ARTICLE IN PRESS
Journal of Purchasing & Supply Management 13 (2007) 216–218 www.elsevier.com/locate/pursup
Notes and Debates
Presenting qualitative research: Convincing through illustrating the analysis process Pa¨r A˚hlstro¨m Stockholm School of Economics, Centre for Innovation and Operations Management, P.O. Box 6501, SE-113 83 Stockholm, Sweden Received 10 July 2007; received in revised form 12 September 2007; accepted 12 September 2007
Keywords: Case research; Qualitative research; Research presentation
DuBois and Araujo, in their well-argued and interesting paper, raise a number of very valid points regarding the usage of qualitative methods and particularly case research. My commentary focuses on the point raised regarding the challenge of presenting case research, especially in the journal article format. My purpose is to extend the advice given by DuBois and Araujo in this respect. The argument made here is that the challenge of presenting case-based research, is to illustrate to the readers the learning process that led the case researchers to their conclusions. Before developing this argument, however, I need to comment on my experience base regarding case research. Over the last 15 years, I have conducted everything from single-case longitudinal studies, spanning several years, to multiple case studies, involving dozens of different companies. I have also been frequently asked to review papers based on case research, for various highly ranked journals. Thus, my methodological bias, in so far it is prudent to speak of such, is and has been towards case studies. Well aware of the fact that we should justify methodological choices purely from scientific reasons, I cannot help but wonder if it is not sometimes so that we are biased towards certain methodologies, for reasons such as competence, personal interests or institutional factors. My fondness of case research apart, I am often concerned when I notice the inadequate way in which case research is at times designed, carried out and subsequently presented.
Tel.: +46 87369573.
E-mail address:
[email protected] 1478-4092/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.pursup.2007.09.008
1. The challenge of convincing the sceptical reviewer steeped in a quantitative tradition Regardless of my bias towards case research, I have lately come to appreciate quantitative research and am currently going through rather painstaking and slow attempts at learning the craft of quantitative research. What particularly appeals to me in quantitative research are the standards which exist for presenting the research. There are generally accepted rules for how to present quantitative research, for instance survey-based research. Just compile the necessary tables with the relevant statistical data and you are all right. Obviously I am simplifying things here: it is not that easy to actually arrive at the tables. Still, the actual presentation of quantitative research is, by comparison, fairly easy. Qualitative case research is not easy to present, particularly not in the space-restricted journal article format. No generally accepted rules for how to present the research exist and there are no simple measures like significance levels and factor loadings. Perhaps we can hope that these rules develop, as the purchasing and supply management field matures in its development. But at present, researchers are often forced to invent presentation formats themselves. It is here that most of us will find that the strength of case research is also a large weakness. Properly conducted, case research generates large amounts of material that can lead to rich insights, but insights that can be hard to convey to others. The advice given by DuBois and Araujo is to ‘‘present the case to persuade the sceptical reader!’’ It is easy to accept this statement. However, I would like to make two qualifications to the statement. The first qualification is
ARTICLE IN PRESS P. A˚hlstro¨m / Journal of Purchasing & Supply Management 13 (2007) 216–218
that the most important readers to persuade are the reviewers. Obviously our main concern should be to do good research, but the reviewers are after all the gatekeepers for deciding what research ultimately gets published. Without being published, our research will never have any readers to persuade. These reviewers are not uncommonly steeped in a more quantitative tradition, particularly if we would like to target our research to USbased journals. There is a clear difference in research tradition between European and US scholars (Drejer et al., 2000). Persuading reviewers with a bias towards quantitative research may therefore necessitate emulating some of the ways of presenting research that are used within quantitative research. Being a seasoned case researcher myself, I obviously agree with DuBois and Araujo in that we must be careful in applying the standards of quantitative research onto qualitative studies. But from a pragmatic point of view, I would argue that the ends could justify the means. If we can persuade reviewers of the soundness of our research through learning from the quantitative research tradition, then we should. What particularly attracts me to quantitative research here can be summarized in the often-used, and indeed often-abused, term ‘‘rigour’’. I am here not advocating the use of the term to indicate ‘‘rigour mortis’’, which is all too often the case—paying undue attention to adhering to methodological standards and in the process ending up with uninteresting and trivial results. I would like to point to the importance of rigour as a way of highlighting that case research is often-times not well designed or conducted (and I am as guilty here as anyone). DuBois and Araujo highlight some of the hallmarks of good case research and there are several papers that provide excellent guides for designing and conducting good case research. The importance of good research design should not be underestimated. Badly designed and conducted research is hard, if not impossible, to present convincingly. Well designed and conducted research is the starting point for good research presentation. 2. Convince through illustrating the learning process that led to the conclusions The second qualification that can be added to DuBois and Araujo’s statement: ‘‘present the case to persuade the sceptical reader!’’ regards what is to be presented. The view DuBois and Araujo seem to take is that: ‘‘the best way to help readers evaluate a case is the presentation of the empirical material informed by the conceptual contribution’’. I would like to extend this line of argument. To my mind, it is neither the case nor the resulting theory that should be the sole focus of the research presentation. Rather, it is the learning process that led the researcher to the conclusions that need to be illustrated. The case is naturally part of this learning process, but not the most important part.
217
Following the advice given by DuBois and Araujo may be appropriate if it is a single case study that we are conducting. Then it may be possible to present the single case in all its richness, allowing the reviewers to test the author’s interpretation themselves. But single case studies suffer from problems of generalization, eloquently commented on by DuBois and Araujo. I am well aware of their points, having myself conducted single in-depth case studies and attempted (and at times also succeeded) in convincing reviewers. There are clear benefits of conducting multiple case studies. The challenge of research presentation is even greater in multiple case studies than in single case studies, since the amount of detail that can be conveyed becomes less. At the heart of the argument made here, to illustrate the learning process that led to the conclusions, is the nature of the theory building process. There is not a one-to-one link between the cases and the theory. We use cases to generate theory. Generating theory requires the researcher to make a creative leap, however small, from data to theory (Mintzberg, 1979). Theories are always abstractions and simplifications of a complex reality. Thus, even if there was space to let the cases ‘‘speak for themselves’’, it would not be the best way of convincing reviewers. What is critical in convincing the reviewers is to provide the linkages between raw data (the cases) and conclusions. Reviewers should, ideally, be able to see how the argument develops, from point to point, until the conclusions are reached. Naturally, the cases and their richness is the starting point here. However, it is necessary to simplify this richness. The need for simplification in case research is necessary also in the analysis process. In fact, there is a strong link between the analysis process and research presentation. Analysing qualitative data is a challenging task, since it often entails working with an overwhelming amount of descriptive data, with the risk of ending up in ‘‘data asphyxiation’’ (Pettigrew, 1990). Coding data is often one of the first tasks in the analysis process. Although we cannot capture the whole coding process, we can illustrate it using examples of how we arrived at different codes in our final paper (Stuart et al., 2002). The codes and not the cases themselves can then form a starting point for further analysis and indeed presentation. It is in making the link between cases and theory clear to the reviewers that we qualitative researchers can learn from more quantitative research traditions. The process of going from codes to the emerging theory, we can attempt to illustrate using tables, figures, conceptual models, numbers, classifications and even statistics. For an excellent example of usage of all these techniques in the cognate field of operations management, see Sousa (2003). Apart from simplifying the analysis process, these tools are also excellent presentation devices. They are particularly powerful if they are combined with logical text, which help show the reviewers how data and conclusions are linked. And the art of writing this logical text is in no small part a matter of rhetoric.
ARTICLE IN PRESS 218
P. A˚hlstro¨m / Journal of Purchasing & Supply Management 13 (2007) 216–218
Graphical formats are particularly helpful in convincing reviewers. Unfortunately, it is near enough impossible to make any general recommendations regarding the details of these graphical formats. They are highly contingent upon the individual research context and the research question. However, textbooks on data analysis can help, for instance Miles and Huberman (1994). Inspiration can also be found from other case researchers, particularly those who have managed to convince more quantitatively oriented reviewers. I would like to stress here the need to find inspiration, not to copy other researchers’ approaches. The most important source of inspiration is still the researcher’s own creativity. Although there are several handbooks on qualitative research, unlike for quantitative research, there are no generally accepted standards for how qualitative research is analysed and presented. But there are good and less good ways of presenting qualitative research, and we can hope that our research community is
on the path of developing more generally accepted standards for good research presentation. References Drejer, A., Blackmon, K., Voss, C., 2000. Worlds apart? A look at the operations management area in the US, UK and Scandinavia. Scandinavian Journal of Management 16, 45–66. Miles, M., Huberman, A., 1994. Qualitative Data Analysis: An Expanded Sourcebook, second ed. Sage Publications, Beverly Hills. Mintzberg, H., 1979. An emerging strategy of ‘direct’ research. Administrative Science Quarterly 24 (4), 580–589. Pettigrew, A.M., 1990. Longitudinal field research on change: theory and practice. Organization Science 1 (3), 267–292. Sousa, R., 2003. Linking quality management to manufacturing strategy: an empirical investigation of customer focus practices. Journal of Operations Management 21, 1–18. Stuart, I., McCutcheon, D., Handfield, R., McLachlin, R., Samson, D., 2002. Effective case research in operations management: a process perspective. Journal of Operations Management 20 (5), 419–433.