Regency A
Adventuring Beyond P < 0.05 (303967)
*Karen Grace-Martin, The Analysis Factor
*Tom Gwise, FDA
*Megan Higgs, Independent consultant
*Dan Jeske, UC-Riverside
* Ruixiao Lu, Genomic Health
*Wendy L. Martinez, Bureau of Labor Statistics
Keywords: p-values, statistical inference, statistically significant
In spring 2019, the The American Statistician (TAS) published a special issue titled “Statistical Inference in the 21st Century: A World Beyond p < 0.05”. The articles in this issue challenge statistical practitioners to “venture beyond p < 0.05” and stop using the term “statistically significant”. While the special issue provides suggestions for methods and practices for moving beyond statistical significance, it is less clear what the best practices are for achieving a world beyond p < 0.05 and how the statistical practitioner should discuss this new paradigm with their collaborators. The challenge is particularly exceptional for collaborators accustomed to using the p < 0.05 threshold. In this panel session, we will discuss adventures beyond p < 0.05 with applied statisticians from industry, academia, and government. We begin by providing a brief introduction and background to the “beyond p < 0.05” movement laid out in the TAS special issue. We will then discuss what this movement has meant for statisticians with different backgrounds and elicit advice on best practices for statistical collaboration and communication in the modern paradigm of statistical inference.