Daniel Spikol, Malmö University

Main Supervisor:

Jalal Nouri, DSV


Jelena Zdravkovic, DSV



For the student, the graduation project is in most universities the final step towards graduation and increased opportunities in the professional career. It is not uncommon for students to struggle with the thesis in order to complete their graduation, resulting in disrupted plans, delays in completion and the worst case, non-completion of their degrees. This outcome is undesirable for not only the individual student or the university but also society at large.

The effects of a failed thesis are waste of resources and the supply of qualified individual in the society, for the student, the effects include lesser possibilities in the career, lower pay. Thus, the dissertation is addressing the problem of the management of resources in higher education, especially in undergraduate thesis supervision.

There are three aspects of the overall problem addressed in this dissertation; a practical problem in respect of too many uncompleted theses, or not completed on time. Another practical problem is increasing of the demand that is not accompanied by increased resources on a comparable scale. The third aspect of the problem is lack of knowledge in understanding the factors affecting the thesis process resulting in dropouts or delays. The lack of knowledge may result in a blind management process.

Two research questions investigate the problem, what factors influence the thesis completion or non-completion that can be useful for management, secondly, what management principles can be helpful to secure high quality in higher education.

The questions are answered through several studies that employ mixed-methods and learning analytics techniques which include data mining and the application of statistics and machine learning algorithms. In total, over 3000 thesis projects have been studied.

The thesis contributes with a better understanding of the factors that influence completion and non-completion of thesis projects, management guidelines for the thesis process, and a novel methodological approach using learning analytics and machine learning to support data-driven decision-making about thesis processes.