Statistical Manuscript Review Tutorial-Statistical Manuscript Review Tutorial

Video Transcription

Video Summary

John Shepard gives an overview of how to review manuscript statistics, especially in urogynecology. He explains the main data types: continuous and categorical, and how distribution determines which tests to use. For normally distributed continuous data, he discusses the t-test, paired t-test, and ANOVA; for skewed data, the Mann-Whitney U test and log transformation. For categorical data, he explains row versus column percentages, chi-square, and Fisher’s exact test. He then introduces regression: linear regression for continuous outcomes and logistic regression for binary outcomes, emphasizing the importance of adjustment for confounders and interpretation of beta coefficients, odds ratios, and confidence intervals. He also covers equivalence and non-inferiority trials, explaining margins and confidence interval interpretation, and briefly mentions survival analysis with Kaplan-Meier curves, log-rank tests, and Cox regression. Overall, the talk aims to help reviewers judge whether authors used appropriate statistical methods and whether their conclusions are valid.

Keywords

manuscript statistics

urogynecology

t-test and ANOVA

chi-square and Fisher exact test

linear and logistic regression

Kaplan-Meier survival analysis