Abstract:

Educational institutions generate a lot of information about many aspects, being one of the most relevant the performance of their students in different courses, however, this data is most of the times presented in a textual or numerical way which difficult its understanding. Consider, for example, the historical registry of students in a certain group of courses, finding how the students are behaving along these ones could be a difficult task looking only the raw data. However, an appropriate graphic representation would help for finding patterns in a simpler way.

For example, finding a course that it is presenting more problems to the students could be a difficult task just reviewing textual data, also, looking just numbers or percentages, sometimes does not give a clear idea of what is really happening. So, it is necessary a different approach of presenting data for its analysis.

Nowadays, the amount of information that is generated and gathered is so big that analyzing it in a textual or numerical way results in a complicated or even an impossible task. Humans have an ability of perception that allows us to understand in an easier way graphic information than textual one. Presenting data in a graphic way that allows their understanding is the study field of Visual Information.

This work presents the analysis of the performance of students over a set of five mathematic courses supported in different graphic representations: graph, box plot and bubble plot. These graphs give a more complete idea about the performance of students than reviewing textual data, analyzing statistics tables and even, other classic visualizations such as bar or pie graphics, which are not suitable for presenting several criteria about one topic.

This study is performed in Mexican Autonomous Metropolitan University (UAM for its Spanish acronym) which basic mathematical branch for all engineering students is composed by five courses:
- Mathematical Workbench
- Precalculus
- Differential Calculus
- Integral Calculus
- Differential Equations

These courses represents one of the most difficult branches in all study plans and students could take many time for finishing them, however this information is not obvious when it is reviewed in a textual way. For this reason, the fact that a course could be generating a called “bottleneck” goes unnoticed.

Students performance is measured more than marks, in time and tries that a student take for approving one or several courses. Graduate time is a problem, many students take two or three times more than the considered time for finishing the basic mathematical branch, and not a small quantity do not even finish it, but again, viewing this in a textual representation or using simple graphic visualizations does not give an idea of the problem.

In the present work, the performance of students is analyzed considering:
Number of tries invested in approving every course of the branch (a maximum of five at UAM), presented with a graph visualization.
Relationship of marks (at UAM, Very Good, Good and Sufficient) and tries per course, presented with a bubble plot.

Number of scholar periods (at UAM, quarters) invested in finishing every course, presented with a box plot. An interpretation of these graphics is performed for finding interesting patterns in a faster way than reviewing the available textual information.

Keywords:

Mathematic courses path, Students performance analysis, Technology in education, Visual learning analytics, Visualizing educative information.