Section_B_Statistical_Methods_OWH_10_25_2019_clean

Supporting Statement B for the State-Level Paid Family Leave Project

1.Respondent Universe and Sampling Methods

The State-Level Paid Family Leave Policy Project funded by the U.S. Department of Health and Human Services (DHHS) Office on Women’s Health (OWH) aims to explore the relationship between state-level Paid Family Leave (PFL) programs and new mothers’ health, health behaviors, and ability to fulfill their roles in the workplace, family and community. Qualitative methods will be used to collect data through 16 focus groups in the four states with fully functioning programs (California, New Jersey, Rhode Island, and New York). In order to explore potential differences in experiences, approximately half of focus group participants will have used state-level PFL. The other half will not have used state-level PFL to serve as a comparison.

Participants must be a convenience sample of mothers with a child under the age of one and be eligible for their state’s respective PFL program. To participate as a state-level PFL user, mothers must have used the entire state-level PFL benefit. To participate as a state-level PFL non-user, mothers must have a baby older than the “state-level PFL threshold” and not have taken any state-level PFL. We define the threshold as the time after which mothers are typically out of the temporary disability insurance (TDI) and state-level PFL window (12 weeks in California, 14 weeks in New Jersey, 12 weeks in New York, and 10 weeks in Rhode Island).

The qualitative approach (i.e., focus groups), use convenience samples, meaning that the results will primarily be used to generate hypotheses rather than answer questions about whether programs have an effect on health and health behaviors.

To recruit 96 women, 48 state-level PFL users and 48 non-users, the project team will adopt two strategies: Facebook ads and partnerships with organizations that provide services to new mothers. Facebook ads will populate the feeds of mothers in the areas where the focus groups will take place. Through the ads, interested individuals will complete a form with their first name and phone number or email address. A member of the project team will then contact the individual to conduct an eligibility screen. The ad will also include the email address and phone number of a member of the project team in case the interested individual would like to contact the project team directly. The project team will also partner with organizations that serve new mothers, including home visitation programs, Family Resource Centers, and lactation/moms groups. These organizations will reach out to their client populations to inform them of the focus group. Some organizations will host a focus group at a location where they typically provide services to women, thus reducing barriers to participation.

The project team aims to achieve diversity in focus group participation by holding the groups in urban and rural areas throughout a given state and partnering with organizations that target populations with diverse socio-economic and racial/ethnic backgrounds.

2.Procedures for the Collection of Information

This project does not entail the use of a sample. All interested individuals that screen eligible will be considered for participation in a focus group. If too many eligible individuals express interest and ability to participate in a given focus group, the researchers will select individuals who together represent diverse experiences and backgrounds (e.g., based on income, race/ethnicity, work history, family make up). Individuals who are not immediately selected to participate may also be considered for future focus groups.

Two researchers experienced in qualitative data collection and analysis will conduct all focus groups with two additional support staff. All focus group transcripts will be reviewed by project team leadership, including the OWH project manager, to ensure fidelity to the protocol, appropriate tone, and familiarity with all data collected.

The project team will then code the transcripts based on the main research questions. Codes will categorize women’s use of state-level PFL and other leave benefits and the potential short-term outcomes of state-level PFL that may influence women’s health, such as sleep and breastfeeding. Codes will also address women’s mental and physical health status and ability to fulfill their roles in the family, workplace, and community postpartum. Finally, codes will capture external factors also related to health, including pre-existing medical conditions, work rigor, and socioeconomic status.

The four members of the project team who will conduct the focus groups will also code transcripts to ensure familiarity with the research themes. All researchers will code two of the same transcripts as a training exercise, identifying inconsistencies across coders and developing guidelines for how to treat discussion topics that were coded differently.

Following the training, each transcript will be coded by two separate coders; differences will be reconciled through discussion after both coders complete their review. Coding pairs will vary among the four researchers to ensure reconciliations are made consistently across transcripts. The coders will also meet regularly to discuss coding uncertainties and if necessary, refine guidelines and recode information. Finally, after all transcripts are coded, the senior researcher will recheck the earliest transcripts to ensure that coding practices did not gradually change over time.

Data will be analyzed by use and non-use of state-level PFL to identify trends among women in each category. Analyses will take into consideration the nuance of leave benefits and external factors to isolate the relationship between state-level PFL and women’s health. Coded data will be linked to participants’ pseudonyms to merge demographic questionnaires and transcript data. As described in Statement A, focus group participants will use pseudonyms during the focus groups and to label their demographic questionnaires. Coding and analysis will be conducted during the several months following the last focus group.

The participating women are not meant to be statistically representative of postpartum women who are eligible for state-level PFL overall. In addition, the project team does not plan to generate quantitative statistics based on the information collected through focus groups. Therefore, the project team will not use advanced statistical analysis to assess findings.

3.Methods to Maximize Response Rates and Deal with Nonresponse

If we find recruitment strategies are not sufficient to recruit 96 women, 48 state-level PFL users and 48 non-users, we will disseminate information about the focus groups through child care centers and other local organizations that serve families, post fliers in breakrooms/communal areas of large grocery, retail stores, and other companies/organizations, and have companies share information about focus groups with their employees.

In addition, we may find that women are more likely to participate in a phone interview than a focus group, given that focus groups require travel and a longer time commitment. Therefore, if faced with focus group recruitment challenges, the project team will supplement focus groups with interviews or allow some women to call into a focus group. Interviews can target potentially underrepresented groups related to income, race/ethnicity, and use/non-use of state-level PFL.

Finally, if recruitment is low for non-users of state-level PFL, the project team may decide to expand eligibility criteria to include all women who worked in the year prior to giving birth, instead of just women eligible for state-level PFL.

4.Tests of Procedures or Methods to be Undertaken

The project team conducted pilot focus groups to test recruitment strategies, the screening process, the demographic questionnaire, and the focus group protocol. Initially, two small separate focus groups of state-level PFL users (four participants) and non-users (two participants) were conducted to test a user protocol and a non-user protocol. Based on the findings, the project team decided to combine the groups to create interaction between users and non-users and to allow women to respond to these different experiences. The combined focus group had four users and two non-users, six in total. Based on this initial pilot, the project team also shortened the focus group protocol to ensure all questions could be addressed in one hour and 15 minutes, revised two questions on the demographic questionnaire that several women found confusing, and revised the screener to capture more details about women’s employment and eligibility for state-level PFL.

The project team then conducted a third pilot focus group with both users and non-users together to test the revised protocol, demographic questionnaire, and screener. Minimal modifications were made to the instruments after this test.

5.Individuals Consulted on Statistical Aspects and Individuals Collecting and/or Analyzing Data

Data Collection Design

The data collection process was designed by staff from OWH, the Assistant Secretary for Planning and Evaluation (ASPE), and Mission Analytics Group, Inc., a contractor.

Brittany Perrotte	Office on Women’s Health Department of Health and Human Services (202) 401-1170, [email protected]
Olumayowa Azeez	Office on Women’s Health Department of Health and Human Services (202) 205-1952, [email protected]
Pamela Winston	Assistant Secretary for Planning and Evaluation Department of Health and Human Services (202) 969-3827, [email protected]
Ellie Coombs	Mission Analytics Group (925) 984-8151, [email protected]

Data Collection and Analysis

The following staff from Mission Analytics will collect and analyze data:

Ellie Coombs	(925) 984-8151, [email protected]
Anna Allison	(415) 432-3894, [email protected]
Trixy Manansala	(415) 852-7138, [email protected]
Amy Lim	(415) 814-1557 (ext: 132), [email protected]