In his new book, The Signal and the Noise: Why So Many Predictions Fail – But Some Don’t, statistician Nate Silver talks about the importance of thinking like a Bayesian. He’s talking about Bayesian statistics, a branch of statistics that gives us powerful tools for making predictions, some of which Silver used to correctly predict 82 out of 83 results during the last election season (all 50 states in the presidential election plus 32 of the 33 Senate races).
What does this have to do with social entrepreneurship? Well, even though Bayesian statistics may sound like an abstract, academic concept to many, social innovators might know it by another name: design thinking. Bayesian reasoning and design thinking aren’t exactly the same thing, but the two parallel one another in a few important ways, and Bayesian reasoning can provide a useful framework for approaching social innovation.
Making predictions using Bayesian statistics is all about having some prior expectation for what might happen and then updating that expectation based on new evidence you receive. In Nate Silver’s case this meant starting off with some likelihood that Barack Obama would win reelection. Then some new polling data would come in and Silver would update his prediction. Then some new economic data would be released and he’d update again. Then new unemployment numbers, and so on until Election Day.
When Tim Brown, CEO of IDEO, talks about design thinking, he highlights three parts of the process: Inspiration, Ideation, and Implementation. Each of these steps finds a counterpart in Bayesian reasoning.
Inspiration comes from observing your target audience and identifying their needs. Ideation is the development of some new idea to help your target audience based on your observation. Implementation is putting your idea and assumptions into place and testing them out. This is another way of saying how you’re formulating some prior expectation for what you think will work, and then testing it out. What makes the process truly Bayesian, though, is its iterative nature. Design thinking is all about testing different prototypes, analyzing why each didn’t work, tweaking, and eventually working toward an idea that successfully meets your target audience’s needs. Likewise, Bayesian reasoning is all about incorporating new information to better understand our world. In Silver’s case, this meant correctly predicting election results. In the social entrepreneurs case, it means coming up with some product or idea that will improve people’s lives.
So maybe Bayesian statistics isn’t as abstract as we may have thought. If you’re a design thinker, you just might be a Bayesian, too.