Statisticians Who Watch Focus Groups
A client had a surprising experience this week when a member of our multivariate analysis team showed up online to watch a live in-depth interview with a registered nurse about how prescribing decisions are made. “Who is that online with us?” the end-client inquired, not recognizing the name. The qualitative manager answered, “He’s on our multivariate team.”
It must have seemed strange to have a statistician taking a keen interest in the qualitative work. Strange, because too often the qualitative and quantitative sides of research do not inform each other in the rich ways they can and should.
On the very same day, we read this interview with Professor Jordan Louviere in the American Marketing Association’s Marketing News magazine:
The field [of product choice research] is really in desperate need of theory and a lot of what goes on in choice modeling is nothing more than statistics. And while there . . . is theory in statistics, it’s not the theory we need. It’s important, but statistics is just a tool. We need theory that tells us and informs us much better than in the past how consumers actually do what they do and what are the best ways to approximate that?
So we have, sort of, a mismatch in the field right now with people. . . who are trying to understand how consumers make decisions and choices and other things, and we have choice modelers, who, by and large . . . are statisticians and the two groups don’t speak to each other.
We are proud that at Versta Research the statisticians doing the conjoint choice modeling participate in, and often lead, the qualitative work. It ensures that the choice models are truly built to reflect and understand the processes at work. Our models are more than rote applications of attributes and levels, analyzed with fancy tools that do Hierarchical Bayes (HB) estimation. They are models of the decision processes our clients care about, designed by quantitative experts who know their stats, but who are immersing themselves in the real world. The end result? Our clients get deeper insight than is typical, and they get better answers to their most pressing questions.
—Joe Hopper, Ph.D.
