One of the values of Learning Together is praxis – a commitment to bringing existing research knowledge to our practice, and to evaluating what we do so that we can challenge existing knowledge and continuously improve our work. Over the last five years, with support from the British Academy, the Cambridge Humanities Research Grant Scheme, and many others, we have worked with our students to understand experiences of Learning Together, and to develop evaluation measures. As part of this process we have conducted over 250 one to one interviews with students, facilitators, lecturers and prison staff.
This summer, the Learning Together team in Cambridge were delighted to welcome a volunteer intern from the USA – Julia Arnade-Colwill – to our team (you can read her welcome blog here). Part of Julia’s work with us involved reading through some of these (anonymized) interviews and ‘coding’ them for themes that appear regularly across different people’s interviews. This is one way of understanding the commonalities and differences in people’s experiences and reflecting on our practices, good and bad. Julia recently shared some reflections with us about her experience of conducting this ‘qualitative data analysis’ as it is known in social science. Beneath this rather impersonal and detached term for a research process, belies a reality that is more accurately captured in Julia’s reflections that describe a highly emotionally engaged and thoughtful process of co-creation. With Julia’s permission, we wanted to share her reflections on our blog. We thought others might enjoy reading them as much as we have, reflecting with us on the relational nature of data collection and analysis. Julia ends her reflections by acknowledging the challenge of all research – how to do justice to the ideas and experiences of people who participate in research, while being aware that the very process of analysis is an act of subjective interpretation and objective meaning attribution. We continue to grapple with this struggle.
Thoughts on Qualitative Data Analysis and Interviews by Julia Arnade-Colwill
Prior to immersing myself in the post-course interviews of Learning Together students, I imagined the interview as solely a matter of reflection. This framework implicitly sorted the interview as separate from the learning that occurred during the course, whereby reflection signifies the delivery of anterior knowledge to the interviewer. It soon became clear to me that this conceptualization fails to capture what is really happening in the process of student reflection, and fails the understanding of learning that is at the heart of Learning Together’s philosophy and practice. The interview is not a delivery, subsequent to the learning, nor solely a means to create new knowledge through its use in research. It is part of the learning itself—a fundamental part of that learning process in which the knowledge of the course and one’s experience changes through reflection. Just as Learning Together believes in Freirean mode of education in which both student and teacher create knowledge and learn from each other, the interviewer and the interviewee are engaged in what seems better described as a conversation than an interview, which alters “known” knowledge and brings new insight into the realm of the known.
The conversational structure of the interview and its impact on shaping knowledge reminds me of Soshana Felman’s work on the intersections of education and psychoanalysis. By invoking Felman here, I do not mean to suggest that these interviews are acts of psychoanalysis, or that the interviewer is psychoanalyzing— only that Felman’s theory of learning and the type of knowledge produced by a psychoanalytic approach to learning shed some light on the learning and knowledge production I “witnessed” in a sense by reading and coding interviews. Felman attests that “knowledge in other words, is not a substance but a structural dynamic: it is not contained by any individual but comes about out of the mutual apprenticeship between two partially unconscious speeches which both say more than they know. Dialogue is the radical condition of learning and knowledge” (Felman, “Psychoanalysis and Education: Teaching Terminable and Interminable). This theory of shared and discourse-driven knowledge clearly harmonizes with Freire’s theories. What Felman adds to the discussion is the use of the unconscious and its relation to the (un)known. Rather than viewing the unknown, or ignorance, as an absence of knowledge, Felman draws on Lacan to envision the unknown as “knowledge which does not know itself.” The unknown is knowledge, but needs to be made known by drawing it from the unconscious to the conscious. This can be achieved through the dialogue— through something like an interview. The most exciting moments in the interviews I read were the moments in which the student unlocked a piece of knowledge that they had previously struggled to communicate and recognize. Oftentimes, these bits of knowledge were the crucial bits—those which provided an alternative framework to understanding their experience as a whole. Sometimes, it was a feeling lurking in the recesses of the obscure that could be made known through new vocabulary co-produced by the interviewer and student. Other times, it was a moment of resistance in the student’s response to a question, a type of unsureness or contradiction that the interviewer could pick up on and disentangle through dialogue. Indeed, Felman’s mantra of learning is the question “Where does it resist?” By locating the point in the interview in which knowledge resists, the interviewer and student could then do the work of making that knowledge known.
These disruptive forms of knowledge offered a rich landscape for coding. The difficulties I faced came from finding the place for myself within this framework of unknown knowledge. There is the first layer of challenge that obviously arises from coding interviews: in moments of “knowledge resistance” that I perceived, I could not directly insert myself into that interview to make it known through dialogue. If I did not understand something and it hadn’t been made clear to me through the interview, I could only do the best with the codes provided to interpret what that knowledge may have been. Of course, the codes themselves representing actors in the next layer of dialogue, my use of them may have altered the representation of what the person was saying at that moment, and what knowledge was invested in it. Following from that idea of the codes as participants in a dialogue, the perhaps trickier challenge I faced was doing justice by fascinating and complex moments in the interview, pieces of knowledge that I struggled to fit within the codes without feeling that I was missing a key element that gave them their complexity. Pertinently, these moments were often those in which knowledge became known in a new way, which I often coded under “productive discomfort,” to describe the phenomenon, but oftentimes struggled to code for the content. Certain forms of ambivalence, whereby the speaker communicated several opposing notions, each of which was true, were difficult to code in this regard. I wanted to code the moments I found interesting and wonderful in order to make them “known” to the research. But did this coding then change the ways it could be known? Without falling back on the regressive notion that knowledge is something individually owned and preserved, I did wonder how to best communicate the speaker’s ideas in a way that could meaningfully shape them within the purposes of coding while doing justice to the speaker.