Transmutability and Interconnectedness: New Criteria for a New Era of Qualitative-Quantitative Collaboration

quant and qual: why can’t we be friends?

First, a very sincere thank you to all of you for maintaining your loyalty to this blog, and even growing it in my temporary absence. August 2008 was a new record for Brandthroposophy at over 40,800 unique visitors. The blog continues to gain steam, and it’s just beginning, I promise you that. I owe its success to all of you, Gentle Readers. The Brandthroposophy Community, you who read, who contribute, who keep coming back.

As those of you who follow me on Twitter already know, the last month has been a
very busy one for me. I am on my sabbatical from York University in Toronto. In
August, I moved to Sydney to begin teaching at The University of Sydney in a
Visiting Professor’s position. The preparing, moving, and teaching have kept me
very busy. I just completed an intensive Ph.D. course on Research Methods with a
group in the Faculty of Economics and Business at “Sydney Uni.” I had students
from a variety of different majors, not only marketing, but finance, organizational behavior, logistics and transportation, business law, accounting, and even economics.

Bernardo, Eva, Amrita, Frank, Zuraidah, Melissa, Aaron, Hee Soo, Howard, Monathip, and Phong: the students were great, and I learned a lot from the course.

After reviewing a lot of material, I decided to teach the course using a brand
new textbook by John Creswell called Research Design: Qualitative, Quantitative,
and Mixed Methods Approaches, 3rd Edition
, just published at the end of August
2008 by Sage.

My entire approach to the course was agnostic. I didn’t favor any type of
approach over any other. I certainly wasn’t going in as a qualitative method or
consumption studies advocate. And I tried to fill the course with readings about
the philosophy of research design as well as how to “do theory” to “write research” and to do it well.

A major concern that I have heard in my field of consumer research and marketing that concerns the so-called “Qualitative-Quantitative Divide” goes something
like this. A number of the researchers who work with qualitative data have
complained that the researchers who do not work with this type of data, or with
cultural theory work, or whatever you want to call it, do not use and build on
the theory development that has been done in the cultural arena. A lot of quant work doesn’t reference qualitative work in the same theoretical or conceptual area. From my experience, I think there is some truth to that.

However, to see things from the perspective of those researchers who do theory-
testing work, a lot of the material that qualitative researchers produce is (1) divorced from the mainstream of quantitative, psychological, econometric work in marketing and consumer research and, perhaps even more importantly, (2) presented in a way that is not easily transferred into propositions, hypotheses, and theory-testing approaches. In the past, I would say that “big ideas” or theories coming from qualitative, cultural work that had these qualities of being connected to extant work and formed relatively easily into scales or other formats easily tested were picked up on and used across the entire field. Two examples would be Russ Belk’s work on materialism, and Al Muniz and Tom O’Guinn’s work on brand communities.

Now this assertion leaves aside a lot of subtle arguments about incommensurability and reification. I know that. We can definitely go back to that, and I plan to, in writing for other formats. But for now what would happen if we were to follow a simple guideline that says we are after semi-solid social rules, trying to track them as they change and flow and interpenetrate in a dynamic environment, and that we can get at them through a variety of means of measurement using many instruments, including the researcher-as-instrument? Our ontology and epistemology would be agnostic and free-flowing, as Creswell seems to be implicitly advocating in his book on research design, a book whose key innovation is the detailed new guidelines and frameworks it gives for combining qualitative and quantitative research into mixed methods approaches.

I am actually not at ease with the way Creswell sets aside these fundamental
differences. He slides those assumptions away all too easily, never mentioning
them, as if he can just smuggle them out the back door. It doesn’t sit easy with me, as my Sydney Uni students already know.

I also think that our unit of analysis is frequently very different. When an
anthropologist studies culture, it can not readily be reduced to a dummy variable
in a regression equation, or turned into a neat set of 7 scale items with
internal validity and a respectable Cronbach’s alpha. Culture just doesn’t work
that way. But it does end up having individual elements and effects, and some of
those may certainly be interesting and testable. They have been in the past.

But those major concerns aside, I’m wondering what if we set ourselves as a
field in consumer and marketing research some new criteria for judging our
research. What would it look like? How would it change?

What if we asked qualitative researchers to think about the “transmutability,” the alteration of the form of their research findings into testable propositions and hypotheses? We already have a wide range of models, many of which look like they have causal elements in them. What if we went the extra distance, just a bit, and started to build our findings into something that looked more like testable propositions, or pre-propositional statements that specified the most important constructs and the relationships between them? After all, that is what theories are supposed to be. And why not make them testable if we can do this without doing violence to our theories or harm to our deeper understanding?

We certainly wouldn’t be the only field to pick up this format. As I learned in my
investigations, qualitative MIS and organizational behavior studies often work
out this way. And even in our own field, we have research work like my esteemed anthropologist collegue Eric Arnould’s study of world systems theory in Zinder province, Africa, that is written in a very similar form.

Now, on the corresponding side of quantitatively-focused researchers, the non-
CCT crowd doing consumer information processing, attitudes and motivation, goal setting, choice modeling, behavioral decision theoretic and other approaches.
What I’d like to suggest to them is that they think about “Interconnectedness”as a criterion for their research. Interconnectedness would mean that their studies draw upon and cite qualitative studies and findings that suggest hypotheses for them to test and verify, or refute and nuance.

In that way qualitative work would do what it does best, open up new areas for
study, and propound new theories that have external validity. Quantitative work
would test those theories, help to verify and refine them across various
contexts. Qualitative research would also be involved in ongoing investigations of the theories across different contexts (social, geographic/national, temporal), nuancing them. Qualitative work would add naturalistic veracity, depth, complexity and richness. Quantitative work would simplify so that we understood the key constructs and verify that the relationships held.

Perhaps not all qualitative research could fit into this pattern, partake in
this idea. Perhaps.

But in my next posting I’m going to take this idea bit further, and play with it. I’m going to see if some of even the most expressive, emotive, introspective and descriptive of consumer research forms can still, perhaps, yield propositions testable and verifiable by other researchers working with quantitative methods.

Controversial? Oh yeah. Career suicide? Probably not. Stay tuned.

Now it’s off to Bondi beach for me. I’ve got some Ph.D. papers and final exams to grade. G?day, mates!