134 Chapter 5 Data analysis In Phases 1 and 2, the (first and second) interviews were deductively analysed to explore awareness of the educational environment, behaviours/ competencies, identity, and mission, using the codebook as developed by Ottenhoff-de Jonge et al.7 (See Addendum 5.2). Two researchers independently coded the interviews. In addition to the researcher who had conducted both interviews (MO), two different research assistants coded the first and second interviews independently. The assistant who coded the second interviews (IvdH) was blinded to the interviewees’ initial categorisation. The codings were discussed jointly to reach consensus. We determined an educator phenotype for each participant holistically, i.e., based on the whole transcript, including all the coded text fragments. A third team member (RvdR) independently analysed half of the first interviews and a selection of the second interviews to further ensure data credibility.18 The first author performed a member check by asking all participants if they agreed with our conclusions on the assigned educator phenotypes. The member check was incorporated into the third interview with those who had matured into a more inclusive phenotype. In Phase 3, we analysed the transcripts of the third interviews. MO and IvdH inductively analysed which factors were perceived by the educators to be instrumental in their maturation during the last decade. This analysis followed an iterative process to promote in-depth understanding of the interviews and the coding. Coding took place independently by MO and IvdH, and discrepancies were discussed until consensus was reached. After further analysis we consolidated the codes into a limited number of categories by discussing the coding of the fragments and categorisation with RvdR and AK. To strengthen the confirmability18 of our findings, the categorisation and interpretation were discussed with the full research team. The team members come from diverse backgrounds, including from different cultures (USA versus the Netherlands), disciplines (medical versus non-medical), and professional experiences (student versus long-term professional experience). This approach ensured interpersonal reflexivity to identify and address any personal beliefs or biases that may have influenced the research process, and intersubjectivity to reach agreement on the interpretation of the findings through dialogue and the sharing of meanings. Regarding contextual reflexivity, MO, AK and IvdH are familiar with the educational context of the LUMC, as medical educators and as a medical student respectively, while NG is familiar with the SUSM context as the Senior Associate Dean for Medical Education. In addition, the first author (MO) participated at SUSM as a medical teacher for a short period of time, in order to gain a better understanding of the practice and
RkJQdWJsaXNoZXIy MTk4NDMw