AG-R06: Statistical Reporting — Effect Sizes, Confidence Intervals, and Correct Test Reporting for GMJ

What to Report for Each Statistical Test and How to Avoid Common Errors

Abstract

The GMJ Author Toolkit (AG-R06) provides a structured, self-contained guide to statistical reporting in accordance with the Georgian Medical Journal’s submission standards. The guide ensures accurate, transparent, and reproducible reporting of statistical results across all manuscript types.

The document is organized into three components: (1) theoretical foundations explaining the three required elements for every statistical result — effect size, 95% confidence interval, and p-value — along with correct reporting formats for common statistical tests; (2) a comprehensive pre-submission checklist ensuring completeness, correct interpretation, and adherence to reporting standards; and (3) a complete worked example demonstrating correct and incorrect reporting for logistic regression, t-tests, and survival analysis.

Particular emphasis is placed on reporting effect sizes with confidence intervals, assessing normality for continuous variables, and avoiding common errors such as reporting p-values without effect sizes or confidence intervals. The guide also highlights the distinction between statistical and clinical significance and reinforces the need for transparent reporting of all results, including non-significant findings.

This resource supports authors in producing clear, accurate, and publication-ready statistical reporting for submission to the Georgian Medical Journal.