For generations, the scientific method has evolved to become more rigorous, more self‑correcting, and more resistant to bias. One of the most important steps in this evolution was the rejection of anecdotal evidence as a legitimate scientific foundation. Today, we believe it is time for the next step in that evolution.

We, the undersigned, call upon the global scientific community — including academic institutions, journals, funding bodies, and professional organizations — to end the use of correlation as an accepted component of scientific reasoning, publication, and policy influence.

Correlation has long been recognized as a limited and often misleading statistical relationship. While it can suggest patterns, it cannot establish mechanisms, causation, or explanatory power. Yet correlation continues to be used as a basis for claims, headlines, policy recommendations, and even entire research programs. This practice has contributed to:

• misinterpretation of data
• public confusion about scientific conclusions
• the replication crisis in multiple fields
• the elevation of weak associations into perceived truths

Just as anecdotal evidence was once considered acceptable but is now recognized as insufficient, correlation should no longer be treated as a meaningful scientific tool. Modern science has access to far more rigorous methods — causal inference frameworks, mechanistic modeling, randomized controlled trials, longitudinal designs, and computational simulations — all of which provide deeper, more reliable insights than simple statistical association.

We therefore urge the scientific community to adopt the following reforms:

• Remove correlation-based findings from being considered publishable without causal evidence
• Require causal frameworks or mechanistic explanations for claims derived from data
• Revise statistical education to emphasize the limitations and dangers of correlation
• Discourage the use of correlation in policy, media communication, and public-facing science
• Promote research methods that prioritize causation, mechanism, and replicability

Science must continue to evolve. Eliminating correlation as an accepted scientific tool is a necessary step toward a more rigorous, transparent, and trustworthy scientific future.

We call on researchers, educators, institutions, and journals to join us in this reform.