This post is from Roman Haas. It focuses on more or less typical problems that appeared to him and his friends during our theses and how to avoid them. They are described by anti-patterns, i.e. there is always a description of a problem and a possible solution for it. The problems are sorted by the moment when you should pay attention to them: the first ones may appear at the beginning of your work, the latter ones appear during your work or at the end.
Rating Criteria Surprise
Problem: You are not happy with the final grade of your thesis because you spent too much time on things that are not (so) relevant for the grade.
Solution: Ask your supervisor at the beginning of the work which criteria will be used to rate the thesis. Does the code that you wrote for the thesis influence the grade? What about the presentation at the end?
Wasting Time on Thesis Template
Problem: At the end of your writing time you realize that there are formal errors like formatting issues and content of the cover page in your thesis.
Solution: Look for a thesis template at the very beginning of your work, e.g. ask your friends which template they used. Do also check at the beginning whether all formal criteria are fulfilled by the template (look on the thesis webpage of the faculty, e.g. here for TUM and the examination regulations, which can be found for CS at TUM here).
Drafting Prose without Outlining First
Problem: Your text is poorly structured and while you write your thesis you do not really know which chapters your thesis will consist of.
Solution: Have early drafts of outlines: One that gives you an overview about the contents of your thesis. Do also work with outlines for the single chapters! Use a template for your document and fill it with subsections that fit your topic. Get feedback from your supervisor regularly! It is absolutely common that you restructure and refine your thesis. See also the posts on how to Spend your Writing Time well and Thesis Architecture.
Problem: You try to find a solution on your given problem before having an overview about existing solutions. After implementing your own solution you find out that several better solutions already exist.
Solution: Get an overview about your topic and related work at the beginning of your thesis, i.e. before you start implementing your own solution. Have a look at the literature search post in this blog.
Citation Style Unknown
Problem: You do not really know how to cite in your thesis. You have questions like “Are direct citations needed?” or “Should I use foot notes?”. Your friends from other universities use a completely different citation style and you are wondering if you should use their style.
Solution: Unfortunately, there is no definitive citation style that is used for scientific work. There are many different styles in use – even for computer science. The most important thing is that your supervisor accepts your citation style, so ask there firsts. Furthermore, stuck to one single style throughout your thesis. To find the style you would like to use take a look at the papers that you read. The TUM university library also provides a citation guideline that presents some styles that are commonly used in scientific work in computer science.
Problem: You remember an important statement from some paper you have read two months ago and would like to reference it in your thesis. You just have no clue, which paper that was.
Solution: Write short summaries of important papers directly after you have read them. Point out the results that are most important for your own thesis.
Problem: You cannot submit your thesis in time because the Infopoint (where CS students at TUM have to hand in their thesis) has already closed.
Solution: Inform yourself in advance, when and where you have to submit your thesis. Pay attention to the opening hours! There are good news for people that would have to submit on a Saturday/ Sunday/ holiday: You do not have to hand in early. It is sufficient to hand in your thesis the next day when the Infopoint is opened.
Did you encounter further pitfalls during your thesis? Please share your experience in the comments!