Inspiration

This project’s goal is to enable the Arizona State University (ASU) computer science Undergraduate Program Committee (UPC) to evaluate whether any anomalies are found within a data set outlining the grade distributions in core CSE (Computer Science and Engineering) classes over seven semesters. From there, the UPC can assess what (if any) course level modifications should be recommended, and present findings to the CSE faculty in an easily understood manner. Note that due to privacy issues, professor names/class numbers could not be included in the visualizations.

Goal of this Project

To develop a set of interactive visual analytics tools to facilitate the exploration of grade distributions in eight core ASU CSE courses and aid in the analysis of any inconsistencies which might exist.

Breakdown of the interface

Average distribution charts: Displays the average grade distributions of CSE classes 205-360. Examples in Figure 1.

Data matrix: Displays the hierarchical clustering of grade distribution data; the darker red the square is, the more dissimilar the two distributions. Figure 2 shows the correlation matrix for CSE 310.

Individual class distribution charts: Displays the grade distributions for every class provided in the data. All classes are grouped by CSE number and semester provided. All distributions can be viewed two ways: with the full display of grades (e.x. A+, A, A-, B+, etc.) or the aggregate display of grades (A-E with no +/- designations). Examples in Figure 3.

What's next

• Yearly or bi-yearly snapshots can be used as opposed to semesterly snapshots to provide a different perspective on the data or to check if collecting by semester creates bias in the data.

• Incorporating professor accountability into the model, such as RateMyProfessor scores and course evaluations, will also help with further analysis.

Share this project:
×

Updates