Distance Education Enrollment Grid Map

By Arie Spirgel, Senior Research Associate
Nova Southeastern University

​As institutional researchers, we spend a considerable amount of time thinking about geography. Where do our students come from? Where should we devote recruitment resources? From where do online students take courses? Maps are an obvious way to visualize this type of information, with the shading of regions, such as states, representing the measurement of some variable, such as enrollment. However, choropleths, as they’re called, suffer a major drawback: Each state takes up a different amount of visual space. One solution to this is the grid map, which assigns equal space to each state and places each one in its relative position. Besides the advantages of grid maps described here, when precision reigns over approximation – which is often the case for institutional researchers and stakeholders – Rhode Island offers exactly the same amount of space to add value labels as Alaska.

To create this grid map, for each public and private institution in the IPEDS Data Center, I downloaded its state, the total number of students enrolled in fall 2013, and the total number of students enrolled in exclusively distance education courses in fall 2013; this allowed me to calculate the percentage of students who took exclusively distance courses, by state. But more interesting than the data itself is the revelation that each state has an equal opportunity to catch the viewer’s eye. Take for instance Rhode Island, which had the lowest percentage of online enrollment, a fact that would be otherwise difficult to discern with a traditional choropleth.

I made this map using the ggplot2 package in R – a free, open-source language and environment – which can produce data visualizations that are limited more by creativity than functionality. Judging by my experience at AIR conferences, interest among institutional researchers in programming languages such as R outpaces their actual use. This is unfortunate because shifting from point-and-click software to R or its counterparts can fundamentally change the way you work as an IR professional. And, as someone with a non-computer-science background, I can attest that R is more intimidating to learn than it is difficult.

 

Fall2013OnlineStudents.JPG 
 

 

 
 

 Comments

 
To add a comment, Sign In
Total Comments: 8
 
Craig posted on 9/10/2015 1:02 PM
I like this. It is a unique, different way to present data in a geographical format. The squares and color make it clean and easy to read and the boxes do equate rather nicely to the location of the actual states.
Jeff posted on 9/10/2015 1:23 PM
The display is very effective for the reasons given. The focus is on the information rather than the geography. (This display could also be done in Excel using conditional formatting.)

More importantly, the information displayed is unique and begs our consideration. The densely populated Northeast and West Coast have more face-to-face instruction. Arizona is surprising in its high rate of distance education, and along with West Virginia and Iowa stand out prominently in this display.
Florence posted on 9/10/2015 1:34 PM
Very nice and innovative variation of regular maps. It looks a periodic table of elements to me at first glance. Now every state is represented at a same size square, thus readers would focus more on the content of the visualiztion instead of the geographic location.
Ghenet posted on 9/10/2015 4:15 PM
This is a great representation of the distance education data. Your navigation R is Very impressive.

The approach you used to extract the information will give us a better tool to further compare multiple years so that we can see the trend with DE. For instance, according to reports based of IPEDS' data, the University of Phoenix online campus had a better DE enrollment in 2012. It went down by about 20% in 2013. While the same academic year ASU & SNHU increased their DE enrollment by about 34% & 90%.
David posted on 9/14/2015 12:11 PM
This is great--thanks for sharing. Is there an R package to create these choropleths?
Heather posted on 9/25/2015 3:48 PM
This is a very effective way to display data; I really appreciate the simplicity of the design, yet it conveys the message well and doesn't, as the author points out, allow secondary variables such as relative size muddy the view. Would love to more about how it was created! Excel would work reasonably well as Jeff pointed out - is that how this chloropleth was created?
Arie posted on 9/28/2015 9:58 AM
Thank you for all of the feedback.

A couple of comments brought up using Excel to create grid maps, and I don't doubt that you could, but my advice would be to weigh the small cost of learning a programming language against the advantage it provides over using point-and-click software.

As I mentioned in the post, I used the ggplot2 package in R to create the map. If you have questions about how to do this, feel free to email me at spirgel at gmail dot com.
Terry posted on 10/7/2015 10:09 AM
This is just plain cool! And very interesting numbers.
​​