Tips for Scientific Presentations

Here are some tips that I think are useful. It's a lot to absorb, but that's because there is a lot to think about when giving a talk. With practice, much of this will become second nature.

• Content.
  • Be sure to hit the main points. What these are will depend on your talk, but you want to think about questions like: What is the motivation for your subject? What are the main assumptions/observational constraints? What are the main theory input/results? How well do theory and observation agree? And be especially sure that your talk makes clear how your subject fits into the material of this course. That is, make clear the answer to: How do your results fit into the big picture of your field, and of astrophysics generally?
  • Hit the right level. This is a difficult art, and not everyone would agree about what the right level is, though clearly it depends on the audience.
    The difficult but important balance to strike is that you don't want to give extraneous detail, yet you don't want to oversimplify to the point of having no content. In thinking about this, ask yourself: what are the most important results? Then ask: what are the key steps leading to these results? Then ask: what is the clearest and simplest way to present these key results.
  • Give physical intution whenever possible. Simple physical arguments are invaluable in helping people to understand and remember the results you present. Pictures/sketches can be very effective too.
• Organization.
  • Make sure your talk has a logical flow. If you need point B to understand point A, then say point B first.
  • Carefully craft the structure of your talk. I find it helpful to think of the "hourglass" model--i.e., moving your focus from broad to narrow then back to broad. Namely: relate your question to the big picture of astrophysics, then give the gist of how you actually go about answering your question, then relate the answer back to the big picture and tell us what new things we have now learned (and any remaining open questions/predictions).
• Presentation.
  • Speaking. You want to speak clearly and loudly enough that people can easily hear you. Look out at the audience whenever possible. Don't speak in a monotone--use inflection in your voice to make clear what points are important. Don't be afraid to pause, or to say "the following result is very important."
    Also, practice the transitions between your slides. This is often a source of trouble for people. You want to make your links smoothly and without too much wordiness.
  • Choreography. While you are striving to make your message clear, remember that you are opaque: don't stand in the beam of the projector. It is good to point to your slides, to help people follow, but do use a pointer, and stand to the side, so that everyone can see what you are pointing to.
  • Slides. It is critical that your slides be neat and legible. Make sure the words and figures are big enough so that everyone, even those in the back, can see them. Practicing your talk provides an excellent opportunity to test this.
    Be very thoughtful about how much you write. I am a firm believer in the philosophy that you want everyone in the audience to understand every mark on your slides. A corollary to this is that you should use great care in showing other people's plots. Often these have much more information that you need to convey your point. If the data is readily available, the best thing is to replot it, so you can present it exactly the way you want, with the labels you want. The next best thing is to make your own marks on the plot to help guide the viewer's eye; for example, you can underline/highlight axis labels and perhaps interpret them, as in "baryon density" or "abundance." If there is more than one curve, you can highlight the ones that are important.
    I am also a firm believer in taking advantage of the full range of the visible spectrum. Monochromatic talks tend to be harder to follow, as color can help you indicate what is imporant, as can underlining, or "highlighting" in yellow (in which case you should recall that a transparency has two sides you can write on). Of course, this can be overdone; too many colors can be distracting. I personally believe in two main colors per transparency, with maybe one more if there's a need (e.g., for highlighting).
  • Timing. It is exceedingly bad for a talk to run overtime; this quickly annoys the audience. Thus, you must be sure that you can do your talk on time. Again, practice will help, as will some forethought as to what you will cut if you find you are running behind.
• Citations and Academic Integrity: Your talks and papers are not required to be original research, and thus all of what you say will draw on your rest of references. You must acknowledge this explicitly and clearly. In particular, you must of course follow the rules of academic integrity, which among other things states that
  • "Every direct quotation must be identified by quotation marks or by appropriate indentation and must be promptly cited in a citation."
  • "Prompt acknowledgment is required when material from another source is paraphrased or summarized in whole or in part. This is true even if the student's words differ substantially from those of the source."
  In a paper, it is clear how to do this (and if it is not, talk to me!). However, in a talk--particularly a short one like this--it is impractical and not useful to flash up a long, full bibliography which nobody can read. Thus the nature of citations becomes more difficult for a talk, but it is still your responsibility to make clear what sources you used. Some things you can and should do:
  • Be sure to explicitly cite any direct quotation. Typically, you will be listing bullet points which probably are not even in complete sentences, so you probably should not even be using direct quotes--but you must cite them explicitly if you do.
  • As you mention key ideas, you can cite them in shorthand by writing on your overheads the names, and possibly years, of the key authors involved. For example, if you are discussing neutrino counting in big bang nucleosynthesis, you would write "Steigman, Schramm, and Gunn 1977". You need not necessarily say every citation out loud, but have them there so that people can find them. Similarly, to help the audience know what is a citation and what is part of the text of your transparency, you might choose to put the citations in smaller print and/or in a different color than the main text.
  • When you present images, figures, tables, or diagrams, you must make it absolutely clear where you got them and/or the information they contain. Even if you create your own figures based on somebody else's data, you must indicate where you got the data.

More tips:

• Astronomy 134 is not quite Carnegie Hall, yet still: practice, practice, practice! In front of anyone who will listen (officemate, significant other, pets, etc.). Bribe them if needed. You will be amazed (I always am) at the difference between running through the talk in your head and saying it out loud. For example, you are much more likely to discover any holes in the logic or flow when talking to someone who has not been thinking about the subject as long as you have in the way you have.
• Use equations sparingly. It is often possible to give a whole talk using only a few or one or even zero equations. If you must equate, keep you notation as intuitive and simple as possible.
• Finally, I find it very helpful to become a "student of talks." That is, when I am at talks (colloquia, seminars, conferences), I pay attention to what works and doesn't work. If I like a talk, or hate it, I ask myself what about the talk leads me to react this way, and I try to learn from this.

Second Opinions

You will find that not everyone would agree with all of the advice above. For other points of view, I commend you to the web:

• Suggestions for Giving Talks , by Bob Geroch. I find this to be extremely useful advice, and it informs much of what appears above.
• Official APS advice
• Official AAS advice
• Qualifying exam for conference speakers