The ASA serves as the central professional organization for statisticians in the United States, representing the field's full breadth—from theoretical foundations to applied practice, from frequentist to Bayesian approaches. While the ASA is not exclusively a Bayesian organization, its sections, journals, conferences, and policy statements have been instrumental in the acceptance and growth of Bayesian methods within the broader statistical profession.
Historical Context
Founded in Boston in 1839, the ASA is one of the oldest professional societies in the United States. For much of the twentieth century, the American statistical profession was dominated by frequentist methods, and Bayesian approaches were often viewed with skepticism. The latter decades of the century saw a gradual shift, as computational advances—particularly MCMC methods—made Bayesian approaches practical for complex real-world problems. The ASA's embrace of Bayesian methods, through the establishment of the Section on Bayesian Statistical Science (SBSS) and the inclusion of Bayesian content at JSM, reflected and reinforced this shift.
In 2016, the ASA issued its landmark statement on p-values and statistical significance, warning against the mechanical use of p-value thresholds for scientific decision-making. This statement implicitly supported the Bayesian perspective that statistical conclusions should be expressed in terms of degrees of evidence and uncertainty, rather than binary significant/not-significant decisions. The statement catalyzed a broader conversation about the role of Bayesian methods in improving scientific practice.
ASA and Bayesian Statistics
The ASA supports Bayesian statistics through several channels. The Section on Bayesian Statistical Science (SBSS) organizes sessions at the Joint Statistical Meetings and sponsors student paper competitions. ASA journals, including the Journal of the American Statistical Association (JASA), The American Statistician, and Statistical Science, regularly publish Bayesian research. The association's continuing-education programs offer short courses on Bayesian methods, and ASA-sponsored working groups have addressed the use of Bayesian approaches in regulatory science, clinical trials, and forensic statistics.
Key Contributions
The ASA's role in the Bayesian story extends beyond organizational support. By providing a "big tent" that brings together researchers from different statistical traditions, the ASA has facilitated the intellectual exchange that has enriched both Bayesian and frequentist methodologies. The association's commitment to promoting sound statistical practice—through guidelines, educational programs, and public engagement—has helped ensure that Bayesian methods are applied thoughtfully and rigorously.
"The ASA exists to serve all statisticians and to promote the practice and profession of statistics. The growth of Bayesian methods is one of the most important developments in our field, and the ASA has a responsibility to support and encourage this growth."— ASA leadership