Notes to OMB

To: Ivelisse Martinez-Beck, NSECE COTR, OPRE

From: Rupa Datta, NSECE Project Director, NORC at the University of Chicago

Date: January 20, 2011

In this memorandum we provide additional detail regarding the points you have specified for elaboration for the package for OMB clearance for the NSECE field test.

1. References to DHHS in materials for respondents.

We have found one omitted reference to HHS in the telephone version of the Household Screener, and it has been corrected. The corrected version is among the files submitted with this memorandum.

2. References to the Privacy Act in language for informed consent.

We have modified the informed consent statement for each questionnaire to explicitly refer to the Privacy Act by name. The revised statement from the Household Questionnaire is copied below, and included in the attached materials. Similar changes are made for all instruments.

CATI/CAPI Household Questionnaire:

Hello, my name is [NAME] and I am calling from NORC at the University of Chicago. We are conducting a study about the experiences and preferences of parents of children under age 13 with regard to the child care or after-school programs that are available for these children. The study is being paid for by the U.S. Department of Health and Human Services, and is designed to help the government understand how private decisions and public policies affect the supply and demand of child and school-age care in our country. We would like to talk with you for approximately 30 minutes about your children under 13 and the child care that you use or would like to use for them.

Taking part is up to you.  You don’t have to answer any question you don’t want to, and you can end the interview at any time.  We are required by the Federal Privacy Act to develop and follow strict procedures to protect your information and use your answers only for research.  If you have any questions about this survey, I will provide a telephone number for you to call to get more information.

CATI: In order to review my work, my calls are recorded and my supervisor may listen as I ask the questions.  I’d like to continue now unless you have any questions.

CAPI: Parts of this interview may be recorded for quality control purposes. This will not compromise the strict confidentiality of your responses.  May I continue with the recording?

                        R CONSENTS TO PARTICIPATE IN THE SURVEY................................. 1

                        R CONSENTS TO PARTICIPATE IN THE SURVEY BUT DOES NOT WANT

               TO BE RECORDED...............................................................2

FAQs: Included on the back of all letters.

NORC Toll Free Number: 1-8XX-XXX-XXXX

You can call the NORC toll free number to take part in the study, learn more about the study, and hear what you will be asked.

How do I know this is legitimate?

This survey has been approved by the U.S. Office of Management and Budget and issued approval number XXXX-XXXX. Without this number we could not conduct this survey.

Who sees my answers?

Everyone who works on the survey must sign an oath that promises they will never give out identifying information about a respondent. Only a few people who work on this survey ever see any personal information. Answers that could identify you or your program in any way are separated from your other answers. Survey findings are put into summary reports that contain no names or other information that identifies you. This information will be used solely for the purposes of this study and will be destroyed when the study is over. Your name and identifying information will never be released to the public.

How do you protect my information?

Your answers are used for research purposes only. All information that you provide will be kept private to the fullest extent provided by the law. If you have questions about your rights as a study participant, you may call Kathleen Parks, the NORC Institutional Review Board Administrator, toll free, at 1-866-309-0542.

3. Multi-mode Data Collection flow for Households

The NSECE questionnaires include questions about prior week enrollment, schedule and other information, and so require fielding outside of major holiday periods, such as the Christmas/New Years’ weeks and summer sessions. For this reason, we want to conduct interviews in a compact time period in the main study, between the second week of January and the second week of May. Given this restricted window for data collection, there is time for each case to be cycled through two different modes, for example, mail-then-field, or phone-then-field. In the main study, there is not adequate time to try the three-step mail-phone-field sequence without risking excessive changes in eligibility (home address, provision of regular home-based care, presence of children in the household) between screening and interview administration or spending out the data collection window without reaching certain cases. For this reason, the field test also pairs modes together. Based on responses rates and yields found in the literature and in recent NORC experience, we calculate that the cost-minimizing approach is to complete screening and interviews by telephone where possible, screen by mail (with field interviews of eligible) otherwise if possible, and finally screen and interview in-person if necessary.

Although telephone administration is more expensive than field data collection, telephone screening and interviewing can be done on a single call, while an eligible household screened by mail must be interviewed on a second contact. Because these households will not have matched telephone numbers, they can only be interviewed in person. If telephone numbers are provided in the mail screener, field interviewers will attempt to conduct the interviews by telephone instead of in person.

In addition, although mail screening is significantly less expensive than telephone screening, expected response rates to mail-only follow-up efforts are also much less. Thus the cost of mail screening includes not only completed screeners but also a substantial fraction of screeners that are not completed and so must be worked in a second mode.

A flowchart depicting the flow of household cases through data collection modes is provided below. In the chart, the home-based providers identified through the address-based sample of households are referred to as FFNN, as these providers are commonly labeled as Family, Friend, Neighbor and Nanny providers.

Household Surveys Data Collection Flowchart

4. Mail version of Household Screener

We have provided a revised version of the Household Screener – Mail administration. The individual items in the Household Screener have all been cognitively tested previously in the Design Phase of the NSECE and several have been previously administered in hard-copy questionnaires. The screener has not been previously fielded as a mail document; we are eager to learn the results of the field test.

Items required for eligibility (items 2, 3a, 5, 6a, and 8) are consistently worded between the mail and telephone versions, although in some cases broken out into multiple questions in the hard-copy (mail) format. Additional questions in the mail screener are intended for non-response analysis and to inform data collection approach and do not have direct analogs in the telephone screener.

A primary objective of the field test will be to understand the rates of screener completion we experience with the mail screener. The next most important criteria for evaluation will be the extent to which we confirmed eligibility status of mail screener respondents when addresses were sent to the field for interview completion, and the extent of differential mail screener response by eligibility status as determined in the mail or field phases of screening. See the proposed table format below.

Final eligibility status % mail response % field response

Household only

Home-based only

Both Household and Home-based

Neither Household nor Home-based

A similar table can be constructed replacing mail response with telephone response to help us assess the relative efficacy of each of the screener modes of data collection with different populations.

These analyses will help us understand the efficacy and cost-effectiveness of fielding the mail screener. They may also help us to refine our screener strategy for the mail study, for example, by identifying tract characteristics where the mail screener may not be effective (such as tracts with high mobility rates or high rates of linguistic isolation).

The next analysis questions pertain to the functioning of the screener as a questionnaire. Here, we will examine rates of item non-response for each of the items. We will also look for ‘out of range’ responses, for example, unusually high numbers of hours worked per week, or adults living in the household. Where the screener item also appears in a later survey instrument (for example, number of children under age 13 in the household or number of hours per week worked by screener respondent), we will compare those responses for corroboration. This review will help us revise the questionnaire as needed for the main study.

5. Household Screener incentive experiments

1. Telephone Screener Incentive Experiment

We have changed the experimental value for the Telephone Screener Incentive to $5 from $10.

6. Provider incentives

1. Home based drawn from household screener.

The expected sample size for this population is indeed quite small, but we nonetheless wish to learn something about data collection from this population in the field test. Our entire data collection approach for the Home Based Provider sample derived from the Household Screener is to mimic the approach for the Household Questionnaire wherever possible. We intend to test whether the Home-Based Provider incentive experiment cases can be pooled with the Household Questionnaire incentive experiment cases. If so, then we are able to make full use of the experimental results. If tests indicate that the Home-Based Provider respondents respond differently to the incentive experiment, that too will be informative. We anticipate making empirically-based decisions about main study incentives for the Household Questionnaire, then choose incentives for the Home-Based Providers relative to what is chosen for the Household Questionnaire.

2. Center-based provider web experiment.

We will offer $35 in the center-based provider web experiment.

3. Gatekeeper incentive for Center-based Provider Questionnaire

The criteria for receiving the gatekeeper incentive include that a single person controls access to the designated respondent; this requires that the gatekeeper not be the same person as the respondent to the Center-based Questionnaire. A gatekeeper would be a school receptionist or facility administrator who would answer phones or greet visitors but who would not have leadership responsibility over the program. The people eligible are the people who would typically have been eligible for the ECLS school coordinator incentives, while the director or principal is generally the questionnaire respondent.

Field interviewers do not have discretion over administration of the incentive. They will document to their field managers that 1) a single person controls access to the designated respondent, and that 2) they have had three prior experiences being prevented from gaining access to the director by the gatekeeper (e.g., gatekeeper refuses to transfer a call or allow the interviewer to leave a voicemail or gatekeeper cancels appointments). They will receive approval from their field managers to offer the incentive (to experimentally eligible cases), and this approval will be documented within the case management system. Evidence of access control and access prevention will be available in the CAPI records of call for prior review by field managers. Central office staff will be able to monitor approvals granted by field managers and any instances of non-compliance with experiment protocols should they arise.

Summary of Revised Incentive Experiments

Incentive Experiment	Sample Type and Timing	Incentive Amount	Sample Size
Household Screener Incentive Experiment - Telephone	HH sample (with telephone match) after 1 refusal, Hang-Up During Introduction, or 2nd unsuccessful contact; randomly divided into two groups: experimental & control	control group:$none	572
		experimental group: $5 via mail	572
Household Screener Incentive Experiment - Mail	HH Sample (without telephone match); randomly divided into two experimental groups	experimental group 1: $1 coin in initial mailing and none in follow-up mailing	609
		experimental group 2: no incentive in initial mailing and $5 prepaid in follow-up mailing	610
Household and Home-based Provider Surveys Refusal Conversion Experiment	HH sample: known eligible field cases (eligible for both the Household Survey and the Home-based Provider Survey) with 1 refusal; randomly divided into 2 experimental groups	experimental group 1: $5 prepaid	116
		experimental group 2: $5 prepaid & $10 promised	116
Home-based Provider and Center-based Provider Web Survey Incentive Experiment	Center-based providers and Home-based Providers (from admin list) who haven't completed web survey by 3rd/follow-up mailing; randomly divided into experimental & control group	control group: none	220/236
		experimental group: $35 gift card promised	221/235
Home-Based Provider(eligible through HH screener) Incentive Experiment	Home-based Providers identified through the HH sample and only eligible for the Home-based Provider survey: who haven't completed web survey by 3rd/follow-up mailing; divided into experimental & control group	control group: none	95/79
		experimental group: $35 gift card promised	94/79
Gatekeeper Incentive Experiment	Sampled Center-based and Home-based providers (from admin. Lists) divided into experimental and control groups (after 1 refusal)	control group: none	220/236
		Experimental group: $10 prepaid	221/235

8