Visual displays should effectively and succinctly communicate key information.
Visual displays in Statistics are often referred to as graphs or charts. We saw in Visual Displays – One Variable how the choice of graph to represent a single variable is determined by the variable type. So too when we wish to display pairs of variables, the variable types of the pair of variables determine which visual display is appropriate.
The visual display’s title, and each of the vertical and horizontal axes must be clearly labelled, informative and unambiguous.
Graphs summarise data to communicate key information clearly and succinctly. Patterns in the graphs displaying data from two variables provide information about the relationship or association between the two variables.
The appropriate graphs to display each combination of variable types are:
Both numerical → scatterplot
- Both categorical → clustered bar chart
- 1 numerical and 1 categorical → histograms vertically aligned (or side by side boxplots)
Before You Watch