�鶹��Ѱ�����

. Patterns and Static is .��

When are we justified in drawing conclusions from data? For Department of Applied Mathematics (APPM) Teaching Professor Brian Zaharatos, that question opens a discussion that statistical methods alone cannot settle.��

Dr. Z's new book, , explores the philosophical foundations of statistical practice. Does probability describe patterns in the world, or express our uncertainty about it? ��What is inference? What allows us to conclude that one thing causes another?��

Released in summer 2026, the book invites readers to examine the assumptions behind how they use evidence to form beliefs and make decisions.��

The project grew out of STAT 4700/5700, Philosophical and Ethical Issues in Statistics, a course Dr. Z developed for undergraduate and graduate students in statistics and data science. The department supported the course’s development, with particular encouragement from Associate Chair Anne Dougherty and the late Keith Julien, then department chair.��

That support gave him room to pursue interests across disciplinary boundaries and bring them into the classroom. The course connected his background in philosophy with his work teaching statistics and data science. Those classroom questions eventually became the foundation for a book.����

“APPM allowed me to find my niche, both in interdisciplinary teaching and leadership work,” Dr. Z said. “Patterns and Static would not exist if not for their support.”��

Dr. Z studied mathematics and philosophy as an undergraduate, earned a master’s degree in philosophy, and completed a doctorate in applied mathematics and statistics. The book brings those interests together around questions that, in his view, deserve more attention in the standard statistics and data science curriculum.��

Students learn how to fit models, test hypotheses, and quantify uncertainty. Yet there are foundational disagreements in choosing and interpreting those methods. Should we evaluate a scientific theory by how probable it becomes in light of the evidence—the Bayesian approach? Or by how rigorously it has been tested and whether those tests would have exposed its flaws—the frequentist approach? These approaches offer different accounts of what makes an inference justified, with consequences for how practitioners design studies and interpret results.��

Patterns and Static introduces readers to the reasoning behind these perspectives and the challenges each faces.��

“I claim that we—statisticians, data scientists, and scientists more broadly—have much to gain from philosophical engagement and reflection,” Dr. Z said.��

In his view, understanding those debates has practical value for people who use statistics. It also offers an intellectual reward of its own: the opportunity to explore questions that remain contested even among experts.��

The book is intended for readers interested in statistics, data science, philosophy, or the broader problem of moving from evidence to conclusions. Although artificial intelligence is not a major focus, increasingly automated analysis gives these longstanding questions renewed relevance. Automating model building still leaves questions about the assumptions, interpretation, and justifications of those models. There is no way to abdicate our judgment to AI. Ultimately, we choose whether to accept AI output uncritically or develop the skills to evaluate it thoughtfully.��

��