Marc Leonti, Raed Binsaeed, Nick Hu
Service-Learning Course Project as part of DS 4200 F20: Information Visualization , taught by Prof. Cody Dunne , Data Visualization @ Khoury , Northeastern University .
Our service partner is the Sustainable Path Forward Task Force from Salem State University (SSU), which has been charged with making recommendations for increasing revenue and reducing expenses. Their goal is to reallocate spending in areas that drive enrollments (increasing revenue) and decrease spending in areas that do not drive enrollments (reducing expenses). The following visualizations explore the spending patterns of post-secondary institutions to search for patterns that may reveal their impacts on enrollment. If patterns are found in these visualizations, a deeper analysis of the underlying data can determine if they are statistically significant. Proper guidance can then be given to make fiscal policy changes that will boost enrollments.
We have included data from following basket of nine schools, as they are appropriate for comparison based on region (Massachusetts), sector (four-year public institutions), and size (1,000 to 15,000 students).
The data for the three University of Massachusetts schools have been averaged and is represented as one entity (gray). The University of Massachusetts: Amherst, Medical School, and Law School have been intentionally omitted from this basket. The average of the MA State Schools is also displayed (pink).
The task force expressed they would like to be able to see the rate of growth of enrollments expressed as the number of Full-Time Equivalent Students, distribution of spending as a percentage of total spending, and time-series information showing the year over year differences in spending calculated per Full-Time Equivalent Student. This has been accomplished with the use of a line graph to show time-series enrollment data, a violin plot to show the distribution of spending in eight major categories, and individual line graphs to show the time-series detail of sending in each of those eight categories.
In each of the line graphs, additional details are displayed when hovering over any data point. Hover over any ⓘ icon for additional suggestions.
The adjacent line graph shows the number of Full Time Equivalent (FTE) Students reported by
each school, which is a measurement equivalent to one student enrolled full time for one academic
year. This is the best way to measure enrollments when making comparisons between universities, as
it captures full-time, part-time, and partial-year enrollments by calculating, based on the
aggregate total number of credit hours, and allows for direct comparisons regardless of school
size or total budget.
ⓘ
We can see with this visualization that most of the schools in this analysis have reported
positive growth over time, but Salem State University (purple) has shown steady negative
growth since 2010. Bridgewater State (red) has the most FTE Students of the MA State schools,
and the UMass schools (gray) have the highest average enrollment among all MA public schools.
Both schools have reported steady growth over the timeframe.
Since these schools
seem to be getting it right, we would look for similar spending patterns between these
schools to find a correlation between certain expenses and enrollment changes.
Below the legend, violin plots show the distribution of all the data in each expense category. Each colored dot represents a school's spending for a particular year, as a percent of their total budget. A gray violin body gives an alternate representation of the distribution, read similar to a sideways histogram. The height of each violin represents the range of the values. The width of each violin represents the frequency of the values. This provides an overview to capture the variance of spending in each category.
Below, the line graphs show each school's spending trends, with each expense category isolated to a single chart. Each line graph corresponds with one violin. The top line graph represents the same data as the far left violin. The bottom line graph represents the same data as the far right violin. The data displayed are the dollar amounts spent annually per FTE student. This provides a direct comparison between schools, and unlike percentages, is not affected by changes in other categories.
These capabilities will allow the end user to explore the dataset and hopefully discover useful insights. While an in-depth analysis will be required to determine if the findings are statistically significant, and is beyond the scope of this project, we hope this tool will help users direct the focus of their research. Depending on those findings it can possible to identify which expenses help boost enrollment, which expenses should be reduced, and ultimately convince decision-makers in affecting financial policy changes.
I would like to thank everyone who helped make this possible: first, to Dr. Kurt von Seekamm from Salem State University, for your unwavering guidance and comical support; to Kyle Langford, for your technical advice and honest critiques; to Professor Cody Dunne, for putting things into perspective; and especially to Leanne, for making me fix my grammar errors, waiting patiently to eat dinner, and warming up my cold feet when I climbed into bed at 3am.
- Marc