All Times EDT
To non-statisticians we sound like the protectors of a secret language: “[My] first original paper [was on] analytic and algebraic topology of locally Euclidean parameterization of Infinitely differentiable Riemannian manifold.” (pace Tom Lehrer). Just as bad is when we talk down to non-statisticians – oversimplifying what we are doing so they will “understand”. That approach hides our assumptions. Somehow we need to be able to talk to non-statisticians in a way that captures what we do in a way that is not too complicated but not too simple. As always, we should follow Goldilocks: our language must select a level of tone and complexity that is “just right”. Our goal should be to explain our methods clearly and accurately enough so our collaborators can enter into a meaningful dialogue with us. We are most effective when we and those with whom we work tell a mutually meaningful story.