I enjoyed the discussion about how much mathematics scientists should use in conveying predictive models to the general public. Prof. Goodman mentioned early on in the interview that in working on some publications regarding the COVID-19 pandemic for the New York Times, she was dissuaded from using mathematical terminology, and was notably asked not to use the word "denominator." Spiegelhalter responded with the well founded argument that the "denominator" is in fact one of the most useful data points in discussions about COVID-19 mortality. He argues that the denominator provides important context with which to understand the mortality rate, including population size and the demographics of the population. It was interesting to hear their later discussion about Prof. Goodman's weather prediction app and the statistics embedded in it, especially given the question about how much mathematical modeling non-scientists are generally able to digest. Spiegelhalter noted that even though the app may have seemed complex at first glance, through a few simple instructions from Prof. Goodman, he was able to understand it fairly easily. In his opinion, most users who come from a non-science background would similarly be able to grasp the analytics behind the program with some guidance.
One question I might pose to the experts, especially in the context of the New York Times anecdote, is whether they believe publications meant for a broader (non-scientific) audience are doing readers a disservice by omitting important mathematical information and parameters, especially when they are so crucial to the conversation. I would repurpose Prof. Goodman's initial question of how much math scientists should include to ask how scientists might be able to present their models in a way that engages and educates readers.