Transkills Course
How to do an Informatics PhD

Pre-Course Reading

For the purposes of this course on Doing Informatics Research, it will be assumed that you have a research project, which you will be expected to analyse and describe. Normally, you would use your MSc by research, MPhil or PhD project. If, for some reason, this is inappropriate, e.g. you have not yet finalised your project, then please make up a suitable project. For instance, use someone else's projects that you have learnt about in your reading of your chosen field.

Now answer each of the following questions for your research project. When you have done this, tackle the exercise at the end. It is essential to the practical exercises in the course that you both do this exercise and bring the result with you.

• What are your aims and objectives? What do you hope to have achieved when you are finished, e.g. build a program to do some task, prove a theorem, model a natural system?
• What is the motivation of your project? That is, why do you think it will make a significant and original contribution to the scientific and/or engineering progress of Informatics? Why is it timely to tackle this project now and do you think it is feasible to achieve in the timescale available to you? Who would benefit from a solution to the problem you have set yourself?
• What are the main pieces of related work? Have other people tried to solve the same problem using a different solution? Has a similar solution to yours been applied to different problems? What is the state of the art in this field? What are the limitations of previous work on this problem?
• What is your novel idea? What do you bring to this project that is new? Why do you think you can extend the state of the art? Can you solve any previously unsolved problems or overcome limitations of previous work? Why might you succeed where others have failed?
• What are your claims or hypotheses? For instance, if you plan to build a computer program, in what way will this program be better than each of the rival programs for solving the same or similar tasks? Will it: exhibit better quality behaviour, be more efficient, apply to a wider range of problems, be more dependable, be easier to maintain, more accurately model some natural system, ...?
• What kind of evidence will be needed to support these claims or hypotheses? Is your evidence experimental or theoretical? Is it amenable to statistical analysis? Will you need to experiment using humans or other animals? Will you require a large corpus of test data? How will your ensure that your test data is representative? What will provide the "gold standard" by which you judge the correctness and/or quality of your project's results?
• At what stage is your project? Are you just formulating your project proposal, is the work in progress or is it finished or nearly finished? What further work do you still have to do or have you identified for others to do after you?

Essential Pre-Course Exercise

In particular you should address two of the above questions with especial care, namely:

• What are your claims or hypotheses?
• What kind of evidence will be needed to support these claims or hypotheses?

During the course, one of your fellow students will read your summary and then give you feedback on it, especially with regard to the above two questions. .