Search

The article reports on the use of statistical testing. It states that among the consequences of statistical testing using fixed levels of significance is the potential for this practice to lead to publication bias. The author notes that history does not provide the sort of randomized experiments R. A. Fisher would have advocated.

The article focuses on the importance of double-blind, randomized controlled study in medicine, physics, and forensic science. Topics discussed are historic examples of blind tests which include judging for the best wine by British purveyor Steven Spurrier and proving the theory of animal magnetism; example in medicine include checking for side effects of statin; need for blind forensic analysis for avoiding unconscious form of bias.