American Statistical Association

American Statistical Association (ASA) is a United States-based nonprofit professional association that promotes the practice, education, research, and responsible use of statistics. Founded in Boston in 1839, it brings together statisticians and other professionals who develop, interpret, teach, and apply statistical methods in academia, business, government, healthcare, and other fields. Its stated mission is to promote the practice and profession of statistics, and its vision emphasizes the use of data and statistical thinking to inform discovery and decisions.

The ASA operates as a professional and scientific association, rather than as a government statistical agency, software vendor, or regulator. It publishes research, convenes conferences, supports statistical education and professional development, maintains ethical guidelines, offers professional accreditation, and speaks on issues affecting statistical practice. Its guidance is relevant to organizations that depend on statistical analysis, data science, surveys, experimentation, forecasting, and quantitative evidence. The ASA's guidance helps establish expectations for statistical work, but the organization does not certify the accuracy of every study that cites its publications or follows its recommendations.

History and development

The association was formed in November 1839 by five founders in Boston, reflecting growing interest in using quantitative evidence to understand public health, population, commerce, and public affairs. Its early history was closely connected with the collection and interpretation of social and government statistics. The ASA broadened its membership and geographical reach as statistical applications spread across universities, industries, and public institutions.

The Journal of the American Statistical Association began publication in 1888. During the twentieth century, professional chapters and specialist sections developed as statistics became integral to scientific research, quality improvement, government programs, and commercial decisions. The association later expanded its journals, meetings, education programs, and policy activity. Its current remit includes statistical approaches used in machine learning and artificial intelligence as well as classical survey design, estimation, and inference.

The ASA's history illustrates the transition of statistics from a narrower set of government and scientific practices into an interdisciplinary profession. Its members work across national borders, even though the association is headquartered in the United States.

Mission and principal activities

The ASA's constitution, effective 27 April 2026, identifies meetings, publications, membership services, education, accreditation, and advocacy as means of advancing statistical science. It also identifies professional ethics and conduct as part of the association's responsibilities.

Activity	What the association provides	Relevance to statistical practice
Research publication	Peer-reviewed journals and other publications	Dissemination and critical examination of methods and applications
Professional standards	Ethical guidelines and statements on statistical interpretation	Expectations for transparency, integrity, and responsible inference
Education	Educational guidance, professional development, and learning resources	Support for statistical literacy and practitioner competence
Professional recognition	Accreditation and other career-oriented programs	Recognition of qualifying practitioners and continuing development
Community	Sections, chapters, meetings, and conferences	Exchange of research, experience, and specialist methods
Public and professional advocacy	Statements and engagement on matters affecting statistics	Communication of statistical considerations in public decision-making

These activities have different kinds of authority. A journal article presents the findings and arguments of its authors; a Board statement expresses an institutional position; and a professional guideline states expectations for statistical conduct. None should automatically be read as binding law or as a guarantee that a particular analysis is correct.

Ethical Guidelines for Statistical Practice

The ASA's Ethical Guidelines for Statistical Practice are among its principal professional resources. The version approved by the ASA Board on 1 February 2022 addresses everyone engaged in statistical practice, regardless of whether their formal job title is statistician. It covers data collection, preparation, analysis, interpretation, presentation, and the development and deployment of models and algorithms.

The guidance highlights several recurring responsibilities:

• Select methods appropriate to the question, data, and intended use.
• Disclose material assumptions, data limitations, sources of bias, and relevant uncertainties.
• Report results honestly, including findings that conflict with a preferred conclusion.
• Differentiate exploratory analysis from analyses intended to test a pre-specified proposition.
• Protect confidentiality and follow applicable legal, organizational, and research obligations.
• Maintain professional competence and disclose conflicts of interest.
• Communicate limitations that could change how stakeholders interpret or use results.

An important distinction is that responsible statistical work begins before model fitting. Inadequate sampling, unclear definitions, selective data exclusions, or misleading visual presentations can undermine a technically sophisticated analysis. The guidelines therefore concern the full evidence process, not simply the choice of statistical test.

The guidelines are professional guidance, not a substitute for laws, contractual duties, research-ethics review, or sector-specific requirements. The ASA's Committee on Professional Ethics maintains and promotes the guidelines but does not act as a tribunal that adjudicates individual ethical disputes.

Statistical significance and interpretation

A prominent ASA contribution to research communication was its 2016 statement on statistical significance and p-values. The statement responded to widespread misinterpretation of statistical tests and described six principles governing their appropriate use. In particular, it explained that a p-value does not give the probability that a research hypothesis is true, that statistical significance alone does not establish the importance of an effect, and that conclusions should not depend solely on crossing a conventional threshold.

For executive decision-making, the implication is straightforward: a result described as statistically significant may still be small, biased, poorly measured, or irrelevant to the decision. Conversely, evidence may be useful even when an individual test does not pass a conventional significance cutoff. Interpretation also requires attention to study design, effect sizes, uncertainty intervals, prior information, data quality, and the consequences of error.

The ASA statement does not prohibit use of p-values. It calls for their interpretation within a fuller account of the evidence. That distinction matters in marketing research, clinical research, product experiments, operational testing, and quantitative risk assessment.

Publications and research dissemination

The association publishes and sponsors journals covering both theory and applied statistical practice. These include the Journal of the American Statistical Association (JASA), The American Statistician, Journal of Business & Economic Statistics, and Journal of Computational and Graphical Statistics. Its publications portfolio also includes journals oriented toward survey methods, public policy, and particular application domains.

Publication provides a route for dissemination and scholarly scrutiny. However, readers still need to assess whether a paper's population, measurement, assumptions, period, and methods are applicable to the decision they face. A peer-reviewed result does not become universally transferable merely because it appears in a respected journal.

The association also publishes material designed for practitioners, students, and the broader public. These resources have different purposes and levels of technical detail; users should distinguish introductory explanations from primary methodological research.

Meetings, sections, and chapters

The ASA supports professional exchange through subject-specific sections, local chapters, and conferences. It participates in the Joint Statistical Meetings (JSM), a major annual gathering of statisticians and data scientists involving several professional societies. JSM includes research presentations, discussion sessions, training opportunities, and professional networking.

Sections address areas of statistical theory and practice; chapters create geographic professional communities. These structures help practitioners compare methods, understand changing professional expectations, find collaborators, and sustain specialist knowledge. Participation does not, by itself, establish that a member's work meets a particular quality threshold.

Education and professional accreditation

Statistical education is part of the ASA's public and professional mission. The association supports instructional guidance, including the Guidelines for Assessment and Instruction in Statistics Education (GAISE), which address statistical learning in school and college settings. The relevant reports should be identified by audience and publication version rather than treated as one universal curriculum.

The ASA also offers professional accreditation. Its Accredited Professional Statistician designation (PStat) recognizes qualifying statistical education, experience, and professional development. Graduate Statistician (GStat) is a distinct early-career accreditation route that recognizes educational preparation for the profession. Accreditation criteria and application requirements are maintained by the association.

Accreditation should be distinguished from membership: joining the ASA is not equivalent to receiving PStat or GStat status. Accreditation is also not an endorsement of each statistical claim, report, model, or product produced by a credential holder.

Governance and public engagement

The ASA is governed through its constitution, bylaws, Board of Directors, and organizational bodies that represent its professional communities. Its Board can approve institutional statements on issues affecting the profession. Such statements express the association's position and should be attributed accordingly.

The association has addressed the role of statistics in data science and artificial intelligence, statistical significance, the integrity of public statistics, professional ethics, and other subjects. For example, its Board statements register includes material on ethical AI principles for statistical practitioners and on the role of statistics in data science and artificial intelligence.

The ASA's advocacy role should not be confused with the functions of a federal statistical agency. A federal agency may collect and publish official statistics under statutory authority; the ASA is a membership association promoting statistical science and professional practice. It may comment on official statistics or professional conditions without itself producing the authoritative government dataset.

Relevance to business, technology, and CIO leadership

Organizations increasingly rely on data-based conclusions when allocating capital, choosing technologies, managing cybersecurity, assessing employee or customer experience, and evaluating AI systems. The ASA's professional work is relevant wherever evidence quality influences consequential decisions.

For a CIO, its guidance offers useful questions to ask of internal analytics teams and external suppliers. Is the study's population defined? Does the dataset represent that population? Were measures chosen before results were examined? What uncertainty remains? Are observed associations being presented as causal effects? Can another qualified analyst reproduce the steps? Is the reporting transparent about exclusions and limitations?

These questions apply to business intelligence dashboards as well as advanced models. A precise chart can still describe an incomplete dataset. A model that performs well on historical records can degrade when business processes, customers, or external conditions change. And a claimed lift from a technology investment may require experimental or quasi-experimental evidence rather than simple before-and-after comparison.

The ASA is therefore a source of professional guidance and research, not a ready-made enterprise data governance framework. Adopting its principles generally requires translation into practical policies, methods, reviewer responsibilities, reproducibility requirements, and escalation processes appropriate to an organization's risks.

Illustrative enterprise example

A hypothetical company evaluates whether an AI-assisted customer-support system reduces service-resolution time. A supplier reports a statistically significant improvement based on a pilot involving selected teams. Leadership considers expanding the system across the business.

An evidence review informed by ASA guidance would first examine how pilot teams and cases were selected, whether resolution time was defined consistently, and whether other operational changes occurred during the pilot. It would compare absolute time savings and uncertainty with service quality, customer outcomes, and escalation rates. Analysts would disclose missing records and assess whether the pilot population resembles the wider service operation.

The review might find that the apparent improvement is real but limited to simple cases. It might also identify insufficient evidence to support a general rollout. In either event, the business decision would be better grounded because the evaluation separates measured results, uncertain inferences, and assumptions requiring additional testing.

Relationship to other organizations and disciplines

Organization or field	Primary focus	Relationship to the ASA
American Association for Public Opinion Research (AAPOR)	Public opinion and survey research methods and disclosure	Related professional body with a stronger survey and public-opinion specialization
American Marketing Association (AMA)	Marketing scholarship and professional practice	Related association; marketing studies may use statistical methods and research standards
Institute of Mathematical Statistics (IMS)	Mathematical statistics and probability research	Separate scholarly society that also participates in collaborative statistical meetings
Data Science	Interdisciplinary methods for extracting knowledge and building data-driven systems	Field of practice in which statistical reasoning and ASA guidance can be relevant
Statistical Analysis	Methods for summarizing, modeling, and inferring from data	Technical discipline, whereas ASA is a professional association

These distinctions prevent a common category error: the ASA is the organization; statistical methods are tools; ethical guidance is professional advice; and application-specific rules belong to the relevant regulatory and institutional authorities.

Limitations and considerations

The relevance of a particular ASA publication depends on the use case. A technical journal article may address a specific model under narrow assumptions; an educational guideline may focus on introductory teaching; and an institutional statement may seek to improve professional interpretation rather than prescribe a single analytical method.

Statistics also contains legitimate methodological differences. Guidance on testing, inference, modeling, and uncertainty should be applied by appropriately qualified practitioners who can explain both chosen and alternative approaches. The association's publications can inform review, but they cannot substitute for access to complete data, relevant subject knowledge, independent validation, or organizational accountability.

See also

• Statistical Analysis
• Data Analysis
• Data Science
• Business Intelligence
• Predictive Analytics
• Market Research