Most charts describe the individual variation of one or more series. In some cases, we can intuit some relationship between two variables and even try to force that impression by manipulating the primary and secondary y-axes. But only when you plot a variable against the other (typically using a scatterplot) can you observe their relationship.

In a basic scatterplot, you associate a variable to the x-axis and another to the y-axis. You can encode a third variable into the z-axis if you can display the scatterplot in 3D. For example, you may want to explore the relationship between education and wealth, defining a metric for education (percentage of the population with tertiary education) and wealth (Gross Domestic Product per capita) and associating them to the x- and y-axis. The relationship will be interpreted depending on the pattern created by the cloud of data points (one for each entity, like countries).

The insights that you take from this are the most relevant ones. But there are complementary insights that can be interesting, and you can unveil them by adding more variables. For example, you can vary the size of each dot depending on population size. Or you can add a categorical variable, using color to encode the continent.

Animation and the connected scatterplot

The most common type of scatterplot displays multiples entities at a given moment. You can use animation to show how the relationships change over time, but you should use it carefully if the temporal pattern is intricate. See below a well-known analysis of a scatterplot (bubble plot) and the change over type.

You can also show how a single entity’s relationship changes over time in a chart known as a connected scatterplot. The scatterplot is not easy to interpret and currently is not a familiar chart type. Still, it deserves more attention because it adds valuable insights beyond the typical display (one or two line charts).