< Back to Volume 13, Number 1
Volume 13: Number 1: Article 1
Significance Levels for the Assessment of Anomalous Phenomena
Robert A. J. Matthews, Department of Computer Science, Aston University,
Birmingham B4 7ET, United Kingdom
Scientific evidence for anomalous phenomena is frequently supported
by conventional measures of statistical significance such as p-values.
However, these measures have been shown to be unreliable indicators
of the existence of genuine effects, and routinely exaggerate the true
significance of experimental data. They are, moreover, especially unsuitable
for the assessment of anomalous phenomena. More appropriate statistical
techniques are available, but pose their own problems when applied to
anomalous phenomena. I outline an approach to hypothesis testing which
allows conventional measures of significance to be retained, while offering
substantially lower risk of seeing significance in chance effects.
Keywords: statistical significance, p-values, Bayesian inference
FULL TEXT:
Significance Levels for the Assessment
of Anomalous Phenomena
To purchase back issues contact Allen Marketing & Management: 1-800-627-0629
