TPC Journal V8, Issue 1 - FULL ISSUE

The Professional Counselor | Volume 8, Issue 1 79 interpreted as follows: small ( h 2 p > .01; d = .20), medium ( h 2 p > .06; d = .50), and large ( h 2 p > .14; d = .80; Cohen, 1969; Richardson, 2011). All analyses were conducted using SPSS version 24.0. Qualitative. The authors conducted focus groups and employed CQR methodology to investigate participant experiences (Hill et al., 2005). Specifically, CQR was chosen because it uses elements from phenomenology, grounded theory, and comprehensive process (Hill et al., 2005). CQR is predominantly constructivist with postmodern influence (Hill et al., 2005), which was a good fit for the project as we were interested in students’ experiences being trained in the aged-up STAC program. Furthermore, we selected CQR because it includes semi-structured interviews to promote the exploration of participants’ experiences, while also allowing for spontaneous probes that can uncover related experiences and insights, adding depth to findings (Hill et al., 2005). CQR was well suited for this study because it requires a team of researchers working together to reach consensus analyzing complex data (Hill et al., 2005). Focus groups were chosen because they allow researchers to observe participants’ interactions and shared experiences such as teasing, joking, and anecdotes that can add depth to the findings (Kitzinger, 1995). Focus groups have potential therapeutic benefits for participants, including increasing feelings of self-worth (Powell & Single, 1996) and empowerment (Race, Hotch, & Parker, 1994). Additionally, focus groups can be especially useful when power differentials exist between participants and decision makers (Morgan & Kreuger, 1993). Three team members (the first and second authors and a master’s in counseling student) employed the CQR methodology to analyze the data. After the data transcription, each member worked individually to identify domains and core ideas prior to meeting as a group. The team met three times in the next month to achieve consensus. Researchers relied on participant quotations to resolve disagreements, to cross-analyze the data, and to move into more abstract levels of analysis (Hill et al., 2005). The team labeled domains as general (typical of all but one participant or all participants), typical (more than half of participants), and variant (at least two participants; Hill et al., 2005). An external auditor analyzed the data separately, utilizing NVivo qualitative analysis software (Version 10; 2012), and reported similar findings with the exception of a minor modification to one domain, which the team incorporated into final findings. Next, the researchers conducted member checks (Lincoln & Guba, 1985) by emailing all participants with an overview of the findings. All participants who responded agreed the findings were an accurate representation of their experience. Strategies for Trustworthiness . As recommended by Hays, Wood, Dahl, and Kirk-Jenkins (2016), we used multiple strategies to strengthen the trustworthiness of the study. First, our process was reflexive with continuous awareness of expectations and biases. Prior to conducting focus groups, we discussed and wrote memos about our expectations and biases (Creswell, 2013). To triangulate data, all three analysts were involved throughout the process and in comparing findings among the team. An external auditor was included to provide oversight and increase credibility of findings. Once all researchers reached agreement about major findings, we elicited participant feedback to increase credibility and confirmability of our findings (Lincoln & Guba, 1985). Findings Knowledge and Confidence The researchers examined changes in knowledge and confidence across three time points (baseline, post-intervention, and follow-up). Results indicated a significant main effect for time: Wilks’ Lambda = .31, F (2, 20) = 6.85, p < .000, h 2 p = .31. Follow-up paired t-tests indicated a significant difference in