Giving Scientific Talks

Things To Do

1. Make a title page that includes your name, department and institution, and the date.
2. Plan your talk as if you were speaking to a physics audience unfamiliar with your work. Typically, you should describe (1) the phenomena you are interested in studying, (2) the apparatus and techniques used to make the measurements, (3) the results obtained and (4) what the results mean. Be sure that what you say is correct and focus on the physics.
3. In general you will have lots of results and will have used a variety of techniques to analyze your data. Determine what must be understood to appreciate your results and focus your talk on those aspects of your study. You will not have time to cover everything so focus on about three things.
4. Talks should be self-contained. Provide all information necessary to understand your presentation - tell them everything they need to know and nothing more. (See No. 1 and 2 below.)
5. It is a good idea to put all items you want to discuss on a viewgraph.
6. Make sure your plots and drawings are simple and correct. You might not want to include everything because that tends to make the sketch cluttered and hard to follow. It is OK to make a picture that is not literally correct, if it is used for the purpose of instruction and explanation of the essentials. Provide all the important features.
7. Provide visual aids on your slides. For example, if you want your audience to see that your data falls on a specific curve (straight line, quadratic, etc.) draw such a curve through the data. Do not ask your audience to imagine these curves because they will probably not imagine what you have in mind. In some cases, you may want to show data without such a curve first. With PowerPoint you can always show a slide more than once, or animate it so that the curve fit appears at a later time. Each time you can make modifications tailored to direct the audience to the point you are making.
8. Always, Always, Always be prepared to answer questions about how you obtained a specific result.
9. If you want your audience to believe your result, you must believe your result. Be UPBEAT! Show some excitement about your work.
10. Regardless of how difficult the subject matter, the best talks always leave the audience with a clear understanding of what was done, how it was done and why it was done.

Things Not To Do

1. Do not assume your audience knows your experiment or your instrumentation.
2. In general, your audience will not have read all that you have read nor have they thought about your experiment as much as you have. Do not assume they know what you know.
3. Do not refer to the manual or other papers and documents. Explain and/or show the equations necessary to understand your experiment.
4. Do not present the material as if you were reporting back to an employer who told you what to do. Your tone should not be one of following a recipe in the lab manual. That is, do not sound like you are giving a report in response to a request. You are learning. Make it sound like you’ve just discovered something new.
5. Do not go through all the gory details of your analysis. Focus on the final results and what they mean.
6. Minimize the discussion of equations. Present only the equations required to understand what you will discuss. If you show an equation explain what it means physically.
7. Never put anything (an equation, data, statement, etc.) on your viewgraph that you do not understand. You must be able to explain and discuss everything on your viewgraphs!
8. Do not try to put too much on one slide. Do not include large paragraphs; sentence fragments and short phrases are best.
9. Do not try to give all the results in a short 10 - 15 min talk.