Key points are not available for this paper at this time.
Abstract We consider several aspects of data analysis that are underemphasized in most presentations of statistical theory and practice. We illustrate some of these with a simple example of Bayesian workflow and conclude by emphasizing shared aspects of Bayesian and non-Bayesian data analysis workflows. The audience for this paper includes statisticians and applied researchers who might not be aware of these commonalities across apparently opposing statistical philosophies. This article is part of the theme issue ‘Statistical workflow’.
Gelman et al. (2026) studied this question.