Center For Biomedical Informatics Seminar
October 24, 2017
12:00 PM - 01:00 PM
Ryan Case Method Room, 1-9576
Anthony Almudevar, Ph.D.
Associate Professor of Biostatistics, University of Rochester Medical Center Center
TITLE: Reproducibility and Statistical Methodology
In recent years, the issue of reproducibility has become especially prominent, driven both by an increased emphasis on data driven science, as well as a number of initiatives aimed at assessing the frequency with which published findings can be duplicated using new experimental data. Some of these studies report distressingly low reproducibility rates, most notably, the Reproducibility Project undertaken by the Center for Open Science.
This seminar will review these findings, placing them in the context of traditional statistical methodology. In particular, we will show how a careful study of the common practice of using preliminary data to estimate sample sizes for future studies can shed considerable light on this issue.
Lunch will be provided.