Aggregating concept map data to investigate the knowledge of beginning CS students

Abstract

Concept maps have a long history in educational settings as a tool for teaching, learning, and assessing. As an assessment tool, they are predominantly used to extract the structural configuration of learners’ knowledge. This article presents an investigation of the knowledge structures of a large group of beginning CS students. The investigation is based on a method that collects, aggregates, and automatically analyzes the concept maps of a group of learners as a whole, to identify common structural configurations and differences in the learners’ knowledge. It shows that those students who have attended CS education in their secondary school life have, on average, configured their knowledge about typical core CS/OOP concepts differently. Also, artifacts of their particular CS curriculum are visible in their externalized knowledge. The data structures and analysis methods necessary for working with concept landscapes have been implemented as a GNU R package that is freely available.

Keywords:

• Concept maps
• Pathfinder networks
• knowledge structures
• cluster analysis
• educational data mining
• OOP

Notes

No potential conflict of interest was reported by the author.

1 For the time begin, please contact the author if you are interested in using this editor yourself. It will eventually be available publicly.