AUTHOR=Gibson Douglas B. TITLE=Revised title: Effect size as the essential statistic in developing methods for mTBI diagnosis. JOURNAL=Frontiers in Neurology VOLUME=6 YEAR=2015 URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2015.00126 DOI=10.3389/fneur.2015.00126 ISSN=1664-2295 ABSTRACT=

The descriptive statistic known as “effect size” measures the distinguishability of two sets of data. Distingishability is at the core of diagnosis. This article is intended to point out the importance of effect size in the development of effective diagnostics for mild traumatic brain injury and to point out the applicability of the effect size statistic in comparing diagnostic efficiency across the main proposed TBI diagnostic methods: psychological, physiological, biochemical, and radiologic. Comparing diagnostic approaches is difficult because different researcher in different fields have different approaches to measuring efficacy. Converting diverse measures to effect sizes, as is done in meta-analysis, is a relatively easy way to make studies comparable.