Rethinking the Use of Box Plots in Data Visualization

The article discusses the author's decision to stop using box plots due to their poor design, making them unintuitive, hard to grasp, and prone to misinterpretation. The author argues that alternative chart types are generally easier to learn how to read and more informative. The flaws in the traditional box plot design are examined in detail, including visual design issues and the requirement for audiences to understand complex concepts like quartiles. The author recommends using alternative chart types such as strip plots and distribution heatmaps, which can communicate a wide variety of insights without the need for complex understanding. The article concludes by encouraging practitioners to consider alternative chart types and to challenge the notion that box plots are the best choice in most situations.