Right, I promised to give a not-so-quick list of seminar do's and don'ts. The context for this is the Honours students giving their 'end of labwork' talks in a few weeks, and there's a few scribbles on a Post-It note I'd like to share.
The thing to remember through all this is the point of a talk (or lecture, or seminar, or whatever you want to call it). You are trying to convey a message, and most of the time that message is "Look how brilliant I am". Seriously. You are using your work to convince people in the audience to think "Hey! This cove deserves a degree!" or "Hey! I'd really like this person to work in my lab", or even "I'm glad I employed this person in the first place". Anything that distracts (or detracts) from your message is a bad thing.
That is why I've come down so heavily on the presentation aspect. It does matter, because people (being human) will focus on what they perceive to be wrong, and you do not want them thinking about the bad when they could (should) be focusing on your brilliance. Now, obviously you're not going to please everyone all the time, but you can certainly put your thumb on the scales.
All these points arise from sitting through various seminars. Unfortunately it's not just students who make basic errors; the pros do it too. A couple of weeks ago an email came round the 'academics' mailing list in the Cage, noting that our courses were getting a bad rap in the national press. We've all got something to learn.
The first point I'm going to make is so obvious I'm surprised it needs to be said. But if you get it wrong it's probably the one thing that will make people think "What a prat". It's so important that I'm going to put it on a separate line:
Spell check your slides.
Honestly, this is not difficult (and I'm not talking about technical terms, I'm referring to common words). On my computer at least, questionable words are automatically underlined, but the ability to check spelling in any document has been standard for essentially for ever in computing terms. And if a word looks all right to you but is flagged by your spell checker (Americanisms are a big gotcha, especially because the Americans themselves are deeply confused about -ise, -ize and -yse), then look it up. The OED is free to use from a USyd computer, so make use of it.
Related to this, if you're writing complete sentences rather than bullet points then check the grammar. This is slightly more difficult, but ask someone in the lab to check your slides for you. And if you're dyslexic you should do this anyway (spell checkers will not tell you if 'their' or 'there' is correct in context, for example).
Let's make that more formal. Even if you are an English geek, you should practise your talk in front of a friendly audience before you do it for real, so that they can spot any glaring errors that you have made. Proofreading shows respect for your intended audience. You can work out the implications of that for yourself.
Also remember that certain sad bastards (ahem) delight in looking for words in protein sequences. Find them yourself first; make a joke of it if you like.
Check that your slides are legible, from the back of the room in which you will give your talk. Easy to do, very few do it. Nigel makes a good point from the presentation point of view that serif fonts are generally bad when projected. Times New Roman is just one example of this.
Similarly, the use of colour can make or break a presentation. I still prefer plain black on a plain white background for clarity, with maybe an institutional crest in a corner and a coloured bar (horizontal or vertical) for relief. Recall that 10% of your audience might be colour-blind. Time spent on designing a flashy theme is time better spent on experiments, writing or going to the pub. You are not trying to sell pharmaceuticals to impressionable physicians, you are trying to demonstrate how clever you are. Just because Powerpoint and Keynote have these flashy templates does not mean you have to use them. Remember, you do not want to distract people from your message.
Which leads nicely into the next point: Transitions. I loathe Powerpoint transitions. Most of us have seen them all before and they're boring and, again, distracting. I have to admit that Keynote has one or two nice ones, but they're likely to become just as clichéd in a couple of years' time.
At the opposite end of the spectrum from yellow on blue Comic Sans with whizzing transitions is Sir Paul Nurse. Sir Paul gave a talk a couple of years ago at my institute in Cambridge, and boggled us all by hand-drawing his slides on the fly, on the OHP. Unless you are a very good artist, I do not recommend you try this, at least not until you're as famous as he is.
Keep your slides simple. One message per slide. One, at the most two, graphs or figures and no more than five bullet points — but not on the same slide as your graph! Highlight key points. Do not write out what you are going to say (more on this further down). Try to make all your slides legible without needing to turn down the room lights. This is difficult and not always possible.
The trick is to imagine that Professor Sleepy has dozed off during your talk, and wakes suddenly during a crucial data slide. She should be able to look at your slide and at a glance understand your point. Consider whether you really need to show a sequence alignment. These are notorious to get right, and it's often better just to leave them out. Think about how you are going to show data: It's good to see raw data, but it's also good to see controls versus experimental in a clear, striking graph.
When you finally give the talk for real — and I hate practising talks; I can never get into the swing of it without an audience but as I say, it is essential to get others to proofread your slides and your verbals — there are more things to remember.
Your talk should be structured. There should be a clear introduction; start large and focus in on your speciality. Contextualize. Explain why you did what you did, and what you were trying to find out. Concentrate on the data, and never, ever merely read from the slide. We can see and read that for ourselves; the slides are there to show data and to remind you what to say, to expand upon the core message that you are showing us. Conclude with a clear message.
I don't think it's important whether you give your acknowledgments at the start or the end, or during the talk. When I've described work that a few people have done I like to start by naming and thanking, so that the audience knows who I'm talking about when I say "David suggested we do this". It's possibly better that students save it for the end, but it's not something I'm worked up about.
So you're standing, giving the talk of your life, and you're facing the screen. Wrong, wrong, wrong. When we used glass slides and projectors this was more excusable, but the chances are that you'll have a laptop between you and the audience. So look at the audience, and glance down at the laptop. Ideally you should not even need the slides to give your talk, so except for pointing out things of interest you should never have to look at the screen. Talk to us! If you find yourself unable to remember your slides, and without a screen in front of you, then stand slightly to the side so you can see the screen out of the corner of the eye. Remember you're trying to impress us, not the screen.
Do not rush. Do not whisper, but neither should you shout; project your voice. This takes practice. Use simple terms, not jargon, whenever possible. Pace yourself, and make sure people can understand you. Look directly at members of the audience to confirm that they are getting it (and don't get too depressed if people are asleep. It happens to all of us). Work on your delivery, on capturing and holding people's attention. Consider what you're wearing; again, give no one an opportunity to focus on you or your foibles rather than your message.
"Hold the bloody pointer still", it says here. Most distracting when the little red dot is jerking all over the place while you're talking. Don't fiddle with change in your pocket (my biggest fault!) but don't be afraid, within reason, to use your hands to describe or emphasize. Don't 'um' and 'er', and don't say 'kinda', 'like' or 'sorta'. You may say 'innit' for comic effect. Don't be afraid to use humour, but you had better make sure it really is funny. Don't forget the names of collaborators and colleagues.
At the end of the talk, make it obvious you've finished. If someone asks a question and you can barely hear them, repeat the question (ask them to repeat it if necessary) so that the entire audience can hear. Address the questioner but talk to the rest of the audience as well (a difficult trick. I'm still working on it). If you don't understand the question, ask the questioner to rephrase or repeat, and then repeat your understanding back to them to check (this rephrasing the question in your own words is a useful trick. It lets the rest of the audience hear the question, confirms that you're answering the right question, and gives you time to think). Don't be afraid to stop and take a drink if you're dry; this also gives you more time to think.
I get a real blast out of giving talks, and one of the most important things to remember is to have fun. It's a crucial skill to learn if you're going to make a career out of this game, so you may as well enjoy it. Finally, realize that criticism of your talk is not aimed at you personally. Take criticism constructively, and learn from it.
. 
.
and was gratified that, after *mumble* years I still remembered how to use one, onehanded. Yay. Go me.
.
