In a scatterplot, each observation is a dot, and the location of a dot is determined by its value on the two variables. So it looks like on the whole, the further out from degree someone is, the higher their salary is. This is as to be expected---obviously, more experienced people get paid more.

## Discrete vs Continuous variables¶

One thing that has been in the background of our lesson thus far is the difference between *categorical* data, also known as *discrete* data, and *numerical* data, also known as "continuous" data.

Discrete data is, loosely speaking, data that takes a fixed number of values in your dataset. For example, in the dataset we're working with, gender only takes female and male, it's a discrete variable. Income, by contrast, is continuous because within its range, it could be anything. (Of course, strictly speaking, this isn't true, because you can't pay people fractions of a penny, and in our data we only have whole dollars. But it's close enough---these categories are kind of rough.) You could make salary discrete by putting it in *bins* --- if we created categories like "under 20,000", "20,000-25,000" and so forth, we could *discretize* that data.

This matters because different visualizations are more or less appropriate for different kinds of data. For example, scatterplots where one of the variables is discrete tend to be pretty useless. Take a look: