(Here’s a blog post I wrote for the Future Faculty Leadership Council of UAlbany  — http://uafuturefaculty.blogspot.com)

Being comfortable with communicating your ideas is absolutely essential in your field and frequently important in your private life as well. Over the course of your career, you may be asked to give hundreds of talks to colleagues or laypeople. Usually, the audience you will be speaking to has never heard of you, has never read your papers, is unfamiliar with the terminology or methods of your work, and is woozy from not having enough coffee on that particular day. Therefore, you should plan for this ahead of time and craft a captivating presentation that will leave everyone in the audience eager to read your paper.

Typically, presentations made in professional settings (in 2013) require either the use of Microsoft Powerpoint, its free equivalent, Openoffice Impress, or using a .pdf  file (Portable document format). The style and flow of the presentation you create should have synergy with every word that comes out of your mouth. The following tips on giving a good talk are not exhaustive, but should be a good general-purpose guideline as most researchers were never trained in how to give these talks besides ‘just doing it’. Additionally, some of these tips can be applied to improve your classroom lecturing skills.

Communicating your research to a professional conference, a colloquium in your department, at a meeting within your own group, and to grant reviewers all require a different approach. While you are preparing for your talk, ask yourself who you are presenting to, as this will affect the focus of your presentation. If you are giving a talk to a broad audience, take special care to include enough introductory information that a layman can follow you. For a departmental or conference talk, the main focus should be your data and methods. For your dissertation or a job interview among people who already know your research well, the discussion of your results is of prime importance. Your motivations and goals for your research should be crystal clear no matter who your audience is. Remember that your papers are the main events in your research career, the presentation is just an advertisement for a single paper. The purpose of your talk should be to give your audience an intuitive feel for your idea and to engage, excite, and provoke them to read your body of work.

Speaking clearly with confidence comes easier to some than others. People who are performers in theater or music may have shed their ‘stage fright’ long ago, and it can certainly be intimidating to speak to a large audience regardless of experience. Rehearsing your talk in front of a mirror may help get your timing issues straight, but it’s just as important to practice giving your talk to actual people who can give you feedback on grammar, style and flow even though they might not have a clue what your research is about. The most important thing is to be natural, in the moment, and know what you are doing. Remember that you know this research topic better than anyone else on the planet!

You are in control of the room when you are speaking and if you act cool, composed, and enthusiastic, your audience will follow what you’re saying better and not lose interest as easily. Don’t over gesticulate and whatever you do, don’t overuse a (laser) pointer. A pointer should be used very sparingly to indicate trends in figures or to quickly show your audience the specific item you’re talking about on a slide. It should never be used to rapidly circle anything or point to words as you’re saying them like a bouncing ball on a karaoke machine. You should not read sentences off your slides, rather, the words should be a carefully-chosen guide to sum up the main ideas that you’re talking about. Speak to your audience (not your slides!) and shift your focus to various people seated. Avoid saying “uh” or “um” or doing things like touching your face and yawning. Plan on using about one minute per slide. Use a professional, readable font- not comic sans!

A picture is worth a thousand words as they say, but the art of displaying your data in figures is highly nuanced. In general, the figures you display for your audience in a talk should be of publication quality. For example, when plotting a relationship quickly between two variables in a study to discern a trend for yourself, a group member, or perhaps your advisor, it may be acceptable to plot the data in Microsoft Excel with all the default settings. However, this should never be done when presenting your research to a broader audience. The reason is that the defaults in Excel tend to obfuscate any meaningful trends in your figure by being visually distracting. There is seldom any need for any background to a figure besides a blank white background and you do not need horizontal lines that span the width of the figure. You end up needing to do a ton of pointing and clicking to arrange your data right, and unless you program a macro, your time would be better spent using a different program (Origin or the python library matplotlib come to mind). Don’t forget error bars if you took several different samples and averaged them. Make sure your axes are labeled with units, your color schemes make sense and that everything is readable. Also. avoid plotting too much information in a single figure. More than four or five data series on a single plot can be visually distracting unless you have a clever way to present them, like adding them one at a time, and the same can be said for more than two figures per slide. Finally, make sure your slides are visually balanced (that is, avoid excessive whitespace) with figures and a small amount of text with a take home message.

If you use anything in your presentation that you yourself did not create, you must cite where it came from. This might be prior work by others that serves as motivation for your research, specific theoretical details that you did not derive, or any pictures, maps, or graphs that came from somewhere else. If you make your slides public, having these citations as hyperlinks at the bottom of the slide in which you reference each bit of information could be helpful. Either at the very beginning or at the very end of your talk you should acknowledge your group, your advisor, and any collaborators or technicians with their affiliations. Having pictures of everyone is a good idea to keep your audience engaged and to jog their memories. It’s also critical to tell your audience where you got the funding to do your research. The individual grant ID numbers are probably too much information, but the agencies themselves  are essential.

As sure as the dawn, you’re going to have raised a few questions in your audience when you finish your talk. You may have questions during your talk and they should almost always be welcomed, so be sure to periodically scan your audience. Questions are a golden opportunity to connect with your audience and further scientific discussion. Be quick and concise when answering questions. You may have backup slides with technical details or additional data in response to some questions if you anticipate them. In the case that someone asks a question that you cannot answer, it is OK to admit your ignorance and tell the person, “I have not looked into that, perhaps we can discuss it later” or something of the sort.

Sources:

K.A. Dunn, “How to give a better technical talk,” guest lecture, CNSE Student Seminar Series (Albany, New York: College of Nanoscale Science & Engineering, University at Albany – SUNY: March 27, 2010).

http://research.microsoft.com/pubs/67052/giving-a-talk-slides.pdf(although this presentation breaks a cardinal rule: use a readable,professional font)