The importance of surprising findings in psychology

David Trafimow

doi:10.3389/fpsyg.2013.00173

Abstract

The importance of surprising findings in psychology

Highlights

Trafimow (2003) used the famous theorem by Bayes to show that if a surprising prediction is confirmed by experiment, it creates more of a difference between the prior and posterior probability of a theory than if an unsurprising prediction is confirmed by experiment
The history of science is replete with examples of extremely important experiments that successfully confirmed surprising predictions made from theories such as Einstein’s theory of relativity
Theory-based predictions rely on (1) the theory and (2) auxiliary assumptions that provide the link between non-observational terms in theories and observational terms in predictions that derive from theories

Summary

Introduction

Trafimow (2003) used the famous theorem by Bayes to show that if a surprising prediction is confirmed by experiment, it creates more of a difference between the prior and posterior probability of a theory than if an unsurprising prediction is confirmed by experiment. If the predictions work out, the theory would be considered to have passed a strong test, and there would be an impressive difference between its prior and posterior plausibility.

Results

Conclusion