Data analysis

Thematic analysis was used to describe Mainline and Black Protestant religious leaders’ attitudes towards abortion and how they provide pastoral care regarding abortion. Interviews were transcribed verbatim by a professional transcription company and de-identified by the research team. Data were managed and organized using Dedoose version 8.0.35, a software package for qualitative data analysis [30]. Data collection, coding, and analysis occurred simultaneously to assess meaning saturation, or a “richly textured understanding” of abortion attitudes and pastoral care practices [31]. Saturation was reached at 20 interviews when a diversity of views and denominations was achieved and no new themes were observed.

Data were read in detail and memoed in order to develop a codebook (See S2 Appendix). Codes were developed through inductive (emerging from the data) and deductive (pre-determined topics from the interview guide) strategies. An inter-coder agreement exercise was conducted prior to coding all data to ensure consistency in the coding process. Weekly team meetings were held to refine code definitions, resolve discrepancies in coding, and discuss reflexivity in data interpretation during the coding process. For example, researchers discussed how our underlying epistemologies, public health training, and differing views on scripture and doctrine, might influence interpretation of data during analysis.

Researchers explored data by codes (e.g., abortion, attitudes & beliefs, and pastoral care), conducted structured comparisons of codes by sub-groups of participants (e.g., by sociopolitical attitudes, denomination, and gender). Patterns in the data were examined and sub-themes within codes were identified. Illustrative quotes were then selected for each sub-theme.