Tips for Better Presentations

  • Be sure to motivate the talk by describing the problem you are trying to solve
  • Consider your audience. If you are addressing interdisciplinary researchers about an esoteric topic, help the audience get oriented by starting out with the big picture.
  • For a short talk, don't spend time reading through the outline of what you are going to present
  • The best presentations use no slides and ask the audience to concentrate on the speaker instead of a screen; keep lights high to increase the alertness of the audience
  • The best presentations use handouts instead of slides, and the best handout is a paper. The next best handout contains the key figures, tables, equations, and references. Put your name and contact information on the handout.
  • If you need to use slides, try to confine them to slides containing figures, tables, diagrams, and key equations; at any rate aim for fewer than 45 projected slides per hour
  • If you use slides, avoid fading and other transition devices, and avoid progressive revelation of bullet points. What's wrong with progressive revelation of bullet points?
  • If you use a handout, do not lead the audience through the handout but just point out key figures, etc.
  • It is OK to move back and forth between slides but try to limit the number of times you do this as it is disorienting to the audience
  • Make constant eye contact with the audience; do not face the screen if there is one
  • Make sure that if you use a laser pointer that the pointer is on no more than 30 seconds per hour of presentation
  • Don't hold anything as a prop
  • For statistical methods presentations it is a good idea to
    • Provide a handout giving all critical definitions used in your talk. Few of us have photographic memories and none of us are as familiar with your topic as you are. If you flash the definition of f(x) for 15 seconds at the beginning of a talk, it is unlikely that we will remember it 15 minutes later. If remembering what f(x) means is vital to understanding what you have done, then it is a good idea to provide a crib sheet defining f(x) that your audience can refer to as needed.
    • Analyze a relevant real data set using a standard or simple approach. Then reanalyze the same data using your method. Explain and demonstrate why your method does a better job than the standard approach.
    • Some members of your audience may be confused by some of your technical details no matter how carefully you explain them. However, by the end of your talk all of your audience should be able to answer the following questions.
      • When should we use your method?
      • What are its strengths?
      • What are its limitations?
      • What are its a priori assumptions?
      • Why is it better than other available methods?

Some of these ideas are from Edward Tufte.

-- FrankHarrell - 27 Feb, 15 Dec 2004, 13 Aug 2010
-- WilliamDupont - 16 Mar 2004
Topic revision: r12 - 05 Nov 2012, JoAnnAlvarez

This site is powered by FoswikiCopyright © 2013-2017 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback