How to give a good talk and avoid common mistakes

Recently, I became aware of a large amount of literature and tutorials online about how to prepare proper slides for talks. While I was paying attention on certain aspects of my slides previously, it occurred to me that there were still many crucial problems in the slides I prepared so far, which must have unnecessarily hindered people in following what I was trying to say. My two ILQGS talks from 2013 and 2015 serve as a good example of the different approach to preparing slides, the 2015 one was my first talk prepared taking into account some of the points that I will discuss below.

In general, many of tips on preparing professional power point presentation can be taken over verbatim also for the scientific presentations that we are interested in. However, there are some peculiarities which scientists should pay attention to, and I will split the tips accordingly. Some of them sound very basic, but they are ignored far too often. In my point of view, it is a sign of respect for your audience if you prepare good slides, or at least avoid the most critical errors in doing so. An audience that feels respected will be more open to what you have to say, while disrespect in form of a poor presentation or unreadable / confusing / eye-straining slides will certainly not further your cause.

General tips:

• Before preparing the presentation, think about what central message you want to convey. Develop your presentation around this central message. Your aim should be that your audience remembers at least this message. 
• Don’t put long texts on your slides, especially don’t simply read text that is on your slides. In fact, put as little text as possible. What happens if you put long text is that your audience will read the slides and at the same time not listen to you. They will be finished reading the slides before you are finished talking, and then have missed what you were saying.
• Choose an eye-friendly colour scheme. Black on white is not eye friendly, the white background will unnecessarily strain the eyes of your audience. Instead use for example a light grey, brown, or orange tone for your background, and some dark, but not completely black, font. Or invert this choice. A classic is for example a blackboard background. An appropriate choice of colour scheme is a one-time investment in your presentation template for latex beamer or similar, well worth the investment. 
• Test your presentation on a projector. Colours may look different than on your computer screen, and e.g. two colours that you used to distinguish some parts of a figure or formula might look exactly the same. 
• When giving your presentation, make short pauses after logical sections of your presentation end, in particular after every slide. This allows your audiences brains to rest for a second and process what you have said. 
• Use pictures and sketches. 

Scientific presentations:

• Target your talk to your audience. Don’t use a talk that you prepared for the specialists in your field for a general physics audience. They will not appreciate it. In preparing your talk, always think of the background that your listeners will have and adjust the technical level accordingly. 
• Repetition is your friend. The topic of your presentation may be complicated, and there is a high chance that not everyone will understand what you are trying to say the first time. Therefore, repeat it in different words at different points of the presentation. The more important the statement is, the more often it should be repeated. Your central message from above should appear in the introduction, the main part, and the conclusion.
• The classic: avoid light green or yellow on white background. It will be almost invisible also in your presentation. 
• Put the slide number on your slides, preferably in some corner of the page and in some non-eye-catching, yet visible colour like 50% grey. People may want to refer to a certain slide when they ask questions. 
• Besides the slide number, don’t put any other recurring information like your name, institution, seminar title, date, etc. on the slide. While the latex beamer templates mostly include such a possibility, avoid it. People will start reading these things during your presentation and stop listening to you. Also don’t put the total number of slides, as people will e.g. start thinking about whether you will make it in time or not. The same is true for talk outlines. It is perfectly fine to have in-between slides where you say where you are right now in your presentation, but this should not be on every slide. Same reason as above. 
• Avoid formulas wherever possible. Whenever you need to use a formula, simplify it as much as possible, even if this means dropping numerical factors or index structure. Remember that if you put something on your slide, it should be important and you convey to your audience that they should grasp it. Always explain to your audience why a formula that you put on your slide is important and what they should remember from it. It may be that some formula is just for an important illustrational purpose, e.g. to show that something can be computed in principle. In this case, say so, and your audience will not waste time trying to understand the formula further, while not listing to you. 
• After you are done with your presentation, go through it again and try to remove unnecessary text and formulas. 
• Clearly explain axis labelling in plots, explain what is shown and what your audience should look for in the data. If your audience is not familiar with certain plots, they will be lost. 
• Rehearse your presentation. You will find better and more efficient ways to explain something once you tried it a few times. This will clearly improve your talk. 
• Know your timing and be one time. If you rehearsed your talk properly, you will know how long it will take. 

I tried to take these tips into account in my last two presentation here and here. If you have any suggestions on what could be improved, please comment!

As a selection of other interesting tips and tutorials, I suggest:

• Ten Secrets to Giving a Good Scientific Talk by Mark Schoeberl and Brian Toon
• Effective Science Talks by Celia M. Elliott
• Suggestions For Giving Talks by Robert Geroch
• Death by Powerpoint by Alexei Kapterev
• Life After Death by Powerpoint by Don McMillan