Notes on Fisher Files, Sequence II, Episode 7 - Giving a talk on your work

Here are my notes on the Fisher Files episode entitled "Giving a talk on your work."

• Why are you giving this talk?
  
  • Affects how you prepare
    
  • Affects what visuals you use
• What messages are you trying to convey?
  
  • Is it a job talk, conference talk, colloquium, or popular talk?
• Don't automatically just give any talk that any old person asks you to do. Like travel, there is a major time commitment.
  
• General advice about talks
  
  • Most important tips
    
    • Talk needs to be interesting!!!
      
    • Talk needs to be informative
      
    • Have a conscious idea of what your message is
      
    • If you don't have a conscious idea of what your message is, you don't have a clear idea of what will come out of your talk
  • Logistics
    
    • Don't go over time or rush at the end
      
    • Target 50 min talk for 1 hour time slot
      
    • If you have a movie, make sure it works on the actual projector. The lighting could be bad, the projector might not be compatible with the movie.
      
    • Bring hard copy of the talk
      
    • Setup computer a minimum of 15 minutes before the talk
      
    • Don't use transparencies that were obviously written on the plane (that insults the audience)
      
    • Have a PDF of your PowerPoint/Keynote presentation as backup
  • Message/scope
    
    • The number of big ideas should match the amount of time you have
  • Structure of talk
    
    • Find a way to make your talk interesting
      
    • Make sure you talk about things you are interested in
      
    • Don't just give a cookie cutter talk
      
    • Audiences can tell if you are truly interested in your talk
  • Style of talk
    
    • Use more pictures/graphs and less diagrams, bullet points, and equations. It makes research more physical, gives deeper sense of complexity.
      
    • Charts and graphs are better than tables and bullet points
      
    • Face audience
      
    • Don't use bullet points, just use keywords and talk about them
      
    • Don't just read off the slide
      
    • Don't feel compelled to explain every detail (most of them are boring anyways)
      
    • Emphasize the details you find interesting (make the talk about you)
      
    • Don't forget to give credit: just mention everyone you worked with. Be generous with credit since it's not good to make enemies and besides it's honest.
• Different types of talks
  
  • Thesis defense
    
    • Demonstrate that you have mastered your field
  • Job talks
    
    • Show how you fit in the department
      
    • Show your analysis skills
      
    • Show how capable you are
      
    • Illustrate deep thinking about physics
      
    • Don't just expand a conference talk
  • Conference talks
    
    • If you are going to give a 10 minute talk, only have one idea
      
    • You have less scope for making it interesting
      
    • OK to use lots of words and put in lots of results that you don't address in your talk because likely your talk will be publicly available
      
    • Proceedings (if there are any) should be an accurate representation of your talk; any ideas that occur after the talk should be footnoted
  • Seminar
    
    • For a seminar (approximately one hour), use one theme and 5 sub themes
      
    • Prepare a 50 minute talk for a one hour slot
      
    • Encourage audience to ask questions during the seminar and not afterwards
      
    • Don't need to show what you did as opposed to what other people did (unlike job talk)
      
    • Can assume audience is well-grounded in your field
      
    • Be a little conservative in assuming what everyone knows (use a few introductory slides)
      
    • Introduce yourself: name, who you are, level, where you are from, and who you work with
      
    • Don't use a cookie cutter structure
      
    • Start generally and narrow focus in first 10% of talk (first 5-8 slides)
      
    • At end of talk (last 10% or 5-7 slides), open up the focus again
      
    • Don't just stop after showing your result, explain why the result is important in the larger context
      
    • Bring closure to talk
  • Colloquium
    
    • Important to know your audience
      
    • Similar to seminar: start broad, focus in and then focus out, *but* do all this more slowly
      
    • Could be a job talk for a senior position (in this case, an assessment of how well you communicate)
      
    • Much less detail than a seminar
      
    • More about what you do, than how you do it
      
    • More pictures, less equations
      
    • Stories helpful (e.g. how the discovery took place, who the people were)
      
    • Need to work much harder in colloquium to engage the audience than in a seminar
      
    • Don't expand a seminar into a colloquium talk; it's easy for the audience to sense (then the audience becomes bored)
      
    • Colloquium much more formal, typically questions are reserved for the end
      
    • Opportunity to reach out to people outside your field and inform them
      
    • Pare down your ideas
      
    • Only communicate the simplest ones, rather than trying to conveying the deep elegance that only a few people understand
      
    • Don't rush
  • Plenary talks
    
    • Broad in scope
      
    • Even more important to give credit
  • Popular talks
    
    • Even more emphasis on explaining clearly, using pictures, and finishing on time
      
    • Important to practice, maybe even video tape yourself