ASA to Release the Recommendations of its Task Force on Statistical Significance and Replication

The American Statistical Association has announced that it has decided to reverse course and share the recommendations developed by the ASA Task Force on Statistical Significance and Replicability in one of its official channels. The ASA Board created this group [1] in November 2019 “with a charge to develop thoughtful principles and practices that the ASA can endorse and share with scientists and journal editors.” (AMSTATNEWS 1 February 2020). Some members of the ASA Board felt that its earlier decision not to make these recommendations public, but instead to leave the group to publish its recommendations on its own, might give the appearance of a conflict of interest between the obligation of the ASA to represent the wide variety of methodologies used by its members in widely diverse fields, and the advocacy by some members who believe practitioners should stop using the term “statistical significance” and end the practice of using p-value thresholds in interpreting data [the Wasserstein et al. (2019) editorial]. I think that deciding to publicly share the new Task Force recommendations is very welcome, given especially that the Task Force was appointed to avoid just such an apparent conflict of interest. Past ASA President, Karen Kafadar noted:

Many of you have written of instances in which authors and journal editors—and even some ASA members—have mistakenly assumed this [Wasserstein et al. (2019)] editorial represented ASA policy. The mistake is understandable: The editorial was co-authored by an official of the ASA.

… To address these issues, I hope to establish a working group that will prepare a thoughtful and concise piece … without leaving the impression that p-values and hypothesis tests…have no role in ‘good statistical practice’. (K. Kafadar, President’s Corner, 2019, p. 4)

Thus the Task Force on Statistical Significance and Replicability was born. Meanwhile, its recommendations remain under wraps. The one principle mentioned in Kafadar’s JSM presentation is that there be a disclaimer on all publications, articles, editorials authored by ASA staff, making it clear that the views presented are theirs and not the associations. It is good that we can now count on seeing the original recommendations. Were they only to have appeared in a distinct publication, perhaps in a non-statistics journal, we would never actually know if we were getting to see the original recommendations, or some modified version of them.

For a blogpost that provides the background to this episode, see “Why hasn’t the ASA board revealed the recommendations of its new task force on statistical significance and replicability?”

[1] Members of the ASA Task Force on Statistical Significance and Replicability

Linda Young, National Agricultural Statistics Service and University of Florida (Co-Chair)
Xuming He, University of Michigan (Co-Chair)
Yoav Benjamini, Tel Aviv University
Dick De Veaux, Williams College (ASA Vice President)
Bradley Efron, Stanford University
Scott Evans, The George Washington University (ASA Publications Representative)
Mark Glickman, Harvard University (ASA Section Representative)
Barry Graubard, National Cancer Institute
Xiao-Li Meng, Harvard University
Vijay Nair, Wells Fargo and University of Michigan
Nancy Reid, University of Toronto
Stephen Stigler, The University of Chicago
Stephen Vardeman, Iowa State University
Chris Wikle, University of Missouri

CHECK DATE OF THIS POST

REFERENCES:

Kafadar, K. Presidents Corner “The Year in Review … And More to Come” AMSTATNEWS 1 December 2019.

“Highlights of the November 2019 ASA Board of Directors Meeting”, AMSTATNEWS 1 January 2020.

Kafadar, K. “Task Force on Statistical Significance and Replicability Created”, AMSTATNEWS 1 February 2020.