TPC_Journal_10.4_Full_Issue

458 The Professional Counselor | Volume 10, Issue 4 digital recorder instead of the Zoom platform, as the interview occurred in person during a professional conference. All demographic information and recordings were assigned an alphabetical identifier (e.g., A, B, C). The last author was the sole individual who knew the identity of participants attributed to alphabetical identifiers. Participant identity was thus blinded to subsequent transcribers and coders. Transcription All interviews were transcribed verbatim by graduate students at the last author’s university, who had no familiarity with participants. Transcribers received transcription training prior to the study and received further training and direction by the last author prior to and during the transcription process. Once each transcript had been completed in full, the last author reviewed transcripts to ensure accuracy and sent the transcripts to the interviewees to conduct a member check. After member checks had been conducted, sections of transcripts were cut and pasted into separate documents for each of four research teams to code and analyze. The research teams were organized by research question (i.e., components of high quality; recruitment, support, and retention of underrepresented students; working with administrators; successful dissertation advising). Transcribed interviews for each research team were uploaded to separate secure folders in a secure encrypted online data management software system. Data Analysis The last author met with members of all four research teams collectively to ensure consistency in the coding approach. The last author developed several guidance documents for the research teams to use and created instructions for coding the data, which included guidance such as each research team meeting to bracket biases and identify any a priori codes prior to initial coding of the data, following Merriam and Tisdell’s (2016) guidelines. Research teams were instructed to identify emergent in vivo codes using verbatim line-by-line open coding when possible to avoid interpreting data too early during the coding process (Merriam & Tisdell, 2016). The focus of coding was to identify themes within and between participants. The four research teams were instructed to meet weekly over a period of several months to code and analyze data specific to their research question. Research teams coded each of the first three transcripts together as a team during weekly live coding sessions using the Zoom online platform, prior to individual team members coding the remaining transcripts separately. Codes were noted on the transcripts themselves, and then the lead team member compiled the codes into the code book. From there, the categories were developed and reviewed by all team members. Discrepancies in coding were resolved using coding consensus, with the research team documenting how they resolved any discrepancies in coding. Weekly meetings were required even when individual team members were coding separately to facilitate sharing their coding experience, clarifying questions about codes, establishing consensus on any parts of the transcript with complicated coding, and following the coding approach with consistency across coders. The last author created a coding database template that each research team was required to use, to ensure consistency in how coding was documented and categorized. These approaches were designed to improve consistency in coding within and between the four research teams. Each of the four research teams only coded and analyzed data pertinent to their assigned research question. A coding team chair was identified for each of the four research teams to ensure that the coding and analysis approach was followed consistently and to organize the work of the team. Each research team organized codes into categories and eventually collapsed codes into themes using axial coding after all 15 transcripts had been coded. Themes also were analyzed by demographic and program characteristics of interviewees to assess the potential influence of background characteristics on responses. Each research team recorded memos during collective team meetings and during

RkJQdWJsaXNoZXIy NDU5MTM1