research analysis

The Qualitative Analysis Trap (or, Coding Until Blue in the Face)

There is a trap that is easy to fall into when conducting a thematic-style analysis of qualitative data. The trap revolves around coding and, specifically, the idea that after a general familiarization with the in-depth interview or focus group discussion content the researcher pores over the data scrupulously looking for anything deemed worthy of a code. If you think this process is daunting for the seasoned analyst who has categorized and themed many qualitative data sets, consider the newly initiated graduate student who is learning the process for the first time.

Recent dialog on social media suggests that graduate students, in particular, are susceptible to falling into the qualitative analysis trap, i.e., the belief that a well done analysis hinges on developing lots of codes and coding, coding, coding until…well, until the analyst is blue in the face. This is evident by overheard comments such as “I thought I finished coding but every day I am finding new content to code” and “My head is buzzing with all the possible directions for themes.”

Coding of course misses the point. The point of qualitative analysis is not to deconstruct the interview or discussion data into bits and pieces, i.e., codes, but rather to define the research question from participants’ perspectives Read Full Text

Giving Voice: Reflexivity in Qualitative Research

Homegoing, the debut novel by Yaa Gyasi, is a moving tale of slavery and its translation across generations. At one poinMinority voicet, we read about a descendant in Ghana who teaches history and on the first day of class stumbles on a lesson concerning “the problem of history.” The problem he refers to is that history is constructed from stories that are handed down over time yet “We cannot know which story is correct because we were not there.” He goes on to say to his students

We believe the one who has the power. He is the one who gets to write the story. So when you study history, you must always ask yourself, Whose story am I missing? Whose voice was suppressed so that this voice could come forth? Once you have figured that out, you must find that story too. From there, you begin to get a clearer, yet still imperfect, picture. (pp. 226-227)

The month of February seems like an appropriate time to reflect on power and what we as researchers are missing in our studies of vulnerable and marginalized segments of the population. After all, with the exception of participatory research, we are typically the ones who control the design and implementation of data collection along with the analysis, interpretation, and reporting of the findings.

Reflection on our role in the research process should be common practice. But our reflection takes on new meaning when our participants are those with the weakest voice. As we sit down with our reflexive journal and consider our prejudices and subjectivities (by asking ourselves the kinds of questions outlined in this RDR article), researchers might do well to pay particular attention to their assumptions and beliefs – What assumptions did I make about the participant(s)? and How did my personal values, beliefs, life story, and/or social/economic status affect or shape: the questions I asked, the interjections I made, my listening skills, and/or my behavior?

Few, Stephens, and Rouse-Arnett (2003) address this in their discussion on interviewing Black women on sensitive topics. As Black women themselves, they felt no less obligated to reflect on their status.

 As Black feminist qualitative researchers, we are particularly attuned to how we become the research instruments and the primary sieves of re/presentation in our exploration of Black womanhood. (p. 213)

By way of this reflection, the authors make recommendations toward the interviewing of Black women on sensitive topics, including such concepts as “contextualizing self in the research process.” The authors also come to the realization that “the diversity of Black experience has been misrepresented [by] traditional family studies orientations,” asserting that “the persistent matrix of intersectionality that Black women endure, succumb, and overcome” cannot be fully addressed if “researchers debate and deconstruct out of existence the ‘critical essences’ (i.e., race, class, and gender) that matter to Black women’s existence and survival in this world” (p. 213).

So, take another look at your reflexive journal. Take another look at your research with the vulnerable and marginalized. And, if not already there, consider adding these queries – so well put by Gyasi – to your journal: Whose story am I missing? Whose voice has been suppressed? Whose story do I need to seek out to help me gain a clearer, more complete picture of the people and the phenomenon I hope to illuminate through my research? How, indeed, have I used my power as a researcher to give center stage to the “critical essences” of society’s minority voices?

Few, A. L., Stephens, D. P., & Rouse-Arnett, M. (2003). Sister-to-sister talk: Transcending boundaries and challenges in qualitative research with Black women. Family Relations, 52(3), 205–215.

Image captured from: https://jennymackness.wordpress.com/category/connectedcourses/

The Important Role of “Buckets” in Qualitative Data Analysis

An earlier article in Research Design Review“Finding Connections & Making Sense of Qualitative Data” – discusses the idea that a quality approach to a qualitative research design incorporates a carefully considered plan for analyzing, and making sense of, the data in order to produce outcomes that are ultimately useful to the users of the research. Specifically, this article touches on the six recommended steps in the analysis process.* These steps might be thought of as a variation of the classic Braun & Clarke (2006) thematic analysis scheme in that the researcher begins by selecting a unit of analysis (and thus becoming familiar with the data) which is then followed by a coding process.

Unique to the six-step process outlined in the earlier RDR article is the step that comes after coding. Rather than immediately digging into the codes searching for themes, it is recommended that the researcher look through the codes to identify categories. These categories basically represent buckets of codes that are deemed to share a certain underlying construct or meaning. In the end, the researcher is left with any number of buckets filled with a few or many codes from which the researcher can identify patterns or themes in the data overall. Importantly, any of the codes within a category or bucket can (and probably will) be used to define more than one theme.

As an example, consider an in-depth interview study with financial managers of a large non-profit organization concerning their key considerations when selecting financial service providers. After the completion of 35 interviews, the researcher absorbs the content, selects the unit of analysis (the entire interview), and develops 75-100 descriptive codes. In the next phase of the process the researcher combs through the codes looking for participants’ thoughts/comments that convey similar broad meaning related to the research question(s). In doing so, Read Full Text