Evaluation Criteria for Master Thesis in Mathematics

Experience shows that there is not always perfect clarity and agreement among the people involved (student, daily and official supervisors, readers) what are the criteria for evaluation a master thesis and its sub-assignments. To some extent, this situation is deliberate: due to the large diversity in master thesis topics available to mathematics students, it is hard to regulate the evaluation entirely. While each person involved in the evaluation of the thesis and its sub-assignments makes their own assessment, this page attempts to provide — based on experience — an overview of criteria that have been applied in the past, specifically for thesis subjects in type theory. This page is not an official document, and does not regulate the evaluation process.

Regulations and timeline for the master thesis are available here. Further information from the Faculty of Science is available here (EN) and here (Dutch only).

The evaluation consists of the following sub-assignments:

• Intermediate report & presentation
• Seminar
• Final dissertation
• Defense

The target audience for each of these texts & presentations are fellow students of the master of mathematics who are not necessarily working on a related topic. Material covered in relatively mainstream courses in the master of mathematics can be considered prior knowledge, type theory generally cannot.

There are three presentations involved. The intermediate presentation and final defense are shorter, making it impossible to discuss all the technicalities in a way that the audience can follow. Hence, these presentations focus on the bigger picture. The seminar is longer and should therefore also cover some degree of technical detail.

• The evaluation roster from the Faculty of Science (NL - EN) is not 100% applicable to thesis subjects in mathematics / theoretical computer science, nor to each of the sub-assignments, but it is a useful tool for both student and assessors to keep in mind what aspects are important.
• The evaluation scale from the Faculty of Science (NL) and the Faculty of Engineering Science (NL - EN) are useful calibration tools.
• In general, the following parts are best treated in each of the sub-assignments:
  • Cite your sources! Failure to do so can have more serious consequences than merely a low grade. A weird citation is always better than a missing one. See also here.
    • In case you consider using generative AI (like ChatGPT), please consult the university-wide guidelines on this topic (NL - EN).
  • Have a clear research question or goal and discuss its motivation.
  • Make a clear distinction between what exists and what is your contribution.
  • Make a clear distinction between what has been done and what still needs to be done.
  • Discuss what has been done. This includes:
    • Familiarizing yourself with the topic (note that the report would not reflect the process of familiarizing yourself with the topic, but rather an introduction to the relevant technical content),
    • Literature study,
    • Actual original research.
  • Be clear about the results of your work. Be transparent about its flaws and virtues. Providing a good understanding of why something does not work or is not so great, is a form of scientific success. Even if, by the end of your thesis, your research question remains completely unanswered or is answered negatively, you can document why this is the case despite having made justified research choices and you still have a thesis.
  • Discuss what are the future plans within your thesis, and (towards the end of the year) what are interesting directions for follow-up research.
  • Discuss related work. The existing corpus of scientific knowledge is almost infinite; discussing related work is a matter of making a reasonable effort, taking into account the length of the text / duration of the presentation and the nature of the work you’ve done so far.
  • Discuss how your work relates to related work; identify the gap in the state-of-the-art that you’re trying to fill.

This document is a best effort in providing insight to how master theses are graded. It is not set in stone and is subject to future changes based on further experiences or feedback from colleagues.

A version history of the html source of this file is available here.