Giving a great scientific talk

By Shaun Hendy 18/01/2010


In the Southern hemisphere, conference season is just a week or two away.  Graduate students south of the equator are beginning to get that sinking feeling.  They’ve been dreading it for weeks, and their data only came through the day before.  Yes, it’s the thing a student fears most:  the conference talk.

Do I need to stress that your conference talk is very important?  When you give a scientific presentation, you present yourself as well as your results. If it goes well, potential post-doctoral supervisors in the audience will take note. If it goes really well, it will put you in the running for the prize for best student talk. If it goes really, really well, your supervisor will think of you when they realise that the all-expenses paid plenary talk in Hawaii they said yes to six months ago, falls on the day of their daughter’s first ballet recital.

Now that you are relaxed (no pressure), here are my top ten tips for a great first scientific talk:

  1. Enthusiam. Let your enthusiasm for your subject come through.  I’ve put this tip first for a reason — your audience will forgive many faults if you can infect them with your excitement for your topic.
  2. Interaction. Interact with your audience (although not by forgetting the laser pointer is on and dazzling the audience with it).  Make eye contact with people around the room from the outset and keep it — you are talking to the audience, so don’t read from your slides.  Avoid referring to notes. Pitch your voice so that those in the back of the room can hear; if necessary, ask whether you can be heard at the back right at the start.
  3. Ownership. Take ownership of your work.  Acknowledge the contribution of your colleagues as appropriate, but don’t motivate your work by saying your supervisor told you to do it!
  4. Audience. Think about your audience and remember that they have come to your talk to learn something. Make sure what you say is clear and understandable. Even the experts in the room will be impressed by an authoritative introduction to the topic. Leave your audience with a key message or two that will trigger their memories in six months’ time.
  5. Timing. Finish your talk within the time allocated, leaving time for questions.  Your audience may need to change sessions or find the toilet between talks.  Give them a fighting chance.
  6. Preparation. Practise your talk a couple of times, but remember that adrenalin will make it faster on the day.  Turn up before your session starts to make sure you know how everything works.  If you have your own laptop, practise making the transition in the auditorium, although this does not always guarantee it will work in the heat of the moment – best to have it on a memory stick just in case.  Consider preloading your talk onto the in-house machine (but check the fonts) or even another speaker’s laptop.
  7. Slides. Use strong, dark colours on a white background.  Better to give people a picture or a graphic before you launch into the detail, rather than after.  The text on your slides should be sufficiently concise to be self-explanatory, but no more.  Don’t go overboard with in-slide animations, they distract the audience and you will regret it as you wile away the time toggling through them.
  8. Equations. You should always talk through the terms in any equation you present, so a slide packed with equations will eat lot of time and bore your audience.  Presenting an unfamiliar equation without defining the variables for the audience is pointless, but remember it will take time to define these in the first place.  In other words, think carefully about your use of equations.
  9. Check your slides. Use a spell checker!  You’ll find it disconcerting (and potentially embarrassing) to spot mistakes when you are presenting.  It’s also worthwhile discovering the correct pronunciation of the name of the professor who wrote the seminal paper in your field – you will thank me when you learn she’s chairing your session.
  10. Questions. Everyone in the audience will sympathise when that awkward question comes up – we’ve all been there – so it’s OK to confess your ignorance or deflect (e.g. ’Lets talk about that in the break’).  Nine times out of ten, your ’awkward’ questioner will have the wrong end of the stick — a chat with them after the talk can be a much better way of finding this out than a lengthy exchange in the auditorium.

Good luck! And yes, it does get easier.


0 Responses to “Giving a great scientific talk”

  • These are great. You can give endless recommendations for talks… here’s a one more idea:

    The “acknowledgements” slide is traditionally the last, but I’ve seen too many first-time speakers (even a few experienced speakers) get to the last slide of the content of the talk and completely forget to show the acknowledgements. My solution is to (also) put the acknowledgements on the first slide. It also shows you are a “team player” from the onset.

  • Great advice Shaun. Personally I find that the talks I gel with the most are the ones that begin with one or two key questions on which the research is focused. Having this question based approach livens things up a little from the standard (and often rather dry) list of aims or objectives.

    My heart starts thumping just thinking about giving a presentation, but it all adds to the post-talk euphoria!

  • Excellent post, Shaun. I’ll pass it on to the folks who coordinate our ‘research skills’ paper for the grad students; it will be very helpful for our students I think. (Some of the worst talks I’ve been to have had every slide crowded with 16-pt text – which has been read out verbatim – & horrible intrusive animations.)
    On the laser pointer – I’d say, ditch it altogether. If you’re nervous about the talk your hands will shake & that translates into quite a lot of movement by the pointer on the screen. Also, I’ve been to altogether too many presentations where the speaker has been so keen to emphasis something that they’ve swirled the pointer around a word or phrase to the extent that it leaves a track like that of a demented moth around a candle flame… If you’re using a laptop or PC at the lectern, then use the mouse pointer if you want to draw attention to a particular point.

  • Some good points.

    One of the ‘sins’ I’ve seen is to have too many slides (similar to Alison’s crammed slide problem). I won a presentation award at one conference where I used just 2 slides to make my presentation. The rest of the time it was keeping the audience interested and engaged.

    Start by making the audience think what you are doing is cool, interesting and relevant. It’s a bit like a marketing exercise. The goal is to get them interested, give them a taste for what you’ve done, and provoke some discussion afterwards.

    Oh, and enthusiasm really counts.

  • If you know in advance what you intend to highlight, you can try build that into the presentation slides. (You’ll want to balance this with getting too clever and fiddly.)

    I agree that people reciting presentations are terrible: best that text on “slides” be short points, with the speaking fleshing it out.

    (I once went to a lecture by a very famous scientist, only for him to look down at the lecturn and read aloud for the best part of an hour. Terrible, truly awful. By contrast another very famous scientist gave a plenary at a large conference using a single blank sheet of overhead transparency as his only prop, drawing on it as he went along, building up the picture he wanted to end with. I think he did it to make a point abut presentations and I don’t think many could pull it off.)

    This reminds me: speak to the audience. As someone who is hard-of-hearing, this bugs me no end. You’re talking to the audience. Face them, lift your head up and speak to them, not the screen or the laptop! You’ll need to know your material well enough to do that, but that’s part of the trick.

    (On the same theme, really the speaker should visible, not hidden in the dark, but many rooms are not set up properly for that.)

  • During the initial interaction, start by thanking the audience for turning up, and also your host/organiser for the introduction ( even if there isn’t one! ), If you want to go back to a slide during presentation or questions, only do it if you can jump directly to it, otherwise just talk directly to the questioner/audience, and suggest they review the slides later when the proceedings are available..

    Have an interesting slide to display for the start of question time – it could be unusual, or it can point to resources for further reading, but don’t use one that you’ve already shown. That way, if you don’t get any questions, you can quickly finish without a protracted silence, and always remember thank the audience for listening, and there may not be such a rush for the door as soon as you stop talking.

  • I confess that I am one of those speakers who forgets their acknowledgements slide at the end. I’ve always been reluctant to put it up front though, because I worry that it distracts the audience from the introduction (particularly if you have a lot of people to thank!). As Rebecca says, putting those key questions to the audience at the start can be very effective.

    One thing I might try in future is to put my acknowledgements just before my conclusions slide.

  • That may be because you have more people to thank than a student! That happens to more senior people… 🙂 Seriously, I hear your point. It can make the first slide very “busy” if you’ve got to many people to acknowledge.

    I like Rebecca idea, though, it gives the audience something to think about while they’re waiting.

  • While it’s in front of me, there’s a short article “How to give a good talk” in Molecular Cell Alon 36(2)165 (2009).

  • The acknowledgement slide at the end of a presentation is one of those things I hate. It completely distracts from the end of a presentation where you should be driving home your important results. Also a “ANY QUESTIONS?” slide is just as bad. You should be leaving up your conclusions for all the see.

  • A great outline of key points to make talks interesting.
    I was curious about your comment of putting dark colours on a light background. I have seen some great talks using that arrangement but I have also seen some great talks with light (white/yellow) colours on a dark (black/deep blue) background and tend to use this myself. I was wondering what your (and others) opinion is on this approach?
    Also, something my PhD supervisor taught me was to have a few extra slides at the end explaining key concepts that people may ask. Although one has to be careful not to make it look too contrived.
    I think your point about ownership is very important. Too many students forget that when they study an area in detail they become the expert in it.
    And knowing your audience is vital. This can be one of the problems in public lectures where a poor understanding of the audience results in confusion and frustration for the audience.

  • One of the best talks I have seen, was one where the lecturer acknowledged everyone at the beginning of the talk. It was done in terms of “I will be talking about subject X, which was carried out by students a, b, and c, and subject Y which was carried out by students b, c, and d.
    Another method, which I have used, is to acknowledge people as you talk about areas where they were involved. You can put their names on the bottom of the slide to remind you, and if a publication was produced you can list this as well.
    There is something really great about a lecturer who sincerely acknowledges the contribution of their students and colleagues, and when it is done with amusing little stories about how the student made a breakthrough or successfully thought outside the box or invented a new piece of equipment it really brings the talk alive.

  • Yellow on blue has issues for some colour-blind people, by the way. Just a loose thought.

    I’ve always preferred dark backgrounds, but after I was approached after a talk by a colour-blind person I’ve had second thoughts on that and am still wondering what the best solution is. Something I keep meaning to look up.

    It is easy to forget you become the expert of your wee niche when you’re a student. In my Ph.D. I became “the” expert on the sequences of CCHH zinc finger proteins and whatnot (in part simply by default admittedly!), but it was hard to see that at the time.

    I prefer the approach of starting with introducing the team myself, and placing credits on the slides when it’s someone else’s work. (Ditto for placing reference citations at the bottom for illustrations of other published work.)

  • drmike: My comment about always using dark colours on a white background was intended for less experienced presenters. I’ve seen some disastorous colour choices from inexperienced presenters (and sadly a few from people with plenty of experience) over the years, so my advice is to play it safe until you have had more experience with how different colour schemes work in different environments.

    As for introducing your team at the start, I always list my co-authors on the title slide and explain who they are and what they have contributed. However, many of us will have funding agencies, user facilities and larger collaborations that need mentioning. These I leave until the end (and unfortunately forget about a third of the time).

  • Ah yes, one must never forget to acknowledge funding agencies 🙂
    I guess with practice, one must start remembering to show the acknowledgements slide more often?