How to Graph Ordinal Variables

Arrange the scale of your ordinal variable in a manner consistent with your hypothesis. Suppose, for example, that you want to display the effect of political enthusiasm on probability of voting. If you hypothesized that a higher level of enthusiasm for politics will correspond with a greater likelihood of voting, do not create an ordinal scale in which 5 equals apathy and 1 equals very enthusiastic. This will result in a negative slope on your graph.

Determine whether the ordinal variable is the dependent or independent variable. The dependent variable is affected by the independent variable. Consider this example from political science: National stability is affected by level of democracy. National stability is the dependent variable; level of democracy is the independent variable. It is likely (though not universally) the case that the ordinal variable will be the independent variable.

Draw an x-y graph, or create one using your graphing software. This graph looks like the letter "L," with the x-axis taking the horizontal line and the y-axis taking the vertical line.

Place the dependent variable on the y-axis and the independent variable on the x-axis. If there is more than one independent variable, you will need to configure a three-dimensional graph or to simply display your regression results in a matrix format.

Create a numerical system to represent your ordinal values. For example, your ordinal variable may be level of education. In this case, you could assign the number 1 to "completed high school"; 2 to "some college"; and 3 to "finished undergraduate school."

Space out the distance between the ordinal ranks (in othe example, the numbers corresponding to education level) evenly on your graph, as you would with an interval variable. Do this even though the distance between ranks in ordinal variables is not equal (i.e., the difference between completing high school and obtaining a B.A. is greater than the distance between obtaining a B.A. and obtaining an M.A.). You should be able to tell that you are working with an ordinal variable because you will graph a curve rather than a line.

Plot your data. Suppose that your dependent variable is income and your independent variable (also your ordinal variable) is a 5-point scale describing an individual's level of happiness at the workplace. If an individual is a 4 on the happiness scale and makes $30,000 per year, place a point above 4 on the x-axis and next to $30,000 on the y-axis, so that 2 imaginary lines going out from the point would intersect perpendicularly with each of those values on the axes.