What can we learn from a Perfect Doctor? A statistician’s view – Biostatistics Seminar October 2nd by Fridtjof Thomas

What can we learn from a Perfect Doctor? A statistician's view
Presented by: Fridtjof Thomas, PhD, Professor in the Division of Biostatistics
Location: Freeman Auditorium, 3^rd floor of the 930 Madison Building, October 2, 2025, 12:00 - 1:00 pm CT

Join us this Fall for the Statistical Reasoning in Biomedical Research Seminar Series by the Division of Biostatistics in the Department of Preventive Medicine! The first talk is titled "What can we learn from a Perfect Doctor? A statistician's view" by Fridtjof Thomas, PhD, Professor in the Division of Biostatistics. Our exploration starts with observing a Perfect Doctor, who "magically" can pick the better treatment of two for any individual patient. We then derive the treatment effect for a small group of fictitious patients of that Perfect Doctor and contrast that estimate with the true treatment effect in our example, as well as estimates based on random assignments of the treatments. We will conclude that observing the treatment outcomes of the Perfect Doctor's patients does not allow us to estimate the true treatment effect, but that we can do so by randomizing treatment assignment. Participants will learn to distinguish between individual, average, and causal treatment effects and to apply causal inference terminology in the context of clinical studies. Participants will analyze the role randomization plays in clinical study designs, which will enable them to critically evaluate different designs of randomized as well as observational studies.

Mark your calendar: Additional talks in this series are scheduled for 10/30/25, 11/06/25, 11/13/25, and 12/04/25. (all 12:00 - 1:00 pm CT at the Freeman Auditorium)