There is an opportunity, says John Griffiths, to change the way we analyse research by building communities of interpretation.
The compactness of qualitative research has been with us for so long that we take it for granted. A solitary practitioner can sample a population thanks to a fieldwork company and then run the project in a linear fashion from proposal to debrief, most projects being small enough to fit inside a single mind.
If it weren't for this, qualitative research would not be the cottage industry it is. Sole practitioners wouldn't tender against international agencies — which they do. It suits us. Yet whenever a market suits suppliers they will persist long after the practice ceases to be relevant or competitive. I think we have got to that point. We take for granted our intermediary status. Today's marketers, however, have access to data which hasn't come from a research intermediary and have more customer contact than fieldwork companies have ever delivered. Customers — not all, but enough — are taking the plunge and are willing to collaborate and to cocreate. So what's an intermediary to do now?
Our greatest added value, analysis and interpretation, has always been half hidden. There is now the added risk that clients, in the excitement of getting answers online before they have even properly framed their questions, will tend to believe almost everything they are told, thus undervaluing our contribution.
Analysis and interpretation is the difference between what the client thinks as they get into the taxi after the last group and what they learn subsequently in the debrief presentation. If we persist in analysing in a linear fashion, though, we have to persuade our clients to wait for the added value that takes us time to produce. If we don't then increasingly we will be judged by the instant debrief on the night.
Calling all Canutes
So here's a suggestion to get the Canutes out of their chairs before the incoming tide washes their feet. Consider whether we could apply co-creation and collaboration to the process of analysis and whether five people might be able to do in one day what takes a single person five. We could do this by splitting the task of analysis and interpretation between them, either asking each one to represent a particular audience, or to follow particular themes. For offline research it would require the fieldwork to be transcribed before analysis begins.
The fundamental difference in this approach is twofold. The unit of analysis is thematic instead of the aggregated and nuanced opinions of a discussion group with a particular sample composition. Secondly, it uses a team of analysts where each one has a different perspective and whose analysis is itself a stimulus to the other members of the team as it is debriefed.
The point of network analysis is that each analysed perspective simultaneously informs all the others. Combining different perspectives amplifies the value by the number of participants. This cannot be done using linear analysis. The primary benefit to the business carrying it out is increased speed of analysis because of the way the network draws insights out through structured interaction between the analysts.
I have run some very simple experiments getting people to analyse transcriptions from different perspectives — sometimes as many as a dozen. The risks in taking a thematic only approach would be only to collect findings according to the preselected themes. But having audience owners counteracts this. The danger with only analysing by units of fieldwork is an over reliance on what respondents say without picking up on the underlying themes.
An online variant would deploy the analysts as scouts going after Facebook, YouTube, Delicious and a range of other sources, then reporting back to the group after having conducted analysis on their separate source material. This is particularly useful for online ethnography projects and could also be used to amplify offline projects with online content at the analysis stage as a corollary. It's difficult to do if analysing serially — but straightforward enough where there is a team of analysts.
What online line research of this kind enables us to do is to quickly gather the context of how an issue is covered and written and talked about online. By using a range of perspectives we can sample a greater diversity including retailers, and even the client company's own web content. This type of online data gathering could still be regarded as part of the analysis, closer to desk research than to online fieldwork.
There are still more advanced levels of this technique involving respondents — or even clients — in the analysis process. This makes it possible for us to apply workshop tools which are rarely, if ever, applied to analysis now. The sophistication comes from the interplay of analysts, not what is done to the data. The simple adoption of group analysis, though, should be quite powerful enough as described.
This should be an area where qualitative researchers flourish because our expertise lies as much in stimulating fruitful group interactions and as it does in "interrogating" respondents. It extends the management of group dynamics from the viewing facility to the room where the analysis is done.
I don't expect most people reading this to change what they do. My hunch, however, is that research agencies with the staffing levels to experiment with this approach will have an advantage over those practitioners who don't.
End to level playing fields
Building communities of interpretation could be the development that ends the level playing field where large agencies analyse in exactly the same way as small agencies despite the very great differences in their human resources. We tested this out in an exercise at AQR's July Online session and we'll be looking forward to testing it out further at the next one.
Copyright © Association for Qualitative Research, 2010