What are you aiming to do? The examiner has set a title which she believes will give you the chance to display your knowledge and understanding of a branch of science and your ability to think creatively within it. It's up to you to convince her she's right! You must therefore first understand exactly what sorts of knowledge and skills she's looking for and then construct an essay that makes it as easy as possible for her to see that you have them. Let's deal with these two aspects in order.

1. Has he understood the question? A good exam question will not be a simple stimulus to regurgitation of a specific lecture but will be an attempt to challenge you to put your knowledge together in a new way just for this occasion. The first challenge, therefore is to deconstruct the question. What main topic area is it addressing? What "spin" or "angle" has been put on the topic? Is it inviting you to integrate two or more areas of the topic? Is there specific proposition which you are being invited to address. If you convince the examiner you have understood these things, then she will be half-way convinced that you have a good and lively grasp of the subject from the outset.

2. Does he know the basics? There will almost certainly be a standard body of facts, experimental approaches and arguments that the examiner is expecting you to display. You might call this the "paradigm" of the field or the "bread and butter" of the essay. Make sure that you do set this out clearly and succinctly in the essay. No amount of original flair and insight will compensate for fundamental holes.

3. Is he aware of the current questions? In most subject areas, research continuously throws up new evidence that puts old ideas in doubt. You need to be able to show that you have been keeping up with things enough to be able to summarize the current areas of doubt and controversy.

4. Can he analyse the data? Where there's controversy, this might be for a number of reasons (e.g. below). Can the candidate spot the experimental or logical errors?

   1. Two groups report doing exactly the same experiment but get different results (rare. More usually one of the others)

   2. Two experiments are ostensibly the same but when you look at the way they do it, they are using materials or methods that are different in a crucial respect that might plausibly account for the differences.

   3. One group has failed to put in the appropriate controls for specificity, quantification, etc. and is thereby over-interpreting things

   4. The two groups are doing quite different experiments, the results of both of which could, with a dispassionate eye, be seen to be consistent with the interpretation of one or other group (or a different and simpler interpretation)

   5. The differences between the two groups is a semantic one. That is, they use different terms which color things differently but are not in fact, disagreeing on any real point of substance. They're just so used to disagreeing that they have got into the habit

   6. The data are good, they're just not interpreted in a statistically, mathematically or otherwise rigorous fashion.

5. Can he synthesize complexity? The issues raised in your analysis are likely to be complex. Can you bring them together, summarize and organize them well?

6. Can he present a clear case? Does he have the rhetorical skills to convince the reader that his analysis and interpretation is sound? Does he convey the excitement of the subject.

How can you make sure that your essay conveys the impression that you have the knowledge and skills outlined above?

1. Take 5 - 7 min to make a plan. This will make sure you have assembled all the arguments and pieces of evidence in your mind and then chosen the order in which to present them. It will save you from forgetting to mention things you wanted to say in the fervor of writing your purple prose. It will make sure that you think about the question and are tailoring your essay to respond specifically to it.

2. Decide upon the main pieces of experimental or epidemiological evidence you are going to present. You will need to back up each of your main arguments with evidence, and so make sure you can do so. You may or may not be able to remember names and dates but the important thing is to be sure that you know what the experiment was and what it showed. You should have developed a little bank of pieces of evidence during your studies (and have them clearly honed and filed during your revision, so that you are able to illustrate most aspects of the fields of interest with either the landmark experiment or recent data of an controversial or significant kind. As with quotes for Eng Lit GCSE, you should try to choose these pieces of evidence so that they have the widest degree of currency.

3. Write a good and concise introduction to the essay lasting ½ to 2/3 of the first page. Why is this crucial? Examiners are human (oh yes!) and they will form an impression, willy nilly, of whether yours is a 1st class, upper second or weak answer before they turn from the first page. If they've put you down as a second classer by then, there's very little you can do in the subsequent pages to push your mark back up through the first class barrier. They also tire and lose concentration, so they will be much more influenced by what they first read (and to a lesser extent by the final few words) than they will be by the bulk of the essay. Your intro, therefore is a chance for you to show that you have all the knowledge and skills they are looking for and the rest of the essay will simply confirm it. To write this section requires that you have produced a clear plan, so these two elements reinforce each other. There are many ways of writing an opening para, but here is a fairly reliable approach.

• Sentence 1. Play back the title to the examiner by explaining what the title means, what the importance of the question is, what angle you are being asked to take, and so on. If there is room for ambiguity in the title (does it include x as well as y) say so, and state your interpretation.

• Sentence 2. Explain the way in which you intend to tackle the question, the model systems you intend to cover, the particular controversies that shed light on the issue, the lines of argument you are going to pursue.

• Sentence 3. Indicate the particular lines of experimental of epidemiological evidence you are going to employ and how they relate to each other and the title.

• Sentence 4. Very briefly, give a flavour of the conclusion you are going to reach.

4. By this point, you should have the examiner in a high state of anticipation, just hanging on the full exposition you have alluded to. Make sure you don't throw this away by rambling from then on. Make sure that the rest of your essay follows the logic of the introduction and introduces the models, examples, experiments in the logical order you have advertised. Make use of headings (underlined) and linking sentences to make sure that the examiner is in no doubt about the point she has got to in the argument (even if she has nodded of for a moment). Make sure to bring in the evidence you planned to use. Use clear and simple diagrams where they help to explain a complex relationship between variables, interactions in anatomical space or time or other things that are hard to represent clearly in words.

5. Keep your eye on the clock and make sure you are devoting only the right amount of time to each essay. Do not spend 2h on one essay and 30 min each on two. You can let your best essay take a few more minutes than your second and third ones but don't let this get out of hand. Within the essay, make sure you spend a proportionate amount of time on each section of the essay.

6. Bring the essay to a close neatly but don't write a long and fatuous closing paragraph ("thus it can be seen that...") unless you really do have a number of threads left to tie together. If you do, a closing para along the lines of "using experiments on Martian guppies, we saw that Bloggs found a relationship between the rate of zarquon injection and defloppulation time but that Smuts, using domestic Nargs found no such relationship. Given that Smuts used a transcutaneous flipple blaster that we have seen gives anomalous results in the well-attested rambunctaceous assay, I believe it is not possible to conclude that the difference in results reflects differences in the rebuistular physiology of the two species and that a better controlled study using standard methods would be decisive" might be called for.

These skills are not natural and, to a great extent, the hope is that you have developed them over the last three years of scholarly endeavor. However, there's nothing like a bit of intensive practice to hone these skills to perfection and I suggest that you do the following:

1. Get lots of practice writing introductory paras. You can go over titles you've already attempted to see if your skills are improving. This should take perhaps 20min for each essay as a revision exercise and makes sure you get slick at making plans as well.

2. Practice timed essays. For most of the essays to which you have written intros, put some 3h periods aside to write full-length essays, to hone your timekeeping and concentration skills. Reduce the time you give yourself for each essay by at least 10min, to account for the preparation time you've already put in to it.