Tips for giving research presentations
A presentation is not your paper in slides
A presentation is more constrained and more straining than a paper. The audience does not have the luxury of pausing, looking things up, and re-visiting the same passage. Time will not enable you to talk about everything. You need to focus on the story. In the case of a lightning talk (lasting as little as one minute), only the story can remain. Shorter talks should focus on making the audience curious so that they want to talk to you (or visit your poster).
You need to anticipate the audience’s background and adapt. If you have a hierarchical structure in your paper, your life will be easier. Otherwise, you need to capture the essence. That is, you need to condense your paper into figures, illustrations, and tables.
Text should be minimal on the slide. It can signal carelessness ("I didn't care to prepare well"), and if you force your audience to read a lot, they won't listen to you. The corollary for online talks is that since people can drop out due to connection issues (their attention is lost, which is easier when you have Zoom open as the tenth tab on our laptop), then a bit more text helps them to follow. But keep it minimal. The slides are not your notes, they are your supporting material. Show what you cannot easily tell.
The same holds for the design. First, graphic design is its separate profession, so don't assume that you automatically will know what works. I like Garr Reynolds' Presentation Zen. I have listened to presentations with amazing content but ugly or "didn't care" styles (plain white with black text). However, it is better to commit this mistake1 than to have a superb design but shallow content. Still, I believe that slide design can help distinguish yourself—at least, it undoubtedly shows the care taken by the presenter.
The number of slides should be minimal. Use 1 slide/minute as a rule of thumb. It is trivial, but I will say it nonetheless: get feedback on your slides (you know, curse of knowledge), and practice, practice, practice—and time yourself.
Regarding time, be prepared for cutting short your talk. Some audiences, especially in North America, tend to ask many more questions, e.g., compared to Europe. If you are not used to this, it can feel intimidating, even though most questions come out of curiosity. Glitches in technology happen frequently, the schedule might change.
Anticipate the questions. If you have additional cool results, put them into backup slides (i.e., ones you will not walk through your talk). It is also impressive when for a question, you can pull out your backup slides.
Not everyone agrees that this is a mistake