Biostatistics Event

Ann Arbor MI 03-29-2018 03-29-2018

In light of recent concerns about reproducibility and replicability, the ASA issued aonSigniand p-valuesat those who are notstatisticians. While the ASA State-ment notes that statistical signiand-values are "commonly misused and misinterpreted," it doesdiscussdocument broader implications of these errors for the interpretation of evidence. In this, we review research on how applied researchers who are notprimarily statisticians misuse and misin-terpret-values in practice and how this can lead to errors in the interpretation of evidence. We also presentdata, perhaps surprisingly, that researchers whoarestatisticians are also prone toand misinterpret-values thus resulting in similar errors. In, we show that statisticiansto interpret evidence dichotomously based on whether or not a-value crosses the conventional.05forsigni. We discuss implications and or recommendations. Light refreshments for seminar guests will be served at 3:10 p.m. in 1690

Statistical Significance and the Dichotomization of Evidence

Blakeley McShane, Ph.D., Associate Professor of Marketing - Northwestern, Kellogg School of Management

icon to add this event to your google calendarMarch 29, 2018
3:30 PM - 5:00 PM
1690 SPH I
1415 Washington Heights
Ann Arbor, MI 48109-2029

Sponsored by: Department of Biostatistics
Contact Information: Zhenke Wu (zhenkewu@umich.edu)

In light of recent concerns about reproducibility and replicability, the ASA issued aonSigniand p-valuesat those who are notstatisticians. While the ASA State-ment notes that statistical signiand-values are "commonly misused and misinterpreted," it doesdiscussdocument broader implications of these errors for the interpretation of evidence. In this, we review research on how applied researchers who are notprimarily statisticians misuse and misin-terpret-values in practice and how this can lead to errors in the interpretation of evidence. We also presentdata, perhaps surprisingly, that researchers whoarestatisticians are also prone toand misinterpret-values thus resulting in similar errors. In, we show that statisticiansto interpret evidence dichotomously based on whether or not a-value crosses the conventional.05forsigni. We discuss implications and or recommendations. Light refreshments for seminar guests will be served at 3:10 p.m. in 1690