Why most published research findings are false

Area of Science:

• Scientific methodology
• Research integrity
• Biostatistics

Background:

• Growing concern exists regarding the reproducibility and validity of published research findings.
• The reliability of scientific claims is influenced by study power, bias, and the landscape of existing research.

Purpose of the Study:

• To analyze the factors influencing the truthfulness of research claims.
• To discuss the implications of widespread research bias on scientific interpretation and conduct.

Main Methods:

• Conceptual analysis of research practices and statistical principles.
• Exploration of factors affecting the probability of true research findings, including study power, effect sizes, and research flexibility.
• Simulation-based assessment of research claim validity.

Main Results:

• Research findings are less likely to be true with smaller study sizes, smaller effect sizes, and increased flexibility in research design and analysis.
• The ratio of true to tested relationships significantly impacts claim validity.
• Simulations indicate that research claims are more likely to be false than true across many study designs.

Conclusions:

• Many current research findings may represent accurate measures of prevailing bias rather than genuine discoveries.
• The conduct and interpretation of scientific research require critical re-evaluation in light of potential widespread bias.