Researchers ’ Positionalities

For any qualitative study, it is important to discuss how the social position of the researcher influences the research process, including data collection, analysis, and interpretations (Creswell, 2013; Saldana, 2018). This research study was conducted by three Latinas, two of whom are first-generation college students. One researcher is a professor in the school of education at a large public university, and the other two researchers are graduate students who are pursuing their doctorates in education. Our scholarly interests broadly focus on equity issues for traditionally underserved students in higher education.

As researchers, we believe that our lived experiences as Latinas who v'ere primarily the first in their family to go to college, were assets that we could leverage during the research process. As such, we drew' upon our collective cultural intuition (Delgado Bernal, 1998). That is, we used our “personal experience, collective experience, professional experience, |andj communal memory” (Delgado Bernal, 2016, p. 1) to gain a unique insight and understanding into how participants described and experienced their identities, and how these identities were engaged to persist in STEM (Leedy & Ormrod, 2005).

Data Sources and Participants

Study participants included first-time, full-time, first-year, and transfer students matriculating at one of four universities in the Northeastern United States in the fall of 2016. The four institutions represent a variety of institutional contexts, including large public land-grant universities, a large private urban university, and a STEM-focused private urban university. While study participants attended a variety of institutions, they all participated in

Table 11.1 Latina Student Profiles (m=8).

the Louis Stokes Alliance for Minority Participation (LSAMP) program, a federally funded program that aims to increase the number of underrepresented minority students matriculating into, and successfully completing, high-quality undergraduate degrees in STEM.

Data collection for this study occurred over several academic school years. During the 2016—2017 academic year, members of the research team traveled to each campus and invited students to participate in the study. Study participants completed an in-person survey that gathered their demographic information, including race/ethnicity, gender, and first-generation college status (see Table 11.1). In the second year of the study, students received an email asking them to participate in a follow-up one-on-one interview. A total of eight self-identified Latina students agreed to participate. Among them, half identified with two ethnic or racial categories, representative of the complex and multidimensional aspects of Latina identity.

The interviews were conducted using a semi-structured protocol that included questions about students’ involvement on campus and who they went to when they needed advice about their future in STEM. Before conducting the interviews, members of the research team tested and received feedback on the protocol to ensure that the questioning route was conversational and that the questions were relevant, clear, and concise (Krueger & Casey, 2009). All interviews were conducted in person and took place in a mutually agreed upon space on the student’s college campus. Each interview was audio recorded with the consent of participants and ranged between approximately 40 and 80 minutes in length.

Data Analysis

To analyze the qualitative data, the research team transcribed each audio file and reviewed each transcript for accuracy. Then, we de-identified the transcripts by replacing students’ names and other identifying information with pseudonyms. Next, we entered the interview transcript data and open- ended survey responses into Dedoose, an online software used for analyzing qualitative data. To begin, two members of the research team used open coding techniques (Saldana, 2015) to develop an initial codebook—that is, a shared understanding of data concepts and themes that would inform future rounds of coding (Denzin & Lincoln, 2008).

Once we defined an initial set of codes, two members of the research team independently and carefully read and analyzed the interview transcripts. Then, we came together to identify similarities and differences in our coding of each interview transcript until discrepancies were resolved. When coding discrepancies were encountered, a research team member served as a peer-debriefer (Saldana, 2015). Reconciling discrepancies among researchers strengthened the codes and increased the reliability of the analysis (Miles, Huberman, & Saldana, 2014). As a multi-member team, peer debriefing and memoing served as critical components in the data analysis process (Saldana, 2015). Finally, members of the research team engaged in a second round of coding to develop themes.