TAS Special Issue: Statistical Inference in the 21st Century: A World Beyond P<0.05

The ASA Symposium on Statistical Inference was held October 11–13 at the Hyatt Regency Bethesda with more than 400 people in attendance. Energized by two days of inspiring presentations and ample opportunities for discussion, the work and conversation continues with a special issue of The American Statistician (TAS).

The inspiration for the special issue is the ASA’s Symposium on Statistical Inference, which followed up on the ASA’s Statement on P-Values and Statistical Significance. The statement called for moving statistical analysis and evidence-based decision-making beyond “bright line rules” toward a “post p < 0.05 era.”

Although the problems identified in the statement have been known for decades, previous expressions of concern and calls for action have not fostered broad improvements in practice. The expectation is that the symposium and this special issue of TAS will lead to a major rethinking of statistical inference, aiming to initiate a process that ultimately moves statistical science, and science itself, into a new age.

The special issue will be online only by late July and remain open access permanently, making it readily accessible to a broad research community and users of statistics.

Generally, papers in the special issue will fall into one or more of the following categories:

  • Conducting research in the 21st century
    • Differences between scientific inference and formal statistical inference
    • Designing research for effective inference (and avoiding common inferential mistakes)
    • Measuring and expressing uncertainty
    • Role of judgment and expert opinion
    • Lessons learned from efforts at reform of statistical inference, including de-emphasis or banishment of statistical significance testing
    • Best uses of p-values
    • Alternatives to p-values, including Bayesian and other methods
  • Using research in the 21st century
    • Understanding decision-makers’ needs for statistical evidence
    • Making evidence available and accessible via, for example, collaborations and clearinghouses
    • Making decisions under uncertainty
  • Sponsoring, disseminating, and replicating research in the 21st century
    • Incentives for better research in research funding and job performance reviews
    • Improving peer review and editorial decision-making, including the structure of the scientific paper in the post-p < 0.05 era
    • Strategies for improving reproducibility and replicability
  • Statistical education and training in the 21st century
    • Introductory undergraduate (or high-school) statistics courses
    • Researcher training and continuing education
    • Materials for decision-makers, journalists, and other such audiences
    • Software and tools that make best practices easily accessible
  • Prescriptions for a post-p < 0.05 era

Key Dates

  • June 15, 2017
    Early Registration and Housing Opens
  • September 12, 2017
    Early Registration and Housing Deadline
  • September 13, 2017
    Regular Registration Open (increased fees apply)
  • October 11, 2017 – October 13, 2017
    SSI in Bethesda, MD