[pocketsand]

Why? They’re non-parametric and make zero assumptions of normality.

[blueflow]

How else would you calculate the quartiles to render the boxes?

[munch117]

Count data points in each quartile. You can do that for any sortable data, independent of distribution.