Amid debate on reproducibility and p-values, experts gather to improve scientific evidence, rigor
ALEXANDRIA, Va. (September 26, 2017) – Following the widespread impact of its historic statement on the overuse and misinterpretation of p-values throughout the scientific community, the American Statistical Association (ASA) is convening scientists spanning many disciplines to help bolster statistical and scientific evidence, move beyond an irrational overdependence on p-values, and improve the rigor and reproducibility of research. This first-of-its-kind event, the Symposium on Statistical Inference, takes place October 11-12, 2017, in Bethesda, Maryland.
"Statistical analysis is key to conducting research in any field of science, and the challenge and pressures of reproducibility are impacting the perception and path of worthwhile research," said ASA President Barry D. Nussbaum. "While the ASA's statement addressed the inappropriate explanations and uses of p-values and significance tests, this conference is an imperative next step to not only further the dialogue, but also drive change that leads to lasting improvements in multidisciplinary research, communicating and understanding uncertainty, and decision-making."
Discussions will center on specific approaches for refining statistical and scientific evidence as it intersects with conducting, using, sponsoring, disseminating, and replicating research. Upon the conclusion of the symposium, recommendations, approaches, and guidelines will be published and made available to the entire scientific community, policymakers, members of the media and the general public.
So important is this focus to the broader scientific community, that it is co-sponsored by the American Psychological Association and American Educational Research Association. Financial support for the conference has been provided by the Central Intelligence Agency (CIA), R Studio, and the ASA.
"Policymakers here and abroad look to American science as a vibrant source of insight and an agent of change that can help steer understanding and adoption of policies that better society. Having a collective and realistic understanding of what will improve research reproducibility will make science across all disciplines stronger and more meaningful," said Marcia McNutt, president of the National Academy of Sciences.
* Steve Goodman, Associate Dean of Clinical and Translational Research; professor of Medicine and Health Research & Policy; co-founder of the Meta-research innovation Center (METRICS), Stanford University, and John Ioannidis, the C.F. Rehnborg Chair in Disease Prevention, professor of Medicine, Health Research and Policy, Statistics, Biomedical Data Science; co-director, METRICS; director of the PhD program in Epidemiology and Clinical Research, Stanford University, will lead the opening plenary session.
* Andrew Gelman, professor of statistics and political science; director of the Applied Statistics Center, Columbia University; Xiao-Li Meng, dean of the Graduate School of Arts and Sciences, Whipple V. N. Jones Professor of Statistics, Harvard University; and McNutt, will provide the closing plenary titled "The Radical Prescription for Change."
A full lineup of speakers and the program schedule can be found on the symposium website. Registration space has been expanded due to demand. Media can attend SSI for FREE but must pre-register by contacting Jill Talley, ASA Public Relations Manager, at [email protected]