John Allen Paulos has written a series of books about how most people have a difficult time understanding the meaning of numbers. Researchers who have relied on numbers to tell a story shouldn’t be surprised by this. Even basic statistics can be hard to grasp let alone the complex Bayesian Math needed for complex efforts like Conjoint. Even though most of our clients are quite numerate, they often present results to those who are not. If we are to play, and help our clients play, an active role in decision-making we have to overcome this problem.
One of the examples that Paulos uses involves our inability to understand risk. In their new book, The Norm Chronicles: Stories and Numbers About Danger, Michael Blastland and David Spiegelhalter have tried to simplify things by boiling risk down to a simple number…the MicroMort. One MicroMort means you have a one in a million chance of death.
On the one hand, this does seem to simplify often complicated actuarial calculations such that we can see that soldiers in Afghanistan face a danger of 47 MicroMorts daily which is of course far more dangerous than your chances of death in a car crash (about 1 MicroMort per day) but far less than WWII bomber crews who were exposed to 25,000. The use of one number certainly simplifies things, but if someone is not great with numbers it might not resonate.
A second means they use is to convert numbers to “MicroLife” terms. So for example, a smoker’s life is cut short by five hours for each day they smoke. Or my favorite stat that your first alcoholic drink each day adds 30 minutes to your life…sadly a drink every half hour won’t get you immortality since each additional drink deducts 15 minutes. While still using numbers, these do at least present them in a clear relatable way. Of course I wonder how many smokers realize they are deducting a year of life for every five years they smoke?
Finding the right mix between numerical precision and understanding can be tricky and not just for research agencies. The key for us is to find the right mix between numerical precision and a clear message. We can’t get hung up too much on things like “statistical differences” (as our Quirk’s Article pointed out). Instead we need to focus on the decisions that need to be made and pull together a narrative that helps drive them. This certainly doesn’t mean we shouldn’t use numbers…just that we need to put them in the context of recommendations.
This can be a challenge when the numbers are simple, but it really becomes hard when using something like discrete choice conjoint. Providing an innumerate client with a simulator will likely take half a day off their life (28 MicroMorts/exposure). Providing them with a precise number (“the pricing research indicated that $43.68 was the optimal price to maximize profits) won’t provide the kind of confidence they need to implement the decision. There is no single correct way to do it (among other things it depends on the client’s individual preferences), but success requires that the answer be put in an easy to understand context and backed by a strong story to build confidence that it is right.
Rich brings a passion for quantitative data and the use of choice to understand consumer behavior to his blog entries. His unique perspective has allowed him to muse on subjects as far afield as Dinosaurs and advanced technology with insight into what each can teach us about doing better research.