On Quant And Qual
Cate Newsom | MARCH 22, 2013
Time and again clients come to us with the same story: “We’re really need a quant shop,” they tell us. “The board likes their numbers.” Well, so do we! In fact, I myself just came back from a Bayesian statistics workshop I won’t soon forget. As inspiring as the workshop was—and if you haven’t spent time with Bayes enthusiasts lately, they’re absolute fanatics—the instructors also reminded me of why there’s a time and a place for quant data, and why qualitative research is sometimes the only way to go.
The example that got me thinking had to do with the stork. Simplifying their message to a point that might make them cry, the workshop leaders demonstrated the importance of understanding the context of your data in order to analyze it effectively. They gave the example of a data set on the number of storks and babies born in local areas to test the hypothesis that storks bring babies. (Their model was also more complex that that, but that was the essence of it.) Sure enough, crunch crunch crunch, a significant p value emerged (along with some lovely graphs) indicating that we could not say that more storks don’t mean more babies.
Sure we’ve heard this message before (the 50’s classic How to Lie with Statistics and Disraeli come to mind). And no, it hasn’t stopped us from using and abusing quant to the fullest. But it sheds light on an issue that’s close to our hearts at Evviva (and the heart of our research practice): there’s a time for all good things, a time to quant and a time to qual.
The best part of qualitative research is that it allows us to find out what questions to ask. It goes in with an open mind, asking about storks and babies, but also notices that cabbage patch out back. And the stork’s status as a revered holy animal in the religious system of the local culture (whose proverbs also encourage at least three children per household). Let me stop with this here. The point is, qualitative methods can let you know when your stork is a red herring, and what line of questioning might produce more fruitful results.
Companies most often come to us when change is in the air—organic growth, mergers, new lines of business, transformation from having a global presence to being a truly global operation. Things are busy and there’s a lot of noise. Storks, cabbage patches, local proverbs and babies everywhere, so to speak. Using only quantitative methods would be like conducting a census in the middle of a crisis; what’s needed is a triage. We need to approach the situation using open-ended questions to find out what’s really going on. Future quantitative work will be informed by what we find and the results will be much more relevant, robust and reliable for it.
