Abstract:

Everyday, a lot of data is generated in a lot of different environments like economics, medicine, business or consumer preferences. Another environment that produces a lot of data and which analysis has grown in the last years is the educational one.

The processing and analysis of educational data for finding aspects of interest and even, detecting problems and identifying their possible causes is known as Educational data mining. Another branch related with this analysis is using visual elements for presenting data in a way that allows finding useful information is known as Information Visualization, applying these techniques in an educational environment is called Visual Learning Analytics.

Visual Learning Analytics also can be separated into different areas, one which considers both information visualization and the use of data mining techniques for finding patterns, the other one presents the data in a more comprehensive way but does not consider a deeper analysis using data mining techniques.

This approach which can be named Visual Learning Statistics can be applied in several aspects which can be of interest at the moment of analyzing different elements related with topics like the academic performance of students.

This work presents the use of several visualization techniques for presenting information about different aspects about engineering students. Main goal is showing the advantages of presenting data in a visual way, which more than looks attractive, the purpose is identifying some behaviors in an initial stage even before applying some data analysis techniques.

Some of the analyzed data is related to the importance of certain students' characteristics in finishing or not their studies.

Another application of visual learning statistics was oriented to present the progress of the students in a certain group of Mathematics courses. Simpler graphics were used for presenting the performance of students before and after the COVID-19 pandemic and helped to determine if it really had a significant effect on them.

Results of different works show that using a visual representation of, originally, raw data could work as a first approach for finding interest patterns which can be used in deeper analysis. Also, considering different visualization techniques more than the traditional can allow obtaining information in a simpler way.

This work is part of a bigger project which considers using several visualization techniques applied to educational environments whose main goal is finding problems that could be affecting the academic performance of engineering students, but also, can be applied to other levels in an educational system.

Keywords:

Information visualization, visual learning statistics, students' performance, educational data analysis.