How to eradicate fraudulent statistical methods: statisticians must do science.

Area of Science:

• Biostatistics
• Toxicology
• Genetics

Background:

• Clinical expression of mutagenic diseases depends on genetic damage and cell viability.
• These two factors act as opposing forces, influencing the dose-response relationship of mutagens.
• The concept of 'context-free' analysis in science is critically examined.

Purpose of the Study:

• To mathematically derive and support the principle that mutagen dose-response curves must exhibit a maximum.
• To expose the fraudulent application of statistical methods in scientific research, particularly within U.S. government agencies.
• To advocate for biostatisticians to engage directly in scientific research to prevent statistical fraud.

Main Methods:

• A novel mathematical derivation was employed to demonstrate the existence of a maximum in mutagen dose-response curves.
• Supporting data were presented to validate the theoretical derivation.
• Examples of fraudulent statistical methodologies used in biomedical research were identified and discussed.

Main Results:

• The study mathematically confirms that the interplay between genetic damage and viability necessitates a maximum point in the mutagen dose-response curve.
• Linear extrapolation of high-dose mutagenic risk data to estimate low-level risks is identified as a common fraudulent practice.
• Several other instances of statistically fraudulent methods in U.S. government-related biomedical research are detailed.

Conclusions:

• The scientific principle that mutagen dose-response curves have a maximum is mathematically demonstrable.
• The misuse of statistical methods, especially in risk assessment, constitutes scientific fraud.
• Empowering biostatisticians to conduct their own science is proposed as a long-term solution to eradicate such fraud.