2013 Change Req-D4

OMB No. 0920-0314 2013-Change-Req.doc

Request for Approval of a Non-Substantive Change:

NATIONAL SURVEY OF FAMILY GROWTH, 2013-2015

OMB No. 0920-0314

(expires April 30, 2015)

Contact Information:

William D. Mosher, Ph.D., Statistician

Project Officer, National Survey of Family Growth

National Center for Health Statistics/CDC

3311 Toledo Road, Room 7421

Hyattsville, MD. 20782

301-458-4385

301-458-4034 (fax)

[email protected]

January 11, 2013

Request for Approval of a Non-Substantive Change:

NATIONAL SURVEY OF FAMILY GROWTH, 2013-2015

A1. Circumstances Making the Information Collection Necessary

This request is for a nonsubstantive change to an approved data collection (OMB No. 0920-0314, expires April 30, 2015), the National Survey of Family Growth (NSFG) conducted by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention (CDC).

On April 26, 2012, the NSFG was approved to collect data for the 3 years ending April 30, 2015. That request for clearance stated that:

“We are seeking approval to:

• Conduct the NSFG for the next 3 years;

• Delete some questions from the questionnaires to reduce burden to the approved 60 minutes for men and 80 minutes for women; and

• Add or modify a small number of questions in late 2013 using a nonsubstantive request. These questions would be similar to the questions and topics contained in the current questionnaire.”

This request specifies those non-substantive questionnaire changes.

This request also asks for permission to conduct an experiment to test a $60 incentive compared with our current $40 incentive. As explained below, our $40 incentive has been used for 10 years now, since 2002, but it is gradually losing its effectiveness as a device to gain the attention of respondents. Our weighted response rate has fallen from 79% in 2002 to 73% currently. And for some groups (e.g., white women), it has dipped to 70% for the first time in the NSFG’s history.

In early 2015, a full new clearance request will be submitted to continue interviewing for the period May 2015-May 2018. The NSFG web site is at http://www.cdc.gov/nchs/nsfg.htm.

Changes

This change request seeks approval for nonsubstantive changes to the questionnaires for NSFG interviewing from September 2013 – April 2015, to address the interests of the 9 existing funding agencies(all in DHHS). These are detailed in Section A.2. Burden is still estimated to be within the approved averages of 60 minutes for males and 80 minutes for females.

A2. Purpose and Use of Information Collection

The National Survey of Family Growth responds to the congressional mandate for NCHS to collect and publish reliable national statistics on “family formation, growth, and dissolution” (Sec. 306(b), paragraph 1(H) of the Public Health Service Act), as well as vital statistics on births and deaths, and a number of aspects of health status and health care. In addition, the NSFG collects data for NCHS and 9 other Department of Health and Human Services (DHHS) programs and agencies.

NCHS alone has published 12 reports on the findings of the 2006-2010 survey; counting outside researchers, more than 70 reports and articles in scientific journals have been published from the 2006-2010 data alone. (Bibliographies of studies based on the NSFG are posted on the NSFG website and updated several times a year.)

To keep the text of this change package brief, we are including several attachments:

Attachment 1 is a spreadsheet showing the questionnaire changes we propose;

Attachment 2 is the Female questionnaire that shows these questionnaire changes in context;

Attachment 3 is the Male questionnaire that shows these questionnaire changes in context;

Attachment 4 is the Agenda for the October 2012 NSFG Research Conference, where we received useful feedback from NSFG data users;

Attachment 5 shows contact information for representatives of the co-sponsoring agencies;

Attachment 6 describes our research on the design of an experiment to test the effects of a $60 vs. a $40 incentive.

After a consultation process in the Summer and Fall of 2012, and the Research Conference in October 2012 (Attachment 4), and a formal meeting of the collaborating agencies on December 17, 2012, we arrived at the list of recommendations for nonsubstantive questionnaire changes shown in Attachment 1. The changes are shown in the context of the questionnaires in Attachments 2 and 3.

The changes are listed in Attachment 1, but the reasons for them are:

• Minor changes that will ensure compliance with DHHS data standards for the Affordable Care Act on questions on disability, primary language, and racial detail;

• Changes to make questions more comparable between men and women;

• Specific agency requests for data related to the Affordable Care Act (an in-depth evaluation of the NSFG questionnaires is being undertaken on this issue for 2015, and any necessary changes will be proposed when a new, full clearance is needed).

A8. Comments in Response to the Federal Register Notice and Efforts to Consult Outside the Agency

In addition to consulting with our new co-sponsoring agencies and programs, we obtained feedback from data users at a very successful NSFG Research Conference held at NCHS in October of 2012 (Attachment 4), and at meetings of the NSFG funding agencies in January and December of 2012. We also met in person with our Atlanta funding agencies in September of 2012, and with the DHHS Office of Population Affairs (OPA) in August 2012 and the National Institute for Child Health and Human Development (NICHD) in Nov 2012. These consultations and our own analyses of the data for the dozen NCHS reports so far from the 2006-10 survey led to the changes described above. Contact information is given in Attachment 5 for representatives of the co-sponsoring programs and agencies (all in DHHS).

A9. Explanation of any Payment or Gift to Respondents.

In the 2002 and 2006-2010 NSFG and the current 2011-15 NSFG, we received IRB and OMB permission to offer a $40 cash incentive as a token of appreciation to respondents. Our response rate in 2002 was 79.0%. In 2006-2010, it was 76.6%. So far in 2011-2012, it is 72.7%. Our interviewers are spending more hours working each case, and visiting each case more times than in 2006-2010, but still getting slightly lower response rates:

2006-10 2011-2012

Hours of labor per interview: 9.1 hours 9.5 hours

Average no. of visits to

households per interview: 7.1 visits 8.1 visits

Response rate: 76.6% 72.7%

In 2006-10 and currently, each data collection year is divided into four 12-week data collection periods, or “quarters.” These 12-week quarters are divided into two parts:

Weeks 1-10 are called “phase 1,” in which $40 cash is offered as a token of appreciation to all main interview respondents for their participation. (No incentive is offered for the screener, which takes about 3 minutes.) Cases that remain as non-respondents are visited an average of 8 times during these 10 weeks. By the end of these 10 weeks, the response rate is about 56%.

About 92% of all NSFG respondents receive a $40 incentive.

Weeks 11-12 are called “phase 2”: with about a 56% response rate at the end of 10 weeks, a sample of about one-third of the remaining 44% is chosen: one-third of 44% is just under 15% of the original sample. This 15% is offered an $80 incentive for 2 weeks, and the number of hours of interviewer time spent on these cases is tripled, to give the interviewers the best possible chance of finding them at home. About half of these cases, or about 8% of the original sample, are interviewed, and receive the $80 incentive. (NOTE: 15-17 year olds are not offered $80; they are offered the phase 1 amount, which is $40.)

We have shown, however, that the $80 group is different from the $40 group (see Attachments C and N in the April 2012 package): for example, childless women, Hispanic males, high-income males, and those at low risk of HIV are more likely to be in the $80 group. Thus, this group (and the $80 incentive) is important for reducing bias.

Unfortunately, however, our response rates are gradually falling. In 2002, the NSFG response rate was 79.0%. In 2006-10, it was 76.6%. So far in 2011-12, it has been 72.7%, despite an increase in the number of visits to households per interview and in the number of interviewer hours worked per completed interview. Thus, increasing effort is getting reduced returns.

Also of concern is that the response rate for white females is now 70% (Attachment 6, table 1) and response rates for males are just 72%. We hypothesize that the $20 increase in the incentive will produce the following 3 effects:

a) an increase in the response rate of 6-8 percentage points, raising the response rate from 73% to 79-81%;

b) a reduction in interviewer effort per case of about 1 hour per completed interview, sufficient to pay for part of the increased costs of the incentive.

c) recruitment into the sample of under-represented groups, including adult white females, and men (especially Hispanic men and white men).

We expect to have results after 6-12 months of fieldwork, and we will share those results with OMB as soon as we have them. We propose to leave the Phase 2 incentive at $80.

The cost of administering and analyzing an experiment, including printing and distributing multiple versions of forms, letters, brochures, and materials, and the likelihood of interviewers making mistakes in administering the wrong versions to some respondents, is significant, so the experiment will not be continued any longer than is necessary to demonstrate significant results.

Further details are in Attachment 6.

List of Attachments

Attachment 1 Proposed questionnaire changes

Attachment 2 Female questionnaire

Attachment 3 Male questionnaire

Attachment 4 Agenda for the October 2012 NSFG Research Conference

Attachment 5 Representatives of NSFG Co-sponsoring agencies

Attachment 6 Designing an Incentive Experiment for the NSFG

5