Appendix B

Caregiver Interview Sample and Methodology

INTERVIEW STUDY METHODS

Study methods were designed to surface the experiences of family caregivers working in the academic sciences, engineering, and medicine whose perspectives have been underrepresented in prior research. A qualitative approach was developed to allow space for the complexity (and often, emotion) surrounding subjective experiences of managing caregiving and career, the multilayered contexts in which those experiences occur, and their consequences for caregivers and the field at large (Sofaer, 1999). The study research protocol and all outreach and data collection materials were reviewed by the National Academies of Sciences, Engineering, and Medicine Institutional Review Board.

SAMPLING

To ensure substantial representation of women of color and caregivers of intersecting marginalized identities, the Committee on Women in Science, Engineering, and Medicine (CWSEM) staff, study committee members, and the RTI International study team partnered to develop a targeted outreach campaign for the study. RTI and CWSEM staff focused on identifying and connecting with member listservs and similar communication tools that centered scholars of color; first-generation college graduates; immigrant scholars; those who identified as lesbian, gay, bisexual, transgender, or queer (LGBTQ+); and those living with disabilities.

Respondents were reached via a series of general and targeted outreach emails and via forwarding to individual contacts. Outreach messages described study eligibility criteria, interviewing approach, and compensation and directed interested individuals to a web-based screening form for the study.

The web-based screening form, developed and hosted by RTI, consisted of a set of closed-ended questions designed to establish study eligibility and elicit information on other relevant life experiences for case selection purposes. To be eligible, respondents had to be working or studying in academic science, engineering, and medical fields (including doctors and nurses in training or academic medicine) and must have had regular, unpaid caregiving responsibilities during or beyond the early COVID-19 pandemic.

RTI reviewed incoming responses daily to identify eligible individuals and select a purposeful sample from among them. Based on the committee’s assessment of perspectives that were lacking in prior research, priority was accorded to women and other caregivers of color as well as those who are LGBTQ+, first generation, immigrant, and/or living with disabilities. RTI also sampled for diversity in

• the career path and stage, with the aim of including individuals at all career stages, on and off the tenure track, and with special priority given to women of color in senior faculty roles;
• the nature of the unpaid caregiving responsibilities, including for whom care was given and what kinds of activities the caregiver performed; and
• the science, technology, engineering, mathematics, and medicine (STEMM) discipline, to ensure participation from subfields identified by the study committee as more heavily male dominated.¹

Outreach, screening, sampling, and recruitment proceeded iteratively based on the evolving characteristics of the available sample. RTI extended the recruitment period by approximately 1 month in order to support ongoing efforts to engage women of color senior faculty.

___________________

¹ The “heavily male-dominated fields” identified by the study committee were physics; computer science; astronomy or astrophysics; and civil, aerospace, electrical, and mechanical engineering.

DATA COLLECTION

Individuals selected for interviews were contacted using each of their preferred modes of contact (email, phone, or both) and invited to schedule a Zoom interview with an experienced qualitative interviewer. Individuals who completed the screening form but were not selected to participate in an interview were thanked immediately and were notified at the end of the recruitment period, using their preferred mode of contact, that they had not been selected.

An informed consent process was administered at the beginning of each scheduled Zoom interview and only individuals who consented to participate in the interview and to be recorded were interviewed. Interviews lasted approximately 1 hour and followed a semistructured guide that covered the following topics:

• Experiences of managing career and caregiving responsibilities simultaneously
• The macro-, meso- and micro-level contexts in which caregivers managed those demands
• Ideas for reimagining success and productivity
• Experiences of joy and satisfaction in career and caregiving

Following the interview, each respondent was sent a thank-you email that included a $75 Amazon gift code, information about the expected release of study findings, and contact information for the CWSEM representative.

ANALYSIS

Interview recordings were professionally transcribed for analysis. A deductive codebook was developed based on the study research questions and early study committee guidance. Inductive codes were developed jointly by the research team to reflect themes that emerged during the interviews. New inductive codes were added by all members of the research team during the analytic process.

In addition, a set of family codes was developed to reflect differences in life experience that the study committee expected might be meaningful for purposes of managing caregiving and career, including the nature of the respondent’s caregiving role(s), the forms of care they provided, career path and stage, ethnic and racial identity, and other experiences of structural disadvantage (identifying as LGBTQ+, an immigrant, a first-generation college graduate, and/or living with a disability).

Recognizing the multitude of distinct life experiences of potential relevance and the complex, intersecting nature of those experiences, the team did not attempt to sort or otherwise “bin” the transcripts according to the family codes (as in a structured, comparative analysis). Rather, analysts referred to screening data to apply family codes to each set of interview notes at the beginning of the transcript review. This information informed the coding and theming process, and analysts made note of any inductive observations related to the family codes.

Themes were described in brief analytic memos shared across the analysis team via a shared master analytic file along with the exemplary quotations associated with each theme.

The interviews were coded with family codes detailing key characteristics of each interviewee as well as deductive codes that emerged from the data. Tables B-1 and B-2 present the codes used in analysis.

TABLE B-1 Family Codes

TABLE B-2 Deductive Codes

SAMPLE CHARACTERISTICS

Reflecting the study aims and sampling objectives, a majority of interview participants were caregivers of color (N = 26). Roughly one-quarter of the sample identified as Black (N = 9), Latinx (N = 11), or Asian (N = 10); half identified as White (N = 21); and smaller numbers identified as American Indian or Alaska Native (N = 3), and/or Native Hawaiian or Other Pacific Islander (N = 2).

Sixty percent of sample members were from immigrant families. A majority of sample members were the first person in their families to graduate from a 4-year college (N = 21).

Women of color senior faculty (N = 8), caregivers working in heavily male-dominated fields (N = 4), LGBTQ+ caregivers (N = 4), and caregivers with disabilities (N = 7) were represented in smaller but substantial numbers. They were drawn from across all career stages, from students to senior faculty and academic leadership, with heaviest representation from graduate students, medical residents, and other early-career scholars (see Figure B-1).

A majority of qualitative interview participants were engaged in caregiving for young (35 percent) or school-age (38 percent) children. Substantial proportions, however, were caring for other loved ones: One-third were caring for a parent or other elder family member (33 percent), and smaller proportions caring for chosen family or friends (15 percent), spouses or partners (10 percent), members of extended family (8 percent), or an adult child with disabilities or other intensive caregiving needs (8 percent).

Participants fulfilled a wide variety of caregiving roles. Most provided support with activities of daily living as well as a range of other high- and low-autonomy responsibilities such as supporting physical needs (75 percent), coordinating medical care (73 percent), accompanying or transporting to activities/appointments (65 percent), managing day-to-day scheduling (53 percent), supporting participation in online school (43 percent), managing finances (25 percent), and managing involved with the legal system (3 percent). Study participants reported having made a range of career changes due to the imperatives of their caregiving responsibilities, including opting for more flexible work (60 percent), reducing working hours (40 percent), and leaving their field (15 percent). A small proportion (13 percent) of sample members had not made any major career changes due to their caregiving responsibilities.

REFERENCE

Sofaer, S. 1999. Qualitative methods: What are they and why use them? Health Services Research, 34(5 Pt. 2), 1101-1118.