Advice on Giving Talks

"At the first stop of a tour in Japan, Albert Einstein gave a scientific presentation that, with the accompanying translation, lasted four hours. Although his audience appeared to be attentive the entire time, Einstein worried about their comfort and decided to pare back the presentation for the next stop on his tour. At the end of the second presentation, which lasted two and a half hours, the crowd did an unusual thing in Japanese culture, particularly in that era. They complained. For Einstein, though, the complaint was a compliment - this crowd had wanted him to deliver the longer version.

When was the last time that you sat through two and a half hours of a scientific presentation and wished that it would go longer? Unfortunately, such responses to scientific presentations are rare. Granted, Einstein was a brilliant scientist, but just because one is a brilliant scientist or engineer does not mean that one is an engaging presenter. Consider Niels Bohr, the great physicist who won a Nobel Prize for his proposed structure of the hydrogen atom. Despite being an inspiration for many physicists, Bohr had difficulty communicating to less-technical audiences. For example, his open series of lectures in the Boston area drew progressively fewer and fewer attendees because "the microphone was erratic, Bohr's aspirated and sibilant diction mostly incomprehensible, and his thoughts too intricately evolved even for those who could hear". "
- Michael Alley (From "The Craft of Scientific Presentations")

One of the jobs that I regularly do as a scientist is to give talks and to listen to the talks given by others. Right at the start I would like to clarify that I am no great speaker. However, my experience is that half of the scientists give terrible talks, and I do think that there are obvious things that every speaker should and shouldn't do which may not make your talks great, but at least will make them reasonably good, and will certainly save you from giving talks that your audience would term bad. Here are my tips on giving good talks (or rather on avoiding giving bad talks).

  1. Ten-times Rule You should put large time and efforts in preparing a talk. This includes deciding the content of the talk, preparing slides, experimenting with different ways of presenting the content, writing down at least some parts of the talk, rehearsing it to see which parts are particularly difficult to explain, and so on. I personally use 'Ten-times Rule' by which I mean that if I am going to talk for 30 minutes, I at least spend ten times that much time (5 hours in this case). This may sound too large to some, but trust me this is a conservative estimate. It is better to spend time preparing a good talk instead of spending more time regretting after the talk.

  2. Relate to the audience One of the biggest mistakes that I have seen most speakers making again and again is to assume that the audience is just like you: they have studied the same subject, they appreciate all the problems and excitements related to what you are presenting, they know all the techniques of your field and so on. Perhaps an extreme analogy would be to ask Albert Einstein to teach the General Theory of Relativity to kids who have just learned how to add numbers. The fact that the theory is extremely beautiful won't help, these children are not going to appreciate it. But this is not just about the age. A subject like Physics (which I work in), is itself so large now that it is impossible for any one person to study all of it. It is divided into various branches, sub-branches, and sub-sub-branches and even experts from a particular area sometimes need to spend time on understanding a concept. A faculty member from my department once gave a talk for other members in the department and after the first slide, started talking in the terminology specific to her field. This inevitably forced me to ask the meaning of the terms she was using, and she made the situation worse by asking me back how come I don't know these. Although I have a Physics degree, surely I have not studied all of it! A worse incident happened in my department when a former student started giving a talk on his Ph.D. thesis assuming that we all were experts in his exact field. The first slide in his talk showed a complicated equation which he announced that it was 'well-known'. You can very well imagine that I (and surely nobody else in the room) didn't even get the gist of the talk.

  3. Pay attention to the body language I have observed some speakers talking to themselves instead of the audience, sometimes mumbling to the extent that you can't even hear them in the middle row of the hall. A related habit is to say at the start of the talk that you don't need mic because you can talk loud. This is almost always false. If you do use mic, you don't have to scream unnecessarily, and also you can make sure that the words you utter are heard by others. Not making eye contact is another habit that can make an impression of diffidence. A friend of mine has habit of rubbing his forehead while talking which makes the audience feel that he is nervous. A good idea is to ask your close friends about this because nobody else would dare to tell it to you.

  4. Don't run overtime There are three principles about this. First, don't run overtime. Second, never run overtime. Third, never never do that. Some speakers think that if they just tell you somewhat more, audience will get impressed. If you couldn't impress the audience in 45 minutes, surely you won't do that in two extra minutes. In fact, the chances are that probably you will annoy them, and they may even forget good things you said in the talk. Some people might have come to attend your talk assuming that it will end at 2:55 pm so that they can can catch that 3:00 pm bus, but if you keep speaking after the allotted time, they may not go out of the room just out of modesty but will probably never attend your talk again.

  5. Prepare good slides My usual experience about the talks that I have attended is that either the speaker hasn't properly thought about the content of the slides or has thought more about cosmetic issues. It is best to tell your inner self that making letters fly off the screen when you hit the 'next' button doesn't impress the audience. Your slides should serve as an assisting tool for your talk, and should not be a representation of your artistic skills. Another mistake is to make the slides 'content-rich' by squeezing as much as possible in the slides. Audience would typically get a minute to look at the slide. If you have filled it with lots of small pictures/images, equations, lot of text, people are only going to get irritated and bored. If you think that people will admire that you know a lot about your subject, you are fooling yourself. Even a genius mathematician will not be able to understand a set of complicated equations squeezed on a single slide. Thus, refrain from this, let there be blank spaces in slides. This is also soothing on eyes.

  6. Don't stretch Question-Answer session Some speakers, especially professors, have this habit of answering questions with too much details without thinking whether the person who asked the question was asking for so many details. Worse, some speakers would just keep discussing with somebody from the audience who happens to be from their field. Others in the room, instead of admiring the knowledge of the two great souls, would just have resenting feeling and ultimately would hate you.

  7. Respect the audience Finally, there is something that (too) many speakers don't think about: respecting your audience. All the points above are ultimately related to respecting people who decided to spend the next hour listening to you instead of watching a movie or chatting with a friend in a coffee shop. A typical scientific talk will have around twenty people attending it. If your talk runs for an hour, you are effectively asking for twenty hours of collective time of others. Thus, if your audience doesn't learn anything new after attending your talk and doesn't enjoy it, you have simply wasted their time and insulted them! As other points here explain, this could happen for a number of reasons including your lack of preparation, not relating your talk to the audience, your body-language, running overtime etc. So respect your audience and give great talks.

Other resources

Back to Home