Steve MacEachern, distinguished professor of statistics, discusses his field, how he got into statistics, and what feature he has to have in his job.

MacEachern earned a doctorate at the University of Minnesota, researching Bayesian methods, a way to use probability to represent uncertainty via prior and posterior distributions informed by data.

He sees statistics as different from data science/analytics because statistics emphasizes modeling and inference, while data science adds large-scale data capture and wrangling, often from computer science.

MacEachern discusses modern statistics’ growing focus on prediction in platforms like search and e-commerce, contrasts hard-science versus human-centered data, and outlines current research challenges driven by AI and massive predictor sets.