‘Membership matters’: applying Membership Categorisation Analysis (MCA) to qualitative data using Computer‐Assisted Qualitative Data Analysis (CAQDAS) Software

Abstract

This paper introduces and outlines a methodology that may be unfamiliar to some qualitative researchers: Membership Categorisation Analysis (MCA). The first section of the paper explains the basic principles of MCA and why it is a valid method for exploring the power of categorisations in texts and talk. Additionally, it explains why MCA differs from other forms of qualitative data analysis. The second section begins with a discussion of why researchers might or might not use Computer-Assisted Qualitative Data Analysis (CAQDAS). Subsequently, a detailed description of how MCA was applied to qualitative data using the CAQDAS software package NVivo is outlined. To provide examples, this draws on a project that used MCA to analyse the interview accounts of twenty-five young people who had taken a Gap Year between leaving school and beginning university. The paper concludes that qualitative researchers should consider using MCA and that CAQDAS is a useful tool to aid its application.2

Using Software for Qualitative Data Analysis: Research Outside Paradigmatic Boundaries

“…The use of auto-coding is described as debateable at best by Lewins and Silver (2007), who argue that it risks limiting the analytical process and lowers the status of any text excluded by the process, thereby echoing earlier concerns that an excessive focus on coding at the expense of other methodological tools risks decontextualizing data (Seidel and Kelle, 1995;Kelle, 1997). Such criticisms rest on the notion that the use of CAQDAS might lead researchers to do too much coding because it is so straightforward to accomplish, leading to "data-fetishism" (Garcia-Horta and Guerra-Ramos, 2009) or a "coding trap" (King, 2010) in which the researcher is surrounded by an excess of codes which distort the rest of the research process. However, it is the researcher and not the computer who defines auto-coding parameters (Odena, 2013).…”

“…Whilst it may still be a matter of debate as to whether the use of CAQDAS is now "routine" (Seale and Rivas, 2012: 432) or "contentious" (King, 2010: 6), or 'critical' or 'instrumental' (Mangabeira, Lee and Fielding, 2004), it cannot be denied that CAQDAS has changed how qualitative research might be done. However, these changes are rather different to those envisaged by early commentators.…”

“…It was decided to proceed with Atlas-Ti primarily because it was felt by the research team that this CAQDAS offered superior and more straightforward tools for the sharing of research work (data, codes, memos and so forth) across the team. The future possibilities of both native mac (at the moment the research team are using Atlas-Ti with windows emulators on apple mac laptops) and ipad versions of the software as also attractive to the team, reflecting once again the importance of personal preferences when using CAQDAS (King, 2010).…”

“…Is it indeed the case that it has now become so (relatively) common for a researcher or a team of researchers to use software such as Atlas-Ti or Nvivo to manage their projects that it does not need to be mentioned in the methods section of a research article (Seale and Rivas, 2012)? Or is it in fact the case that the use of such software continues to be contentious, somehow causing losses in some aspects of the research process or otherwise generating problems of theory and/or method for the researcher (King, 2010)? Does the lack of reference to software in methods sections denote a greater familiarity with and consensus regarding the use of such software, or does it simply reflect the fact that the writing up of much qualitative research is instead characterised by scant regard to method and theory (Tight, 2004;Trowler, 2012;Tummons, 2012), whether or not qualitative data analysis software has been used?…”

Recognising Adulthood? Young Adults’ Accomplishment of their Age Identities

“…They were analysed using a three stage MCA administered via NVivo qualitative data analysis software (see King, 2010). This involved: identifying key membership categories, devices and predicates (CDPs) within the sample; mapping the sequence of these CDPs within each account and across the sample; and identifying how certain CDPs were treated as anomalies or disruptions by the co-participants in the talk that required some form of explanation or 'repair' work.…”