Why Most Published Research Findings Are False: Author's reply to Goodman and Greenland
2007
Why Most Published Research Findings Are False: Author's Reply
Editorial
Author Information
Author(s): John P. A. Ioannidis
Primary Institution: University of Ioannina School of Medicine
Conclusion
The author argues that many published research findings are not as credible as they are presented, primarily due to biases and the misuse of statistical significance.
Supporting Evidence
- Many positive findings in observational studies have p-values between 0.01 and 0.05.
- Bias can significantly affect the credibility of research findings.
- Replication of studies is necessary for genuine scientific progress.
Takeaway
Sometimes, scientists say their findings are more certain than they really are, which can confuse people and lead to mistakes.
Potential Biases
There is a strong presence of bias across various research designs, which can affect the credibility of findings.
Limitations
The article discusses the limitations of current research practices, including bias and the over-reliance on p-values.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website