Skip to main content

Research

The Department has a lively research culture, with research groups in many different subject areas. See the research section of the department web site, including the specific page for research students.

What is a PhD

The aim of doing a PhD degree is to become an independent researcher, capable of generating, pursuing and communicating novel research ideas. The PhD degree is at a higher level than Bachelors and Masters degrees, giving you the title of Doctor, and it can open up many academic and industrial career possibilities.

Doing a PhD primarily involves carrying out an original piece of research, under the guidance of an advisor, and writing it up in a dissertation, which is then examined by a viva. A PhD degree normally lasts for three years. It can sometimes last longer, but you have to beware of running out of funding, exceeding University limits, giving the department a bad reputation, or damaging your chances of a successful outcome.

PhD students become competent communicators, by taking a full part in discussions with their advisors and research groups, by producing written reports, and by giving presentations. They also take part in the wider research community associated with their subject area, by taking part in long-distance discussions, publishing papers, and going to conferences.

More formal information can be obtained from the engineering faculty postgraduate handbook and from the university postgraduate regulations.

Applying for a PhD Place

There is information about applying for a PhD place on the PhD admissions page. To be accepted onto a PhD degree, you typically need

Choosing a Subject Area and Advisor

Prospective PhD students should make an effort to become knowledgeable about the research activities in the Department, as a well-informed student can more successfully choose an advisor and an area of specialisation. The following points should be kept in mind when seeking an advisor:

Please see the department webpages for a list of example PhD projects by research group and supervisor.

Funding

There are many different sources of funding. For example, you can fund yourself, or you can be funded by a grant from your country if you are coming from abroad, a general grant held by a particular research group, a grant from a company, scholarships, and so on. It is also possible to have a mixture of sources of funding.

The Department of Computer Science recognises that most students require financial assistance to remain at the University. Assistance to fund you or to supplement your funding is available in the following forms.

What is Expected of a PhD Student

Your main overall task is to develop your own research skills in a self-motivated way. The key skills are to be able to solve a problem that has never been solved before, to identify which new problems are interesting enough to be worth solving, and to describe a problem and its solution clearly and convincingly, both in speech and in writing. At the same time, you will be adding depth and breadth to your computer science knowledge and taking part in the general life of the department.

Any initial project or area of the literature which advisors give them is likely to be only a start. It is rare for an advisor to give you a problem whose solution will form your whole PhD project. In any case, that goes against the idea that a PhD is supposed to be original work. Whatever you are given as a starter will have to be expanded into a project of your own, or you may end up in a different topic altogether.

Tasks

Your first job is usually a literature search in your chosen area. You must be thoroughly conversant with the literature, so that you know what has been done already, and what sort of new ideas would be of interest to the community. Quite soon, you should expect to become the literature expert rather than your advisor, at least in your own narrow area.

Next, you will probably want to investigate any initial problem suggested by your advisor, or find a suitable initial problem for yourself. This may or may not expand into your thesis topic.

The most difficult task you have, which you should never lose sight of, but which may well take a couple of years in all, is to decide on a thesis topic and bring it into sharp focus. Where will this thesis topic come from? It may emerge from immersion in the literature, tackling initial problems, suggestions made by your advisor or group, discussions on the net or at conferences, web searches, or just a flash of inspiration.

Use a weblog or whatever suits you to take notes of everything you do. Write down annotated references to things you read, ideas you have or find out about, experiments you do, installation or configuration sequences you have to go through, records of tests or demos, and failures and blind alleys as well as things that worked. A note can be just a date and a single sentence, but you must do it at the time, because you can't recreate lost information later.

Back everything up, including code you write, notes you make, the text of your dissertation. If you work in the department or can arrange direct or broadband network access, you can keep your work on one of the departmental servers which has regular tape backups. Otherwise, you can arrange for work to be regularly copied onto a departmental server.

Making Progress

There is a particular difficulty in measuring your own progress, because research is naturally open ended. You have an advisor, and there is a progress monitoring system which is described below, but you still need to set your own goals and milestones. Advisors can help, but you need to chase them, because they are too busy to chase you.

One problem in measuring your work is common of many PhD students. You will probably feel after completing a piece of work that it is trivial, after all you can now understand it. You need to bear in mind how hard you thought the problem would be when you started on the work. This is a symptom of gradually becoming an expert in a field, you start to feel that things are easy when actually they are hard.

A PhD requires full time dedication and year-round effort, just like your advisor shows. Good work habits, organisation and time management are also important. Although the academic life offers considerable freedom, you should still aim for working hours which are regular, not too short (say 40 hours per week minimum), not too long, include the main part of the day when other people are around, and include a weekend break for refreshment. You should expect to have to put in extra evening and weekend hours from time to time in emergencies, but not too often. If you do take time off, make sure people know about it and can contact you if appropriate. In the end, though, it is results which count, and as long as you perform well, you have a lot of flexibility in these matters.

It is important to avoid distractions which eat up too much of your time. Common problems are getting involved in too much teaching or other peripheral activities, trying to tackle too big a problem or too many problems, getting bogged down in programming, doing too much throw-away experimentation, or giving way to perfectionism. The best approach is usually to tackle a small problem, finish it, publish it, and get feedback on it which helps shape your future direction.

You may need grit to get through your PhD, because it is the biggest individual project you will ever have tackled. It is common to get the mid-way blues because you have nothing tangible with which to prove the amount of progress you have undoubtedly made, and the final form of your thesis may not yet have come into focus. You have to be resigned to the fact that research can be like that, because if you knew where you were going, it wouldn't be research. This is where feedback from other people and the monitoring process become really important.

What happens if you start to overrun, and the end of your funding is looming? You don't want your PhD to drag on because motivation may wane, it gives the department and the university a bad reputation, and you risk having to face some difficult questions about it in your viva. You need to face it and talk about, not let it drift, because there are several solutions. First, and usually best, there may be a way of wrapping up what you have already done or narrowing your sights to get a swift conclusion. If the research really needs more time, it may still be best to get your PhD out of the way and then get involved in a research proposal to continue the work at leisure. Alternatively, there may be ways of finding a little extra money or getting sponsorship to tide you over, or you may prefer to suspend your studies while you earn some more serious money and then finish off later.

Professional Development

A PhD student rapidly develops into a colleague. You need to be highly self motivated, and have high standards. You can take on as much responsibility for your own progress as you want. In the end, you decide what research topics to pursue, what papers to publish, which conferences to go to, when to submit, and so on. However, you obviously need to cooperate with the university while using university resources, with your funding organisation while being paid, with your advisor while getting their support, and with your research group while being involved in it.

You need to learn to take criticism. This will come from your advisor, research group group, reviewers of papers (whether the papers are accepted or not), hecklers at conferences, and so on. Open and direct criticism is absolutely vital to any discipline calling itself a science, and is the only way to keep standards high and make sure that the subject moves forward. You need to be aware that this criticism is never intended to be personal. Even "what you have done is rubbish" is very different from "you are rubbish", and usually just means "your work is irrelevant to my work". The purpose of criticism is to test ideas, to improve the quality of the work or its presentation, to make sure that publications which reach a wide audience take into account as many different points of view as possible, and to point out where the research could go next. Accepting, considering and learning from criticism, without rejecting it or getting into an argument about it, is a key skill.

You need to publish. Without publications, the field you are in does not progress, and the work you have done is wasted. Your PhD dissertation itself does not not usually count as a publication, so you should expect to publish succinct summaries of the best work from it elsewhere. Publications are also the most significant indicator of personal success, and it is one the issues you will be asked about in your viva.

While selling yourself and your work, you also need a little humility. They say that "the more you know, the more you know you don't know". With any tough research problem, two heads are better than one, and three heads are better than two. Get used to showing your ignorance by asking lots of polite and succinct questions at every opportunity, and looking for opportunities for cooperation.

What is Expected of a PhD Advisor

A PhD advisor is often called a supervisor, but that word is used less often now because it suggests someone who tells a student what to do, which has the wrong emphasis. The main task of an advisor is to give advice.

Advisors help students to sort out practical issues to do with money, space, equipment, accounts, software, tools, materials, travel and so on, if only by telling students who to go to. They also help with outlining or writing papers, suggesting when and where to publish, suggesting what conferences or workshops to go to, planning and preparing the thesis, preparing for the viva, and giving references or other help towards a student's future career.

An advisor helps a student to find a thesis topic by discussing the overview of the general area, giving a set of initial references from which to start a literature search, suggesting a problem or two which may help get a student's investigations going, and continually discussing possibilities.

Advisors try to be available for meetings and discussions whenever needed, tell students when they are going to be away, and arrange for someone to take their place if they are going to be away for a long time. They offer encouragement when things go badly, and praise when things go well. They help with planning and time management to make sure individual students stay on track, and they take part in the general monitoring activities which are designed to make sure all students make good progress.

An advisor criticises what the student does, particularly written work and presentations. The criticism should not be personal, but constructive, and prepares a student for the shock of their first criticisms from the wider community, particularly from reviewers of the student's first paper.

An advisor helps someone to move from being a PhD student towards being a colleague, by sharing experience, doing joint research, and developing mutual respect in one another's expertise.

PhD Progress

You should be able to :

During the course of your PhD you will have to submit reports to your advisor. These reports are used for the PhD monitoring process, are a record for you and the advisor, and will, together with papers that you published, form the basis for your PhD submission. You may have to give presentations at local seminars, or presentations to the research group.

At the end of your PhD you will write up your dissertation. Ultimately, it is your decision what you write in your thesis, and when to submit it, but your advisor will give you feedback on your thesis, and will tell you when your thesis has the right structure and contents to submit.

These are rough guidelines on the major milestones, and when they should be accomplished, for a typical PhD student. In general, the sooner a milestone is completed, the better.

Twice a year, you will be subject to a PhD review. The review has the following purposes:

Every PhD student is subject to the half-yearly review even if you have just started. The only exceptions are if you have submitted a full draft of your thesis to your advisor, or if your studies are suspended.

Reviews take place every Monday at 13:00, two or three speakers per day.

Your report should consist of a list of bullet points stating your achievements since the previous review (up to half a page), a plan for the coming 6 months (up to half a page), a list of problems encountered, a list of seminars attended, and the status of (planned) publications. Under the last point you should list technical reports, literature surveys or papers that you have written, and, if applicable, which parts of your thesis are completed.

When you have completed the first year of your PhD, you should write a mini thesis and submit that with the two pages above. This thesis should contain a detailed research plan (around 10,000 words), including a literature survey and showing a comprehensive understanding of the field. If you need to transfer from an MSc to a PhD, then this report will be used as the transfer report. This report should convince the reviewer that you will be able to complete your PhD in three years.

If you have completed 30 months of your PhD, a detailed outline of the dissertation is required, together with a longer account of the technical results obtained so far and a timetable for completion. Individual chapters of the dissertation or papers intended for publication may be submitted as well.

Feedback from the review will be provided to you by an appointed reviewer. It is intended that this feedback consists of constructive criticism of your work and suggestions for improvements or changes as are considered appropriate. A student may be also asked to come to a meeting with the Head of the research group to discuss progress.

Viva

When you have finished writing your dissertation you will hand it in to the university, and you and your supervisor will have to find appropriate internal and external examiners. The internal and external will decide with you on a date for your viva. During your viva, you will have to demonstrate that you know the subject area in general, the literature, your work, and be able to defend your thesis. After your viva, the external and internal will meet and let you know whether they were happy with your performance.

Postdoctoral Study

After your PhD you could become a postdoc. This means being a research assistant or similar on a research project. Typically, this means being on a fixed-term contract, so it is usually seen as a temporary post. However, some people seem to manage to be postdocs indefinitely, lurching precariously from one short-term contract to another, or relying on unreliable pieces of funding. This is probably not a good long term career decision, so why would anyone want to become a postdoc?

One reason is that if your goal is to become a full-time academic, e.g. as a lecturer, this gives you an opportunity to spend some time doing mainly research, and relatively little in the way of teaching or admin. It is a time to improve your publication record, and to develop experience which you can draw on later.

If you have other goals in life, a postdoc offers an opportunity to get into the habit of life-long learning. You can learn new ways of doing things, new tools, new ways of thinking, new ways of running things, new ways of being independent. It may also be an opportunity to build relationships for the future. It gives you considerable freedom to develop in in the way you want.

As a postdoc you must contribute to the research of the group, by doing good relevant work and publishing it, and there should be a "plausible relationship" between the work you do and the source of your funding.

The Practice of Research

Authorship and Acknowledgement

Who should be listed as authors on a paper, and in what order? This varies, but there are several guiding principles.

Having said all that, it is never wrong to discuss the authorship issue with potential coauthors.

In some subjects, overlapping research is dealt with by being secretive about your research and then trying to make sure that you publish first. In Computer Science the prevailing attitude is that overlapping research is dealt with by everybody being completely open about what they are working on, so that where overlap occurs, cooperation can develop. Be open about what you and your group is doing. Display everything on the Web. CS may be the only subject where almost the entire range of research effort can be followed on the Web.

Ownership of Research Results

During your research you will develop results that may have commercial value. The university is involved in the exploitation of research, and involved in joint industrial work. If you identify opportunities for commercial exploitation of your work, you must first discuss those with your advisor or the Head of Department. The results from EPSRC funded research are usually owned by the university, whereas results from work sponsored by industry are subject to the particular contract. Results from research include:

Publications

Publications of members of the Department are maintained in an electronic archive on http://www.cs.bris.ac.uk/Publications/. You are expected to add all your publications to this archive. Publications in the archive are included in the Research Report of the Department, and in the University Research Directory.

Work that has not been published externally can be published as a Technical Report. Technical reports are submitted and archived electronically (using the URL above). A standard cover page and a report number are automatically generated. A technical report is recognised as a disclosure in that it carries a publication date, showing that the work was performed at that date. Material that can be issued as a technical report includes: literature reviews and progress reports of PhD students, deliverables for externally funded projects, accepted papers that will not appear in print for some time, and in general any reasonably polished text that you want to refer to in other papers. A technical report is not a proper publication, and as such it does not limit you in publishing the work in a conference or journal later on.

Other resources

A really useful guide to research is "Working with Norman Ramsey, A Guide for Research Students" by Norman Ramsey, Harvard University 2002.