When to Use
Bivariate proportional / graduated symbol maps combine two datasets (usually numerical data) into a single hybrid map symbol. They are very efficient because the size of the symbol tells you one thing and the color/fill tells you another. These maps exemplify the power of bivariate maps by allowing us to easily examine geographic relationship between two variables (like income and education). They inherit many of the strengths and weaknesses or univariate proportional symbol maps, outlined here. Like single-variable graduated symbol maps (in which the size and color of the symbol show the same data), an important decision here is whether or not to group your data into classes or to show unfiltered “raw” data (assuming your data aren’t already classed for you, in which case, the decision is moot).
This is a 5-class x 5-class bivariate graduated symbol map that uses natural breaks classification method and sequential color scheme. While this map uses orderable (ordinal or ratio) data for both size and color, color on a bivariate proportional symbol map could also encode categorical data represented by a qualitative color scheme.