Bioelectromagnetic research and statistical power

Abstract

Bioelectromagnetic field research has some way to go before the scientific community at large is convinced that weak, extremely low frequency electromagnetic fields affect biological systems. This paper argues that researchers are doing themselves and the discipline an injustice by running studies where statistical power is very low. The most frequent outcome of low statistical power is a null result, even when there are experimental effects to be detected. It is suggested that future research must address the issue of statistical power, both through increasing it in individual studies and through conducting meta-analytic investigations of existing research.