Dissertation Proposal Guidelines
The purpose of the dissertation proposal and examination is two-fold: first, to establish that the student is qualified to embark on PhD-level research in the chosen field; second, to make sure that the student has a good plan for finishing the PhD, and that the committee (especially the advisor) and the student agree on the scope of the future research.
The proposal should include the following sections:
- Background knowledge
The problem statement should be comprehensible by faculty in the department who are not in the particular specialty of the student and the advisor. The thesis proposal should provide ample references to existing literature in the field and should provide a clear explanation of the problems to be worked on. Typically, the background description for a thesis proposal takes 5-10 pages, with fields with a rich literature requiring more than newer fields. Note that this is shorter than the "prior work" chapter of a finished PhD thesis, but can serve as the nucleus for such a chapter. - Evidence of Preparation
The proposal should explain how the courses taken to fulfill the degree requirements have prepared the student for research on the selected topic. The student's transcript will be available to the committee at the qualifying exam. In some cases, the student will have already worked on some of the problems and have solved some of them. This evidence of ability to work in the field is extremely persuasive, but is not a substitute for citations to the literature. Results should only be summarized in this section; publications or technical reports may be attached as appendices. - Research Plan
The research plan should include detailed descriptions of what the student plans to do (theorems to formulate and prove, systems to build, experiments to perform, and so forth). There should be specific milestones and a time line for when the student expects to meet each milestone. The research plan should schedule writing up of results along the way, indicating possible conferences or journals to which the work should be submitted. The research plan should describe not only the problems to be attacked, but the approaches to be used in attacking them. The details do not need to be fully worked out (indeed, they shouldn't be at this stage), but there needs to be enough information for the committee to estimate the likelihood that the approach will work and to estimate the difficulty of completing the thesis. The research plan should be 5 to 20 pages long.
Drafted by Kevin Karplus and Martine Schlag, March 2001