How can we change beliefs? A Bayesian perspective

Area of Science:

• Philosophy of Science
• Medical Research Methodology

Background:

• Bayesian philosophy explains how new evidence updates prior beliefs to posterior beliefs.
• Randomised Controlled Trials (RCTs) are a key source of evidence in scientific reasoning.

Purpose of the Study:

• To explore the capacity of RCTs to alter established beliefs within a Bayesian framework.
• To examine the limitations of RCT evidence in mechanism elucidation and paradigm shifts.

Main Methods:

• Analysis of Bayesian reasoning principles in scientific belief updating.
• Evaluation of the evidential limitations of RCTs concerning intrinsic effects and placebo hypotheses.

Main Results:

• RCT evidence can falsify the placebo hypothesis but cannot identify specific mechanisms of action.
• Subjectivity and differing prior paradigms, as seen in homeopathy debates, hinder belief updating.
• Accepting new paradigms requires open-mindedness and consideration of all hypotheses, superseding mere Bayesian calculations.

Conclusions:

• While Bayesian analysis provides a structure for belief updating, RCTs have inherent limitations in overturning established beliefs.
• The inherent subjectivity and differing paradigms, particularly in controversial fields like homeopathy, present significant barriers to belief consensus.
• Genuine paradigm shifts necessitate an open-minded approach that transcends purely mathematical Bayesian updating.