AAA

Graphical analysis of educational data using Python

Jakub Swacha

Abstract

In practice, the high quality of educational processes and content can hardly be achieved and maintained without monitoring. And since tracking the results is not enough, it is necessary to find existing flaws and identify their causes. In this paper, it is argued that even small data sets such as students' test results can provide valuable information useful to improve further iterations of teaching and learning outcomes verification processes and educational materials. Next, the appropriateness of graphical analysis for this purpose is pointed out taking its simplicity even for non-statisticians and its ability to visualize entire small data sets with high precision into account. However, the primary aim of this paper is to provide practical examples showing that the Python programming language (with a selection of specialized modules) can be used in a convenient and effective manner for graphical analysis of small educational data sets. For the purpose of demonstration case study, actual test results were used. Analysis examined: correctness and confidence of students answering respective test questions, correlations between results based on question type and relevant content area, and also similarities of answers given by different students. Also, a Python-based software environment for graphical analysis has been compared to Microsoft Excel.

Bibliography

AUTHOR

Jakub Swacha

University of Szczecin

About the article

DOI: https://doi.org/10.15219/em64.1239

The article is in the printed version on pages 13-21.

pdf download PDF

pdf read the article (Polish)

How to cite