How to give a paper presentation
Derived by Gregory Dudek from material originally developed by Hong Zhang, University of Alberta (with permission). Hong, howeever, may well disagree with some of the comments or directions this has evolved in.

Summary
This article gives some suggestions how to prepare and deliver an oral presentation for an academic audience, in particular for a conference presentation or in a class. It is intended for students who have to give such presentations for their courses, but the principles described apply equally well to most scientific and business presentations. It giving presentations makes you uncomfortable, you might naturally seek to avoid situations where you wil have to do them. My advise is the contrary: try go get in situations where you will get the practice you need to become more proficient, comfortable and confident.

Contents

Summary

Contents

1 Introduction

2 Interpersonal factors

3 Structure and planning

4 Timing

5 Preparation

6 Visual aids

6.1 Slides

6.2 Computer and video displays

6.3 Visual cliché

6.4 Grammar and spelling

7 Delivery

8 Questions

1 Introduction
Giving oral presentations is a critical skill in both the research environment as well as in many business contexts. For that reason, oral presentations are sometimes expected as a part of graduate or (less frequently) undergraduate courses. As Hong Zhang eloquently put it: professors don't make presentations part of a course ``because they like to see students squirming nervously (although this is part of the fun), but because the ability to give such a presentation is an important skill, and one that can only be developed by practice.''

2 Interpersonal factors
When you are giving an oral presentation, remember that everybody listening is hoping you will do well, since that will allow them to get more out of it. Futhermore, the listeners usually experience a substantial degree of empathy with the speaker: if you stumble or forget something, they will be on your side and almost always sympathise. Don't let small errors unnerve you, they can actually help establish rapport with the audience (so long as you don't do over-react yourself). When something goes wrong, it's fine to comment on it (briefly). You want to establish a connection with the audience. You might respond to them or your surrounds. You should try to avoid a deadpan delivery that would be unchanged if you were giving it in an empty room.

Try and establish a connection with the audience: look at them, ask questions, be aware of their reactions. On the other hand, you can also think of yourself as a lion tamer somtimes: you need to encourage the audience to react appropriately, sometimes by cajoling them to think, ask or get involved, and sometimes by politely getting them to stay quiet.

In general, try and assure that you are dressed as least as well as your listeners, that you are polite, and that you remain professional.

There are a few things that will cause the audience to lose sympathy with the speaker, and thereby lose interest in the talk. Here are some examples of behaviour that will probably be perceived negatively by some or all of the audience:

We tend to make judgements about other people very quickly and often on grounds that a little extra thought would show to be irrelevant. The cliché that `first impressions count' is as true in giving presentations as it is everywhere else. If you start well, it will give the listeners a good impression, and they are more likely to take your presentation seriously and pay attention.

3 Structure and planning
When planning your presentation, bear in mind that most people have an attention span of about ten minutes for subjects in which they aren't particularly interested. You should not expect people to be able to concentrate for longer than ten minutes without some change of subject or in the method of delivery. The issue of timing is discussed in more detail below.

If your presenation is poorly structured or difficult to follow, people won't be able to concentrate even for the ten minutes they normally manage. This means that you need to have a very clear structure to your presentation, so that the audience knows how it is progressing. A good general plan might be something like:

This structure is not very inspiring, but it is practical and will not confuse the audience.

Give each section of the talk a title, and list these explicitly at the start of the talk: ``First I will describe... Next I will explain...'' This is not a waste of time, even if you have only ten minutes. It will make it easier for the audience to absorb what follows.

4 Timing
The shorter the talk, the longer it will take to prepare. This is because it is harder to say everything you think is important in the given time. Most business and some scientific presentations have to be limited to about ten minutes. In my experience, the majority of beginners tend to over-run, so keeping to time is a way to make a good impression.

There is only one foolproof way that I have found of keeping to time: prepare your presentation on the basis that you have even less time. For example, for a ten-minute presentation I would aim for eight minutes, on the basis that I usually over-run by about 20%.

If you find that you are running out of time towards the end of a presentation, you should decide which of these emergency measures you wish to take:

An option that you don't have is to carry on regardless, and hope nobody minds your over-running.

5 Preparation
The most common problem with student presentations is that the presenters underestimate how long it takes to prepare and practice. I find that it takes me a whole day to prepare a ten minute talk. I usually want to rehearse the whole talk about ten times before I feel ready to deliver it.

In my experience, the shorter the talk, the longer the preparation (roughly). This may seem strange, but if you have only a short time you have to think much more carefully about what to include and what to leave out.

The original guidance from Hong said you should practice the presentation at least once in front of another person, who should ideally be as critical as possible. . This is fine for an unimportant or informal presentation. One of my very renowned colleagues requires his students to do a full practice of any conference talk, after the slides have been fully complted, a minimum of 10 times. This is not unreasonable for an important talk, although it might sounds surprising. Some of these can be done without an audience, but they must be done aloud and standing up.

5.1 Knowing your material
It goes almost without saying that you should thoroughly know your material and the background ideas of the work you are presenting. In particular, however, you need to be especially ready to talk about some things in detail. In order of increasing importance these are: any items listed on slides, anything to say to call attention to verbally, any equation on any slide, any items on a slide that you explicitly refer to verbally, any equation you explicitly refer to verbally.

6 Visual aids
Most people who give talks like to support them with visual material, usually slides. You should take the same trouble with the preparation of these as you do for the spoken material. Bear in mind that your visual aids should support the presentation. They should not duplicate what you are saying, and you should avoid the temptation to read them to the audience.

6.1 Slides
A talk with slides is the usual way of presenting scientific results and business plans, so it is important to be familiar with this type of presentation in particular.

You need first to decide how many slides you will show. For a short talk I find that one a minute is about right. For a ten minute presentation, I will normally produce ten slides with the most important information on, and another two or three with supporting material. I normally try to anticipate the sorts of questions that people will ask, and prepare supporting slides to illustrate my answers to these questions. Usually I find I don't need all these extra slides, but it doesn't take long to make them.

For a one-hour presentation you can take things a bit more slowly; in this case one slide every two or three minutes is probably about right. Of course you should not follow these guidlines dogmatically; if you find you need to show more slides then that's what you should do. If a slide contains complex experimental or statistical results then it will take longer to explain. In this case you may find you can only show a few slides in a ten minute presentation.

The trusty old overhead projector was once the normal method of showing slides in the academic environment; now it's the data projector. A common mistake is for the speaker to stand in front of the screen and obscure the audience's view. Often the projector and screen are deployed so badly that it is impossible to avoid blocking someone's view.

On the whole it is possible to convey more information on colour slides than monochrome ones, but it may be that it isn't worth the extra effort to produce them.

On an overhead projector slide, good text sizes are about half an inch for ordinary text and one inch for titles. If the slide is about eight inches high, this means you should expect to get about seven lines of text on a slide. The `rule of seven' states that people are best able to assimilate an image that has about seven entities in it, so this is a good size from a readability and a psychological standpoint. If you photocopy slides from pages of a textbook or other printed material, you can be sure that hardly anyone will be able to read it.

6.2 Computer and video displaysi You should check in advance that the equipment will be compatible with the software you use (for our class, my laptop runs Mac OS X). I have seen presentations fail to start at all, because the presenters prepared their visual material on computers with very high resolution screens. When transferred to the much lower resolution of the video projector the material did not fit on the screen. The solution is simple: check before you go, and take back-up material in case it doesn't work.

6.3 Visual cliché
If you only produce presentations occasionally, you may feel the urge to over-produce the visual material, using all sorts of fancy graphics and symbols. I suggest that you think quite carefully about not doing this. The first time I saw a bulleted list where the bullet was a pointing finger I though it was quite stylish. Now I have seen it hundreds of times and I find it annoying. This is an example of a `visual cliché': an image that appeared charming and amusing when first used, but now only serves to annoy. I currently feel that one should only use images at all if they are part of the content of the presentation. The fact that software like PowerPoint or Keynote allows you to produce slides whose background is a mountain range in no way suggests that you should do so.

If you are using colour you have to be even more careful. Bear in mind that a fair proportion of people are colour-blind to a certain degree, so some colour combinations (e.g., red and green) should probably be avoided completely. A popular colour combination is yellow text on a dark blue background. In fact this is so popular that I now avoid it.

In addition, be wary of the fact that the projector or room lighting my be inadequate. In MC103 this year, there is a lot of ambient light leading to poor contrast on the screen. In general, I prefer black text on a white background. It's a bit boring, but it works reliably and also prints out welll on paper. Never ever used blue/violet text, nor pairs of foreground and background colors with similar luminance.

Here are some other examples of visual cliché, particularly common in computer-assisted presentations:

6.4 Grammar and spelling
Most academics, most scientists and many businesspeople are fussy about grammar and spelling. Some mistakes are worse than others, and each person has his or her own particular foibles. With me it's apostrophes: I get very grumpy when I see apostrophes used incorrectly [-hz].

Other people find other things objectionable. The only way to avoid annoying anybody is to make sure your grammar and spelling are faultless. You should bear in mind that university lecturers read a great deal, and will be used to reading material where the writer has taken a lot of trouble over the grammar, spelling and presentation. To give a presentation where you have not taken trouble with these things is like turning up for a job interview wearing beach shorts and T-shirt (unless you're applying for the job of lifeguard).

If your grammar and spelling are not very good, in the short term you need to find someone who is suitably expert and ask him or her to check your material. This can be a tedious and unpleasant job, but it is important. Also, you should encourage listeners to actually point out issues with your oral presentation. (In the long term, you need to address the problem directly.)

7 Delivery
Not everyone has an orator's speaking voice; however, everyone can be understood if he or she speaks at an even pace and faces the audience. The two most common delivery problems in our community are speaking too quickly and facing away from the audience. Less intrusive but still annoying are fidgetting and repetitive gestures.

When practising, you should ask your colleagues to shout at you if you do these things. In a large lecture theatre hardly anyone has a loud enough voice to be heard while facing away from the audience. If your voice is so quiet that you can't be heard even while facing the right way, then you need some sort of amplification. You should check that this will be provided if you need it. Not everyone has my fog-horn voice, thankfully.

Many people fidget during presentations. Partly it comes from not knowing what to do with your hands: people find it surprisingly uncomfortable to have their hands on public display and not to do anything with them. The same applies to a lesser extent to feet. I once attended a lecture where the speaker -- an eminent professor -- paced the entire width of the lecture theatre for the whole duration of his talk. It was particularly noticeable as the talk was being filmed, and the camera operator had to chase the speaker from side to side. I guess the speaker probably did not even know he was doing it (but the puffing cameraman should have given it away).

The advice that most experts give is that you should stand with your feet about shoulder-width apart, and with your hands clasped behind your back. You will have to move, of course, to change slides and point at things, but other than that you should keep still. Personally I find this advice extremely dificult to follow. When I talk in social situations I am quite animated: I wave my arms and point at things and generally jump up and down. I find that I can't entirely suppress the urge to do this during presentations. However, if you have a tendency to fidget with your hands, you should certainly not put temptation in your way by, say, having a pen in your hand.

Some universities make video recordings of student presentations so that people can see themselves as the audience sees them. Personally I think this is a good way to destroy any self-confidence that a nervous speaker has developed, so I don't do this.

There is only one foolproof way to get these problems of delivery under control, and that is by practising over and over again, perhaps in front of critical but sympathethic colleagues.

Many people find that they need written notes to help them remember what to say. When I first starting giving presentations I used notes. However, I soon found that I never looked at them, so I stopped using them. If you need the reassurance of having notes, then that's fine. The advice I received as a student, and that I still think is sound, is to make your notes on small cards rather than on paper. You can keep the cards in you pocket or otherwise out of sight, and only refer to them if you need to.

You should not read your notes out loud. Even if you have to read your notes and then speak afterwards, this is better than reading out loud. There are two reasons for this. First, if you are looking at notes you can't be looking at the listeners, so they won't be able to hear you as easily. Second, it will annoy the listeners, who will rightly think that if they wanted to read a presentation, they could read it themselves. Thirdly, without making eye contact you won't be able to determine if there are questions or other concerns from the audience.

You should practice your talk sufficiently that you can avoid reading your notes out loud, even if this means that you say less and make a few mistakes.

Another piece of advice I received as a student, and that I still follow diligently, is to memorize the first and last line of a talk. I would not try to learn by rote a complete talk, even a short one, but to learn the first and last line is a sound idea. At least your talk will start and finish smoothly, even if you make occasional mistakes in the middle.

I find that by practicing a talk aloud several times, I also memorize certain nice phrases or chunks of text that sometime re-emerge during the actual presentation. Be sure to practice the entire talk, end-to-end, several times.

8 Questions
At the end of a presentation you will normally be expected to take questions from the audience. If you have prepared properly and know the subject this will not present any problems for you. In a course presentation, the instructor may well ask questions that are very difficult to answer. The reason for doing this is to see how you handle them. The correct response to any question for which you don't know the answer is ``I'm sorry, I don't know''. You should never waffle or answer a different question. I can't stress this strongly enough. Time is short in a presentation session, and no-one wants to have it wasted by waffle. There is absolutely no shame in saying that you don't know the answer to a question.

Similarly, if someone asks a questions that does not seem to be related to your presentation, it is perfectly acceptable to say ``I'm sorry, but I can't see how your question is concerned with this subject''. Of course you should not be rude to the audience, but at the same time you should not be expected to answer irrelevant questions.

I find that I can often predict the sorts of questions people will ask about my own talks. I usually practice my answers to these questions in the same way that I practice the rest of the talk.