«The Unwritten Rules of PhD Research Gordon Rugg Marian Petre THE UNWRITTEN RULES OF PHD RESEARCH THE UNWRITTEN RULES OF PHD RESEARCH Gordon Rugg and ...»
• Did you get their name and title right?
• Does your question show that you’ve done some homework?
• Does your mentor/supervisor think that your question looks interesting?
• How long would it take a reasonable human being to write a reply to your question? (If it’s more than ten minutes, then consider rephrasing the question.)
• Is your message so long that it scrolls off the page when the addressee opens it? (If so, shorten it.)
• Does the message show you in a good light, as someone who can spell, write clearly, think and generate interesting questions?
• Does the message offer them anything (e.g. access to data), and if so, can you deliver on that promise?
At a meeting or conference Have something interesting and relevant ready to say. A compliment is handy, but be prepared to follow it up with a question, otherwise the conversation will die. (Even the most eminent researchers can be embarrassed by compliments, especially if they’re too gratuitous.) It’s best to have a question prepared that requires a multi-word response: for instance, ‘Professor Katz, I was intrigued by your paper in Nature on semi-stochastic systems. I wondered whether you had tried applying that approach to trade networks?’ Use the opportunities that the meeting provides. If your person asks a good question during a session, you might catch them after the session and remark on the question and its implications. If you see your person talking to someone you know, you might ask the person you know to introduce you. If you see your person in a loosely arranged group, you might stand visibly on the periphery until you get a chance to make a contribution (a short question or a joke is good) or ask if you may join the group. A good time to catch people is as a session breaks up, before they’ve found their way to coffee or lunch. But don’t keep them from refreshment – offer to walk with them.
52 THE UNWRITTEN RULES OF PhD RESEARCHHave a business card to hand, and perhaps a copy of a summary of your research (previously read and approved by your supervisor).
Don’t assume that you are beneath notice, or, worse, beneath interest. Here are some home truths about Great Researchers to help you put them in
• they are usually great because they love ideas and asking questions – so they usually have an appetite for nifty ideas and good questions;
• they are usually just as susceptible to deft ﬂattery as the rest of us;
• they were once research students and many of them still remember that.
People you should remember to include in your network Most of the section above refers to contacting researchers about research.
There are other categories of very useful people that it’s easy to overlook, so we’ve included a short section about them here.
Mentors Mentors are Wise People who take an interest in your personal, professional and intellectual development. They’re the people who teach you the ‘unwritten rules’ and who can see the ‘bigger picture’. In theory, your supervisor should be a mentor, but it doesn’t always work out that way. A mentor is a more experienced researcher who will show you the ropes from this perspective of success and informedness. A mentor can show you the things you don’t know how to look for. This is particularly useful for the things that you don’t know that you need to look for or do, such as getting the right things on your CV as early as possible – a friend won’t always know what the right things are, and some of the right things are counter-intuitive. Other students are often good at identifying things actively wrong with your institution, but bad at identifying things which are passively wrong (i.e. things not being done which ought to be done). Mentors are useful for this, among many other things.
Secretaries and other support staff Always treat support staff – secretaries, technical support, custodians of facilities – with respect. Never underestimate their value. Never confuse salary level with worth. Support staff are the keepers and collators of useful information – they are the ones holding together the department, they are the providers of services and assistance, they are often the gatekeepers to things you need.
Consider: if you really wanted to know the inside story about government policy, would you ask the prime minister, or the civil service? Journalists know all about this, and successful journalists always get on well with the secretaries of the people they investigate.
NETWORKS 53 Wonderful People One invaluable resource is Wonderful People. There are a few people who have invaluable skills such as improbably excellent social or professional contacts, or encyclopaedic knowledge of one or more literatures, and who are helpful and pleasant. Such people should be cherished and appreciated. As a new researcher you will probably not know any people ﬁtting this description (or more probably, not realize that you know them). When you do start meeting them, treat them well; they should be declared living national treasures.
Librarians and secretaries are also often wonderful, and friendships with them are almost always well worth cultivating.
6 Reading... those frightful parts of the Pnakotic Manuscripts which were too ancient to be read.
During a PhD, you have to do a lot of reading. That reading needs to be the right reading, and you need to make correct use of it. This chapter covers these topics.
The chapter overlaps considerably with the chapter on writing. This is for various reasons. One is that you are doing the reading so that you have the information required for writing (for instance, writing your thesis). Another is that if you know about the concepts and structures used in writing by experienced academics, then you will be able to make much better use of what you read.
We have included a fair amount of detail about online searching, as opposed to internet searching. If you’re going to be a good professional researcher, you need to be aware of the differences, and able to use the tools of your trade efﬁciently and well. Few students are strong in this area; few students do outstanding work.
We have also included a fair amount about different types of papers and about different types of research. If you understand these concepts then a lot of things about the literature start to make more sense. It’s also worth reading this part if you’re thinking about writing a paper, or about planning some research with a view to publication outside the dissertation – if you know what you’re aiming for, then you can plan and conduct your research much more efﬁciently.
Finding the right references
You have to convey the right message when you are writing, and that involves some hard work beforehand reading what you need to read (the ﬁrst golden principle: don’t lie, in this case by pretending to have read things which you haven’t read). However, there is no point in overkill. One of our usual examples of strong academic writing contains eight references in one paragraph on the ﬁrst page. Do you think that page 24 of that thesis contains the same number of references per paragraph? It doesn’t, because it doesn’t need to – the writer has by then already cited practically all the references he or she needs to. How do you know what references you need to cite?
The easiest solution is to ask your supervisor, politely, where to start. If your supervisor doesn’t know, ask someone else, politely, and keep your supervisor informed, in case you start blundering in where angels fear to tread. You need to send out the signal that you’re a hard-working individual who will make good use of the advice, rather than an idle brute who can’t be bothered to do their own research (mentioning what you’ve already read and asking where you should go next is a good start). Your supervisor can be invaluable here.
If you are lucky and virtuous, your supervisor might say something along the lines of, ‘The person to talk to about this is X; I’ve emailed them, and they’re happy to give you some guidance. Here’s their email address’. This is an
encouraging sign, and is academic shorthand for the following things:
• here is something which will save you a lot of effort;
• here is a chance to make contact with a major player in this area;
• I trust you to enough to let you speak to important players in this area by yourself.
(It does not mean, ‘I am too ignorant or idle to provide guidance on this by myself’.) If your supervisor offers you this opportunity, then grab it with both hands.
The researcher’s core literature Most good researchers carry around in their memories a good, accessible database of relevant papers. That means, out of all the reading they do, they maintain a working knowledge of about 100–50 papers. As they continue to read, the core adjusts, shifting to follow developments in the discipline or to follow their changing interests. But some of that core will persist for years.
Interestingly, the lower limit for a good bibliography on a dissertation tends to
56 THE UNWRITTEN RULES OF PhD RESEARCHbe around 100 citations. One of the things a good doctoral student will accomplish is to amass a ﬁrst ‘core’ literature. (Of course, there’s huge variation, but the numbers don’t really matter – the idea of keeping a selection of pertinent literature accessible in memory does.) So, where is a good place to start looking for references? Your supervisor is likely to remind you about literature reviews in the papers you’ve already read, and also about review articles. You may also be pointed towards some online searching, with keywords either supplied by your long-suffering supervisor, or included in the articles which you have already found. The following sections cover these topics in more detail. We’ve included some discussion of how papers as a whole are put together, not just literature reviews, to save duplicating this in the section on writing.
After that, you have the problem of evaluating the quality of what you’re reading. This is not always easy for the average student. If you ﬁnd a paper impossible to understand, is it because the paper is far too brilliant for you to understand, or because it’s a pile of pretentious, obfuscatory garbage? A later section of this chapter describes some ways of reading between the lines of academic writing, so that you are in a better position (a) to evaluate what you’re reading and (b) to improve the quality of what you’re writing.
Academic papers and dissertations normally begin with a literature review.
There are good reasons for this.
The ostensible reason for a literature review is to set the scene for the work described in the paper – explaining what has been done previously by other researchers etc. This is done via standard referencing conventions, so that interested or sceptical readers can locate the original sources and read them to check the alleged facts in the literature review, if they so wish. The second, and equally important, reason for a literature review is to demonstrate that you have done your homework thoroughly, so that readers are assured that they will not be wasting time wading through the rest of what you have written.
The literature review needs to have a structure, since even the best academic prose is pretty hard reading at the best of times. The structure is also a way of demonstrating that you have a clear understanding of what you are doing and why you are doing it. It is your responsibility to make your work understandable; it is not the reader’s responsibility to make sense of a pile of references indiscriminately grabbed from the internet and then tacked together with semi-coherent prose.
The usual structure, and one with which we have no quarrel, is one which begins with the earliest work in this area and proceeds via the most important past work up to the present. Your references will therefore usually begin with READING 57 old seminal references, then continue with more recent key references and assorted examples of less important references, and end with very recent foundational references. Some readers may notice the similarity to the Whig view of history; this is probably a suitable subject for a paper which would be viewed as quite amusing by at least four readers...
One widespread source of confusion is the link between literature reviews and introductions. Institutions and people differ. Some favour a completely separate literature review and introduction; others favour a complete integration of the two. The best advice is to ﬁnd out whether there is a speciﬁed formalism for your venue (including PhD regulations). If there is, follow it; if there isn’t, use whichever approach you prefer. There’s no point in getting into a war on this topic.
At the heart of your literature review is a good plot. The story should start with a problem of some sort (for instance, a dragon laying waste the land, in a good legend, or a problem in the domain, in research). The literature review and/or introduction then follow the steps taken by previous work in an attempt to resolve the problem. The literature review and/or introduction end at the point where you, the hero or heroine, enter the scene, armed with your enchanted sword/improved research methodology. The rest of the paper/dissertation follows your adventures, to the point where you emerge triumphant. If you do not emerge triumphant, then you should have got your experimental design right before you started, and it is your problem. (Interested readers might wish to try reading Propp’s work on formalisms in Russian folk tales, or Campbell’s examination of archetypes, or a good book on experimental design, depending on their precise problem.) This strand is known by various names, such as ‘plot’, ‘red thread’ and ‘narrative spine’, and is viewed as extremely important by most experienced and able writers.
Novices usually have a lot of trouble with narrative spine. The situation normally improves with practice, if you deliberately work at it, but will not automatically improve otherwise. There are various ways of helping yourself with this issue. One simple way is to use top-down decomposition. This involves starting with a very short list of key points in the story – half a dozen
brief sentences at most. For example:
Elicitation of software metrics via card sorts
• Choice of metrics for software is difﬁcult
• Card sorts should have advantages over previous methods for choosing metrics
• What happens if you use card sorts in this area?
• Card sorts do have advantages over previous methods Once you are happy with this top-level structure, you then break down each part of it into smaller parts, and keep on repeating the process as necessary.
58 THE UNWRITTEN RULES OF PhD RESEARCHYou should end up with a set of section headings, subsection headings etc.
which will give you the main structure.
In practice, readers tend to get lost quite easily, even in a well-structured paper, because of the sheer volume of information which should be in there.
(If there isn’t much information, this is usually a danger sign.) The wise writer therefore uses bridging text and signposts. Bridging text is used to join two sections of a paper or other document. It usually consists of a closing paragraph or two at the end of a section, summarizing that section, telling the reader what will be in the next section and explaining how the previous section leads on to the next section. A signpost is a piece of text ﬂagging (i.e. indicating) something which will be mentioned later.
At this level, you should be making extensive use of journals as your main source of information. Although textbooks and the internet are useful starting places, they are usually not appropriate as main sources of information because they tend to present simpliﬁed accounts.