Visualizing multidimensional health data poses challenges in selecting methods that effectively reveal patterns and separations. This study evaluates five visualization techniques for maternal health risk data: scatter plot matrix, parallel coordinates, RadViz, principal component analysis (PCA), and multidimensional scaling (MDS). Both standardized and normalized data are used to assess group separation effectiveness. Direct visualization methods and PCA show limited separation, especially for medium-risk. MDS with Manhattan distance and standardized data provides the best separation. Results show that the visualization method determines the ideal scaling approach, with no single technique universally optimal for multivariate health data.
This work is licensed under a Creative Commons Attribution 4.0 International License.