Search:
View  Edit  Attributes  History  Attach  Print  Search
Menu Main / Going Overseas to do a Postgraduate Degree in Computer Science: A Guide for South African Students

HomePage

Forum

Researchers

Going Overseas for a Postgraduate Degree

Studying in South Africa for a Postgraduate Degree

The 10 Simple Rules Series


WikiSandbox

PmWiki

edit SideBar

Going Overseas to do a Postgraduate Degree in Computer Science

A Guide for South African Students

Contents

  1. Introduction
  2. Why Study Overseas?
  3. Which Degree and Where?
  4. Getting Funding
  5. Getting In
  6. Making a Decision
  7. Responsible Representation

Introduction

This document was originally written by George Konidaris in roughly 2006, and is intended to be, as the title would suggest, a guide for South African students who are interested in getting a postgraduate degree in Computer Science overseas. It is now a community resource, reflecting the collective experiences of several people.

We decided to set up this guide as a way to save that collected experience, because the two most important hurdles you have to overcome before you decide to get a postgraduate degree overseas are fully realising that it's not only possible but actually very feasible, and then understanding how to go about doing it. This information is usually obtained by word of mouth, which immediately leads to the obvious problem that people who have recently decided to study overseas are by now usually very far away. Every time one of us goes back home, we make a point of visiting our home university, and talking to the current batch of Honours students, but that isn't really enough.

To make a decision like this, you need to know that it's an option early, and you need to get a good grasp of the shape of that option. That's what this document is for. When many of us decided to study overseas, it took us a very long time to collectively gather some of this. We've written this guide so that you can get a very good start in one place.

It's worth mentioning that much of the information here will port to other disciplines, particularly the other Mathematical Sciences and the Engineering disciplines, and maybe even further. We can't vouch for that, but most of the general advice and all of the scholarships we've listed are not specific to Computer Science.

A note: the material on this page, particularly on scholarships, ages very fast. If you find errors or think you can update the material here, or make it more useful, please do.

Why Study Overseas?

There are a couple of good reasons to study overseas.

You want to become an academic, in which case a PhD from a good overseas University will in all likelihood significantly boost your career. It will also benefit you in other ways - it will expand your view of academia, give you exposure to other methods of research, teaching, fundraising and administration, and generally broaden your horizons.

It's a career boost, even if you plan to go straight into industry. South African employers like the sound of a foreign degree, especially in something useful, and they also like the implied life experience. Having the extra experience and a broader knowledge of the subject is always good.

It's a fantastic personal experience. Living and studying overseas, meeting people from all over the world, and learning about the way that things work overseas (in industry, academia and life) is tremendously personally enriching.

You want to study for a little bit more (or you're not quite ready to start working) but you want a bit of an adventure. This is pretty much the same as the last point. :) Many people only start to really enjoy their studies when they start to really get into the heart of a subject, or when they start doing research. Some people who start out this way decide after their experience that they really do want to become academics (although obviously many of them decide that they really don't).

There are also one major downside. For however long you decide to study overseas for, you're going to be away from your family and your old friends. It's unlikely that you'll be able to afford to fly home more than once or perhaps twice a year. You should be aware that when you leave home and come back much later, everything may be completely different to the way it was when you left it, for all sorts of reasons.

Which Degree and Where?

The UK MSc degree is mostly a single year Master's degree, almost always half coursework and half research project (although some are coursework only). These degrees are intensive, and a good opportunity to get deeper into an area you're interested in. In the UK they're often used as conversion courses for students wishing to switch disciplines, since you can go directly into a PhD there with just an undergraduate degree. The UK also has another Master's level degree, called an MPhil, that is usually a two year research only degree similar to a mini-PhD. Sometimes this degree is conferred when a PhD student submits a thesis that isn't of sufficient quality or depth.

The British PhD (or DPhil at some Universities, notably Oxford and Sussex) is a three year research only degree, during which your only goal is to produce a thesis (although if you do not have an MSc you may be asked to do some courses to complete your background). You're usually given a year to come up with a research proposal, and then two years to complete it. You should be warned that many people don't finish in three years, and end up taking four or five. This happens for several reasons, not least of which is the freedom given to you during your degree - you are required to submit yearly progress reports, and that's about it, unless your supervisor feels like they need to keep a close watch in your progress. Your funding, however, will almost certainly only last three years (although most universities don't charge tuition fees for longer than that). Nevertheless, finishing in three years is definitely feasible if you put your mind to it, don't slack off, and maintain focus (anyone who thinks that is easy has never been to Edinburgh). The quality of the UK PhD varies; the best ones are as good as the best in the world, but some of the lower end are probably below what would be considered sufficient in the US.

In the US, the MS is usually a professional degree, and terminal in the sense that it does not lead to a PhD or get your credits towards a PhD unless it is awarded during your enrollment in a PhD programme. It's much easier to get into a very good MS programme than it is to get into a very good PhD programme, but it's harder to get funding for since the relevant School us unlikely to fund you.

The American PhD lasts for between four and six years, and usually consists of a breadth component and a depth component. The breadth component usually consists of coursework covering the primary areas of CS (often Systems, Theory and AI) and takes up about the first two years. It is usually terminated by a set of Qualifying Exams. Failing your quals gets you kicked out of the PhD programme (many offer a second, but never a third, chance), and passing them gets you into PhD candidacy. The coursework you do during this phase usually gets you an MS along the way to your PhD (although it sometimes requires a small research project that is considered your Master's thesis), and if you already have a recognised MS or MSc then you can often skip some of it. After candidacy you are required to write a thesis, which takes at least two years, but possibly up to four.

Canada is somewhere between the two systems. You are required to get an MSc before you go onto a PhD, and it is similar to the American MS in that it takes about two years and has substantial coursework and research components. It's usually easier to get funding from your target Department though, because it is required for a PhD. After an MSc you apply separately for a PhD, which requires a few more courses and is followed by a thesis, with the whole degree taking another three or four years.

An Australian PhD is one of the easier options to get into. You only need a good 4-year degree with a substantial research component, so a South African Honours or engineering degree is usually sufficient. However, international PhD tuition is very expensive and there are few scholarships. If you have a very good track record, you stand a small chance of winning a scholarship. A few well-funded research institutes have limited funding for international students but this is the exception. Australia has a deep hierarchy of universities so take care to select one with research strength, particularly in your area of interest. Completion time is nominally 3 years, but can stretch, though funding for more than 3 years is even more difficult. Australia is a better choice for postdocs: a postdoc position is reasonably well paid, and your job is to build your track record, not to be skivvy labour for academics. Australian Masters degrees are mostly conversion degrees, sometimes of low quality.

Most people proceed by deciding what kind of experience they want - if they're in it for the life experience, or want to do a Master's degree to help them decide whether to do a PhD, they mostly go for a UK or Canadian MSc at an interesting place and a University that matches their broad research interests and career goals. Students who want to do a PhD are generally more concerned with individual supervisors (advisors in the US) in terms of research match, reputation, personal interaction, ideas and personal dynamism.

It's worth noting that it's sometimes difficult to switch Universities between your MSc or MS and PhD, especially from the UK to the US (although not necessarily vice versa). Inevitably, you'll like one place and one University more than the other, and one academic culture more than the other (American academic culture is very different to the British tradition, which is what we're used to in SA).

Getting Funding

Note: you may wish to check out this list of funding opportunities for African students, courtesy of Shakir Mohamed.

The first thing you should do when looking for funding is go to the Financial Aid and Scholarships office at your University and ask for all the information they have about scholarships for studying overseas (at Wits this is called the International File, and most of it is online here). It usually has quite a few decent funding opportunities, and will tell you when your deadlines are. You should also go to the NRF and have a look at their list of international funding opportunities, as well as their fellowships for PhD study overseas. It's essential to plan well ahead - scholarship applications are often due more than a year before you would actually start your degree (so you should start planning at the beginning, not the end, of your Honours year). Also, it always pays to check out the funding opportunities available at the Universities you want to go to - some have their own special scholarships for international students, students from third world countries, etc., which you might be eligible for.

The funding options open to you depend heavily on where you decide to study, and which degree you decide to go for. What you're looking for is colloquially termed a "Full Ride", which means tuition plus a stipend that's sufficient to survive on. Most scholarships will also pay for your flights and some relocation expenses.

If you're going to the US for an MS, then you'll almost certainly have to get a scholarship, most likely a Fulbright. Since the MS is primarily a professional degree in the US, most schools expect applicants to find their own funding, or pay their own way.

If you're going for an American PhD, then you can also apply for a Fulbright, but most good schools will find funding for you if you get in anyway - either in the form of a Teaching Assistantship (TA), where you have to do things like tutor a class, grade tests and exams, etc., or a Research Assistantship (RA), where you're essentially paid for your contribution to a particular active grant. Some, but not many, Schools will also be able to offer you a Departmental Fellowship.

This leads many students to think that they don't need to worry about funding at all. However, you should be aware that the funding situation in the US isn't quite as wonderful as it sounds. Students with TAs, and sometimes those with Departmental Fellowships, often come under pressure to switch to RAships, so that the funding can be used to get more students. Students with TAs often find that they want to switch to an RA anyway, because they don't like teaching or because it takes up too much time. Students who have RAs are essentially being paid for their research skills, but they're almost always funded through research grants obtained by their advisors. These grants typically come from DARPA, the NSF, the NIH or NASA.

In such cases, you will be expected to work on the topic and the deliverables promised in the grant, and you may find that you have significantly less time than you thought to do your own research. This isn't a problem if you don't have a very strong and narrowly focused research interest, and in such cases it may actually help you find a research direction of your own, but if you have something very specific you're interested in you might find this frustrating (on the plus side, many advisors will be happy to help you write a grant covering your specific interests). You may also find that some funding agencies will impose constraints on your research, or that you find you're uncomfortable working for them (e.g., DARPA has recently displayed a tendency to see Universities as cheap contractors for building Killer Robots). For these reasons, I suggest that if you want to get a PhD in America you do your level best to get a Fulbright, because it will make these issues disappear.

If you decide to go to the UK for an MSc or an MPhil, then once again you're most likely on your own for funding. The most likely sources for Masters and PhD level funding are:

  • A Commonwealth Scholarship
  • A Chevening Scholarship (usually Masters level only)
  • A Rhodes Scholarship, if you want to go to Oxford
  • A Cambridge Trust Scholarship (including the Gates Foundation scholarships) if you want to go to Cambridge
  • A Dorothy Hodgkin Postgraduate Award, which is new but likely to be competitive (PhD only)
  • A Patrick and Margaret Flanagan Scholarship (administered by Wits or Rhodes), for women only. This can also be used in the US but applications for study at Oxford, Cambridge, Edinburgh, St Andrews, or Trinity College Dublin are preferred.
  • The Scottish International Scholarship Programme, which offers scholarships for permanent residents of certain countries (one of which is South Africa) who would like to do a one year Masters at a Scottish University.
  • The Scottish Informatics and Computer Science Alliance, which offers PhD studentships to a number of Scottish universities. These studentships provide support for 3.5 years, but involve some work with SICSA.

All of these scholarships offer you a Full Ride, except the Patrick and Margarent Flanagan which sometimes only funds two years of study. Some South African Universities also have their own scholarships available.

It's also possible (but slightly harder) to get your target University to fund you for a PhD. This usually happens through the combination of an Overseas Research Scholarship (which pays for the difference between overseas and EU (or home) fees, but can only be applied for through one UK University) and a University or Departmental Scholarship, which pays for home fees and a stipend. It is also occasionally possible to apply for an open Research Associate position at the Department in question (these are advertised at jobs.ac.uk). A Research Associate post is a paid position where you're required to work on a particular funded research grant. Since UK PhDs are so short, it's almost certain that your thesis will be on work you've done within the scope of the grant, so it's a good idea to only consider grants that cover research areas you're interested in.

If you want to go to Canada, then you can often get a Full Ride for an MSc or a PhD from the Department, according to something akin to the American system, but probably with less pressure. You can also apply for a Commonwealth Scholarship to go to Canada, although far fewer of these are granted than the UK version.

If you want to go to Australia, your options are much more limited. Check if your home university has any scholarships, and ask the university in Australia what they can offer. If your track record is very good, there's a small chance that they can help.

Getting In

There's an awful lot of information on the Internet about how to get into a good CS postgraduate school. Most of this information is about American schools, but also applies to Canadian Schools, although they are generally slightly less competitive.

The American PhD application process at a very good school works something like this:

  1. Everyone without reasonably good GRE scores gets cut.
  2. Everyone who didn't do very well in their previous degrees (this means a first or sometimes upper second for Honours, ditto for an MSc) gets cut.
  3. Everyone who doesn't have references that say that they're very good, that they have serious research potential, and that they're dedicated and interested, gets cut. References that say that you're good at coursework are not good enough; these schools are being asked to bet a large amount of money on your ability to do research, so they naturally prefer references that say that you have some research experience, or that you're at least likely be good at it.
  4. This process usually leaves about four times as many students as the School can actually take. What happens next depends on the Schools; some schools (not many, an example is MIT) give your application packets to potential advisors, and they pick the students they're interested in. For these schools, you want your Statement of Purpose to be neatly aligned with your potential advisors research interests. Some of these Schools (notably USC) go even further, and require some sort of student-advisor relationship to be established during the admissions process. At such a school you should very specifically target a particular advisor, read their papers, and perhaps contact them directly when you apply. Other Schools (e.g., Carnegie Mellon) try to work out which students have good research potential instead and leave the advisor-student matching process until later.

So if you're serious about getting into a top-ranked American PhD programme, then you need to:

  1. Do well at your Honours degree.
  2. Get good scores for your GRE. PRACTICE. At the very minimum use all the free practice material you get to familiarise yourself with the style of questioning. If you can afford it, get a study guide and do as many example questions as you can.
  3. Sometimes it helps to get an MSc, even if the programme website says you don't need one, and do as well as you can. You can even, if you like, do an MSc in the UK with the aim of getting into a good PhD programme in the US. An MSc helps because it gives you more research experience, a better chance to get very good references, and a strong shot at publishing something. Which brings me to ...
  4. If you can, publish something. Having just one decent publication will almost certainly lift your chances dramatically, because it's proof that you are capable of doing research, and it distinguishes you from the majority of applicants.
  5. Target particular advisors in advance, and write your Statement of Purpose with them in mind. Even in schools where potential advisors don't see it during admission, it's impressive to have thought about how and where you'll fit in, and it'll help you decide where you would be able to pursue your interests.

GradDecision has some very good advice on this whole process.

Obviously everything is much less competitive for Masters level courses, and at Universities with lower ranked Schools.

In the British system, if you have a good undergraduate degree, decent references and you can get funding then you shouldn't have any problem getting into to whichever school you like. The problem then shifts to how to impress funding agencies, which depends on the agency, but is usually centered around an impressive coursework record first, then solid references, then a good statement of purpose. Publications count, but not as much. If you need to get into a good Department in the UK and get funded, then you're in a competitive situation very much like that in the US, without of course having to do a GRE. Once again, just one good publication can make your application.

Making a Decision

After you've applied to a whole bunch of schools, and gotten into some fraction of them, you're faced with deciding where to spend the next few years of your life. For some people, this is an easy decision, and for others, it's an incredibly difficult one. Here are the factors you'll need to take into consideration.

Supervisor (or advisor) name and research match. Your supervisor is going to be the most important person in your academic life for the duration of your degree. Most students aim to find someone with a very similar research agenda to their own. It's also often a good idea to get a well known supervisor, someone who is recognised as good in the research community, because your name will be associated with theirs for quite some time. On the other hand, well known supervisors often have less time, less energy and are less enthusiastic about way out ideas than younger ones. Sometimes it's useful to land up in a Department with more than one potential supervisor. It's a good idea to contact potential supervisors by email once you've been accepted (this is often OK in the UK before you've been accepted). Your supervisor name and research match is much more important for PhDs than for MScs.

Department strength also obviously determines the academic value that you get out of your degree, and the amount of department name recognition that you'll get. This is very often field-specific; many strong Departments are in otherwise weak Universities. This is mostly important if you want to impress people within Computer Science.

School strength and name recognition is mostly important if you want people outside of Computer Science. If you want to get a good job after your degree, then having a degree from a famous University might count more than having a degree from a lesser known but technically better one.

Location and Setting are an important and often neglected aspect of your choice. Is it a country you're excited about living in? Is it an area of that country that you want to live in? Is it in a city, or a small town? Are there lots of international students? It's important to know that you'll be happy with the place that you live before you commit between one and five years to living there. If you're primarily after the life experience, then this aspect might even be the most important one.

Funding. When you're comparing offers, make sure that you understand and have clarified the funding situation with each. Email the University contact you've been assigned to and get them to tell you everything they know about the funding situation. Stipends, conditions, fees and general area living expenses vary widely.

Departmental Atmosphere can also be something you want to look out for - some Departments are very competitive and not very friendly. Others are extremely collaborative. You should know what kind of situation you're walking into before you make your choice.

You should make a decision based on some kind of weighted sum of the above factors. Everybody has their own weights, but not everybody actually knows what they are. If worst comes to worse, then a good piece of advice is to flip a coin, and then if you find yourself wanting to make it best of three, you know where you want to go.

One extremely important piece of advice: if you're deciding, visit all of your potential Schools. US programmes will often pay for you to do this (although in our experience they will place a cap on how much they will pay for visiting international students (e.g., CMU caps at $800, which is about all you will be able to expect, and may not be enough to pay for your whole trip)), but UK programs won't. When you visit, it's important to speak to your potential advisors and have a look around, but it's also incredibly important to speak to their students, and other students in the Department generally. Find out what it's like working for them, how happy their students are, how quickly they finish, and how much leeway they're given. If an advisor wants you as a student, they'll come across as friendly and helpful, but the kinds of personality quirks that you'll encounter much later might significantly affect the course of your degree. This seems to be less true in the UK than it is in the US.

Responsible Representation

As a South African student overseas, you're going to meet many students from other countries who are curious about South Africa and Africa in general, South African history and politics, what's it's like to live there, and why things are the way they are. For many of these students, you are their only source of information about SA.

That makes you something of an informal, mini-ambassador, which requires a fairly high standard of knowledge about your country. I would suggest that you read an introductory book on general South African history and South African politics, and that you make an effort to generalise your perceptions of South Africa and reexamine your opinions about major issues to make sure that they're fair and reasonably grounded in fact, or that you at least understand alternative points of view.

One of the most frustrating things about being a South African overseas is having to repair the damage done by other South Africans who have thoughtlessly shot off their mouths, presenting their own (often unfounded) point of view as fact, or spread common but untrue misconceptions. Don't forget that you know what it's like living in South Africa as you, which is almost certainly not the general case. Be as fair and as balanced as you can.