Guidelines for students doing M.Sc. projects and third-year final projects
- General guidelines:
Prepare yourself to take notes; this should not have to be stated, but unfortunately there have been a lot of students in my experience that has made the mistake of coming to appointments without pen and paper.
Read the course literature. The course literature is excellent and brings up important issues and guidelines that improves your chances of a successful project. By reading the course book, the actual time spent in meeting is used more efficiently. In my experience, student who read the course book often spend their time more efficiently than students who has not read it.
Be prepared. Consider questions to bring to the appointsments, submit material for review as soon as possible. Again, this improves your chanses by making more efficient use of the supervision.
- Writing guidelines
- Personally, I am a believer in the idea that "To think is to write". The following are advantages of write as soon as possible:
- You can review your own material, a process that can improve your ideas, starting new threads of reasoning etc.
- Your supervisor and peers can give meaningful feedback on your ideas. The earlier the writing process starts, the earlier it is possible to provide meaningfull feedback.
- If you have any problems with writing, then the chanses that these are discovered and managed early are improved.
- How to start writing early:
- Review the articles that are likely to be background material in your report. A review should contain the following parts: (i) the purpose of the work presented in the article, (ii) the motivation of the work, (iii) the objectives of the work, (iv) contributions of the work, (v) future work, (vi) important concepts, and (vii) interesting references that the work addresses. Part (i)-(iv) is usually part of any peer review, while part (v)-(vii) is specific for the project course. In particular, part (v) is significant since this address open problems that can serve as the basis for your problem definition.
The text in the reviews can be copied into your report and then refined. Further, you can get feedback on the reviews.
- As soon as you have an idea of a problem definition, start constructing a layout.
- Scheduling guidelines: (To be continued):
- Create a page budget of your report. This might sound crazy, but my experience is that it is an important tool of getting the material in your report to the proper place. My recommendation is that given that you should produce X pages in a typical report, then X/3 pages should be spent on the problem definition, the method (research method such as analysis, simulation, experimentation, survey, case study as well as solutions, methods, techniques, architecture, algorithms, implementation issues), and the results (including application of the method on the problem, discussion, related work).
The introduction and the summary can be viewed as extended abstracts of the problem, method, and results; where the introduction can be written using examples, incomplete definitions etc. to introduce a reader to your work, whereas the summary is written as if the reader has read the rest of you work. In my view, the summary is part of the conclusions with contributions and future work as additional sections. Since the introduction and conclusions contain an overview of the whole report, it is my recommendation to save this as long as possible.
The page budget should be used as a sanity check rather than rules for how many pages that each section should contain. In a typical report, the problem section is usually larger than the results section, but the size should not be unproportional. - To bring substance to you report, you need to reference others work. From my experience of typical third year final projects, the report reference 15-30 sources (journal papers, conference papers, work shop papers, technical reports, web pages as well as books). To find these references, it is often necessary to cover through 3-7 times the number of sources used in an actual report.
To plan the survey in the beginning of your project, you can try the following method. First, you should browse the sources to check if they are relevant. From the seeming relevant sources, you should read and preferrably review them to discover new sources. The new sources must be browsed in order to see if they are relevant. This process is performed until you have covered an area in a sufficient manner and are able to define the problem.
To plan your work, an estimation I use is that browsing an article takes about half an hour, whereas reviewing it may take 1-2 hours. So, given these estimates, you may have to browse 200 sources spending approximately 100 hours and reviewing approximately 30 of them for another 60 hours.
Given the schedule, you have approximately 6 weeks for the problem definition. That is, approximately 240 hours to spend on the problem definition. Given the aforementioned review, then you have approximately 100 hours for writing your report. - My advice is to plan the first version of the report in two-three major revisions, where each revision is ended by a review (at least once by you supervisor, preferrably more). Assume that you spend two iterations, then you have approximately 50 hours per iteration on writing (1.1 weeks each).
Two reflections on planning. First, if you have less that two weeks left and have not started writing, this is an indication that you have problems. This problem grow the closer you get to the deadline. Second, to handle problems in writing, in reading and reviewing sources, in defining the problem, you need to start as early as possible with these tasks to enable your supervisor to help you out.