Evaluating Multivariate Normality in Medical Datasets: A Case Study with R

Multivariate normality is a crucial assumption in many multivariate statistical methods, influencing the validity of medical data analyses. This study aims to develop and evaluate the multivariate normality of a dataset comprising biochemical parameters, specifically Total Cholesterol (TC), Urea, Creatinine (Creat), and Uric Acid (Uric). Using R and the MVN package, we developed a syntax to test for multivariate normality, applying Mardia’s skewness and kurtosis tests. The results indicated that the dataset meets the criteria for multivariate normality, with significant p-values confirming the assumption. Ensuring multivariate normality is essential for the validity of multivariate analyses in medical research. Our findings demonstrate that the biochemical parameters analyzed conform to the assumption, supporting their suitability for advanced statistical analyses. This study highlights the importance of verifying multivariate normality and provides a practical guide for researchers using R.