To identify an appropriate subsample for analysis, we are going to use exploratory graphs to visually sort through the data. This is one approach that will be supplemented by a measure of interdependence called correlation. A correlation coefficient measures the linear dependence between two variables; in our case, tuition revenue and instructional expenditure per student.
First, let’s consider the graph below. What trends do you observe among the different

institutional types? Are there any apparent outliers? Any shared traits among the outliers?

It might be hard to identify trends using the static graphs. Below are two dynamic graphs that sort institutions by their operational structure (private, public, for profit, not for profit) and by the highest degree they award. By clicking on the label in the legend to the right of the graph, you can add or remove groups from the graph.

#### Dynamic Graphs

**Arizona Institutions (2009-2010). Public/Private/For-profit**

#### Arizona Institutions (2009-2010). Highest Degree Awarded

#### All Institutional Types

]]>