper and Internet questionnaires returned, we expect a 90% response rate during the telephone operation. The telephone operation will be conducted from September 23 to

October 14, 2013.

The estimation procedure for the FHWAR Pre-Screener Test will follow the usual statistical principles used for other surveys. We will compare the three data collection methods using weighted and unweighted response rates within the two states.

There are no unusual problems requiring specialized sampling.

These sample sizes within the two states and the three different treatment groups will allow us to detect changes in response rates between the groups as low as 1.6% at the 90% confidence level.

This is a one-time data collection to test methodologies that could be utilized in future iterations of the survey.

Question 3. Methods to Maximize Response

Each panel will receive a letter with wording that will encourage household respondents to complete the survey. We plan to mail the initial pre-screener packages in mid August. Two weeks after the initial mailing, we will mail a follow-up pre-screener package to those households that have not returned their paper questionnaire or completed the Internet Survey. Two weeks after the follow-up mailing, we will mail a final package to households that have not yet responded to request their participation. Households that receive the follow-up mailing or the final package will remain in the same panel throughout the test. (Panel One receives the paper questionnaire, Panel Two (the Internet push) will receive an Internet invite for the initial and follow-up mailings, and a choice of completing the survey via Internet or paper in the final mailing, and Panel Three will receive a choice to complete the paper questionnaire or respond via the Internet.)

Once we have completed both parts of the test, the three panels will help us determine whether there are viable alternatives to collecting phone numbers through a simple pre-screener questionnaire. First, we would like to decide if the pre-screener is feasible. That is, if we can obtain a 40 - 50% response rate for the mail operation (any combination of paper and Internet questionnaires) and if 75% of the telephone numbers obtained reach our sample addresses, then we will consider the pre-screener a viable methodology for use in the 2016 FHWAR. Obtaining phone numbers from the respondents themselves could potentially increase future contact and response rates.

If we determine that the test is feasible, we want to determine which pre-screener method works the best in terms of response rates, data quality, timeliness, and cost. We will compare the response rates for the three panels. The comparisons will help us answer two questions: 1) Did we collect valid phone numbers? And 2) Which method yielded the highest response?

If either mode improves our success in obtaining accurate telephone numbers for sample households, we may improve contact and response rates and reduce the costs for conducting the 2016 FHWAR. A mail pre-screener operation is less expensive than the telephone research operation we conducted for the 2011 FHWAR, and we could potentially conduct more interviews in CATI with accurate phone numbers provided by household members.

Question 4. Tests of Procedures or Methods

The FHWAR Pre-Screener is a test to determine if a new methodology can improve our contact rate in CATI. This is the first time we have introduced a short pre-screener for the primary purpose of collecting contact information. The questionnaire contains approximately 10 questions that have been taken and modified from the 2011 FHWAR screener. The Census Bureau’s Center for Statistical Research and Methodology will conduct an expert review to identify any content problems prior to production.

Question 5. Contacts for Statistical Aspects and Data Collection

You may consult the following Census Bureau individuals for information regarding sample

design and data collection:

Sample Design: David Hornick 301-763-4183

Aref Dajani 301-763-1797

Data Collection: Denise Pepe 301-763-3785

Aniekan Okon 301-763-6192

List of Attachments

Attachment 1 Table of materials for the National Survey of Fishing, Hunting, and Wildlife-Associated Recreation (FHWAR) Pre-Screener Test.

Attachment 2 FHQ1 – Panel 1 initial letter sent with the paper questionnaire.

Attachment 3 FH-W1 – Panel 2 initial letter requesting that a household member conduct the survey by Internet.

Attachment 4 FHW-QW1 – Panel 3 initial letter providing the choice to complete the questionnaire on paper or Internet. Paper questionnaire is included in mail package.

Attachment 5 FHQ2 – Panel 1 follow-up letter sent with the paper questionnaire.

Attachment 6 FH-W2 – Panel 2 follow-up letter requesting that a household member conduct the survey by Internet.

Attachment 7 FHW-QW2 – Panel 3 follow-up letter providing the choice to complete the questionnaire on paper or Internet. Paper questionnaire is included in mail package.

Attachment 8 FHQ3 – Panel 1 final letter sent with the paper questionnaire.

Attachment 9 FH-W3 – Panel 2 final letter requesting that a household member conduct the survey by Internet but also includes a paper questionnaire to provide the respondent a choice to complete the survey either by paper or Internet questionnaire.

Attachment 10 FHW-QW3 – Panel 3 final letter providing the choice to complete the questionnaire on paper or Internet. Paper questionnaire is included in mail package.

Attachment 11 FH-PS(T) – Questionnaire that accompanies Panel 1, Panel 2 (third mailing) and Panel 3 mail-outs.

Attachment 12 FH-JA1 – Job Aid that provides the script the interviewers will use during the telephone operation (part 2 of the pre-screener test).

Created by DSD/SSB Page 5 1/28/2021