Examining students’ course trajectories using data mining and visualization approaches

• Published in: International Journal of Educational Technology in Higher Education Vol. 20; no. 1; pp. 55 - 18
• Main Authors: Maqsood, Rabia, Ceravolo, Paolo, Ahmad, Muhammad, Sarfraz, Muhammad Shahzad
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.12.2023; BioMed Central, Ltd; Springer Nature B.V; SpringerOpen
• Subjects: Colleges & universities; Computer Appl. in Social and Behavioral Sciences; Computer Science; Computers and Education; Course trajectories; Data acquisition; Data Analysis; Data mining; Education; Educational data mining; Educational Technology; Foreign Countries; Hierarchical clustering; Higher Education; Humanities; hybrid and online higher education: supporting students’ complex trajectories; In person; Information Retrieval; Information Systems Applications (incl.Internet); Institutional Evaluation; Law; Learning Trajectories; Markov chain; Pattern Recognition; Private Colleges; Research Article; Statistics for Social Sciences; Students; Transcripts (Written Records); Undergraduate Students; Visualization; Pakistan; Markov chain; Educational data mining; Hierarchical clustering; Course trajectories
• ISSN: 2365-9440; 2365-9440
• DOI: 10.1186/s41239-023-00423-4

The heterogeneous data acquired by educational institutes about students’ careers (e.g., performance scores, course preferences, attendance record, demographics, etc.) has been a source of investigation for Educational Data Mining (EDM) researchers for over two decades. EDM researchers have primarily focused on course-specific data analyses of students’ performances, and rare attempts are made at the domain level that may benefit the educational institutes at large to gauge and improve their institutional effectiveness. Our work aims to fill this gap by examining students’ transcripts data for identifying similar groups of students and patterns that might associate with these different cohorts of students based on: (a) difficulty level of a course category, (b) formation of course trajectories, and, (c) transitioning of students between different performance groups. We have exploited descriptive data mining and visualization methods to analyze transcript data of 1398 undergraduate Computer Science students of a private university in Pakistan. The dataset includes students’ transcript data of 124 courses from nine distinct course categories. In the end, we have discussed our findings in detail, challenges, and, future work directions.