At the start ... How to write a good (no, great) PhD dissertation
- Papers appearing in the top conferences (not necessarily journals) in your field over the past 2-3 years
- Look at the best papers in those conferences
- Look for taste in research, taste in presentation style, amount of work that it takes to have a best-paper award
- Theses of the ACM Doctoral Dissertation Award winners and honorable mentions over the last 2-3 years
- Your advisor’s thesis and recent university theses
- Helps you to understand how much “work” your advisor will expect
- Ask your advisor what he/she is proudest of and what he/she would do differently, if given a chance to re-write the dissertation
Defining the problem
- Paragraph 1: What is the problem?
- Not more than 3-4 sentences telling the reader what the problem is, in as simple English as possible
- Paragraph 2: Why is the problem hard?
- What has eluded us in solving it?
- What does the literature say about this problem?
- What are the obstacles/challenges? Why is it non-trivial?
- Paragraph 3: What is your approach/result to solving this problem?
- How come you solved it?
- Think of this as your “startling” or “sit up and take notice” claims that your thesis will plan to prove/demonstrate
- Paragraph 4: What is the consequence of your approach?
- So, now that you’ve made me sit up and take notice, what is the impact?
- What does your approach/result enable?
Writing up ... How to Write a Thesis: A Working Guide
- Hartmann's hints
- Title. The title should be succinct, focused and objective, giving, if possible, the scope of the thesis.
- Abstract or Summary. Examiners will look here to find out whether it is new knowledge; and if so what.
- Introduction. Remember that the introductory pages are important because they create the first, and perhaps lasting, impression on the examiner. Use flow diagrams, headings, sub-headings etc., to create and sustain interest.
- Literature Review. This should be a critical synthesis of the state of the knowledge. Especially important are the areas needing further investigation: what has not been done, as well as what has been done, but for which there is a conflict in the literature. The examiner finds out how the candidate thinks from reading this section.
- Experimental Chapters. The hypotheses must be framed carefully and experiments designed thoughtfully
to test it.
- General Discussion or Conclusions. You may afford to be speculative here.
- Examiners ask the following questions when reading a thesis:
- Has the student read all the references?
- What questions does this thesis raise?
- What richness does it contain that can spawn other work?
- What is the quality of flow of ideas?
- Keep in mind that examiners read a thesis in instalments and display a natural benevolence, i.e., they do not set out to read a thesis with the aim of failing the student.
- Read the whole thesis to pick up repetition.
- Read your thesis for ideas and read it again for editing.
- A common structure used for experimental chapters
- Introduction/Aim - What did you do and why?
- Materials and Methods - How did you do it?
- What did you find?
- Analysis of the results
- Discussion - What do your results mean to you and why?
- Comment on the results - what are they?
- What meaning can you rest from them?
- Are they in accord with accepted theory?
- Do the results uphold your assumptions?
- How do you treat unexpected or inconsistent results?
- Can you account for them?
- Do your results suggest that you need to revise your experiments or repeat them?
- Do they indicate a revised hypothesis?
- What are the limitations in your methodology?
- How do your results fit in with the work of others in the field?
- What additional work can you suggest?
- Conclusions - What new knowledge have you extracted from your experiment?
- A thesis must tell a story clearly and convincingly.
- The structure impart logical continuity to the thesis in much the same way that links in a chain confer on it integrity and strength.
- The hypothesis is all important. It is the foundation of your thesis.
- Some different explanations:
- The hypothesis defines the aim or objective of an experiment, that if some likely but unproven proposition were indeed true, we would expect to make certain observations or measurements.
- A hypothesis is an imaginative preconception of what might be true in the form of a declaration with verifiable deductive consequences.
- Hypotheses are the larval forms of theories.
- ‘In every useful experiment, there must be some point in view, some anticipation of a principle to be established or rejected’; such anticipations are hypotheses.
- Key features:
- Fits known facts (know the literature).
- Is testable (done the experiments).
- Example from engineering.
- Engineers invent rather than discover, but they discover knowledge along the way.
- ANN + hardware to sort good apples form bad.
- what is good? what is bad? what is accuracy?
- discover works well with green apples, but not red ones.
- discovered new knowledge.
- suggest revised hypothesis as a starting point for new research.
- have demonstrated the hypothesis ‘It is possible to sort good green apples from bad green apples, with an accuracy of better than 90%, using ANNs and suitable hardware’.
Writing a (Computer Science) Paper
- Motivation – a real issue?
- What is the research context?
- What is the state-of-art?
- Hypothesis / Problem
- What is broken/missing (the ”grab")
- What are the research question? (more abstract than what you have done)
- Thesis or Problem statement
- Goals and methods
- What are the operational goals of this paper?
- And how were they achieved?
- Paper overview
- Outline of the rest of the paper
- Main body
- Section organization reflects how your argument unfolds
- Each section should have a main point
- Each paragraph should have a main point
- Tell them what you’ve told them
- some people only read abstract, intro and conclusions
- Relate back to general area
- Introduce future work
- Types of venue
- Workshops are great for initial ideas (hard to attend from NZ)
- Conferences for work showing that have done something, have results and can reflect upon it
- Journals don't cost anything to attend (good) but require more than conferences
- Industry events are great for getting feedback and promoting yourself
- Choosing a venue
- Applying for funding to attend
- Responding to rejection