Question? Leave a message!




How to give a scientific presentation

How to give a scientific presentation 17
How to give a scientific presentation Karl W Broman Department of Biostatistics Medical Informatics University of Wisconsin – Madison www.biostat.wisc.edu/kbromanNote This is not a scientific presentation 2Note This is not a scientific presentation What I say is the truth (but not everyone will agree) 3Note This is not a scientific presentation What I say is the truth (but not everyone will agree) This isn’t a particularly good talk 4Why give a talk  You’re compelled to  For practice (eventually, you’ll be doing it a lot)  To get a job (an indication of teaching ability)  To tell people stuff  To become known  To get people to read your papers 5Basic principles  Know your audience (don’t overestimate them)  Take it seriously (you’re being judged)  Keep things simple (have one or two key points)  Tell a story  There are lots of ways to be good (and bad)  Know yourself (adopt another’s style at great peril)  Go to lots of talks and pay attention  Images over text 6Multiple imputation 0 16 22 40 56 Genetic map: Observed data: Imputations: = AA = AB = missing 7Multiple imputations 8Imputation LOD curves 10 8 6 4 2 0 0 10 20 30 40 50 Map position (cM) 9 LOD scoreMore basic principles  Humor can be good  Start with an application  Nobody wants to hear the technicalities  Give a good summary at the end  Pay attention to your environment  Don’t bullshit; admit ignorance  Don’t go over the time limit  The content is key 10Ways to annoy me  Spend a lot of time on a useless outline slide  Unlabeled graphs; unreadable or meaningless labels  Use graphs/tables straight from a paper  Use a serif font (e.g., Times)  Include lots of unnecessary digits  Include raw computer output  Read the slides verbatim  Cover everything you’ve ever done 11Outline  Background  New method  Simulation results  Application  Conclusions/future directions 12Ways to annoy me  Spend a lot of time on a useless outline slide  Unlabeled graphs; unreadable or meaningless labels  Use graphs/tables straight from a paper  Use a serif font (e.g., Times)  Include lots of unnecessary digits  Include raw computer output  Read the slides verbatim  Cover everything you’ve ever done 13A bad figure 14A better figure 70 60 Treated 50 40 30 20 Control 10 0 0 20 40 60 80 100 Dose 15 ResponseWays to annoy me  Spend a lot of time on a useless outline slide  Unlabeled graphs; unreadable or meaningless labels  Use graphs/tables straight from a paper  Use a serif font (e.g., Times)  Include lots of unnecessary digits  Include raw computer output  Read the slides verbatim  Cover everything you’ve ever done 16Ways to annoy me Spend a lot of time on a useless outline slide Unlabeled graphs; unreadable or meaningless labels Use graphs/tables straight from a paper Use a serif font (e.g., Times) Include lots of unnecessary digits Include raw computer output Read the slides verbatim Cover everything you've ever done 17Ways to annoy me  Spend a lot of time on a useless outline slide  Unlabeled graphs; unreadable or meaningless labels  Use graphs/tables straight from a paper  Use a serif font (e.g., Times)  Include lots of unnecessary digits  Include raw computer output  Read the slides verbatim  Cover everything you’ve ever done 18A bad table 19Ways to annoy me  Spend a lot of time on a useless outline slide  Unlabeled graphs; unreadable or meaningless labels  Use graphs/tables straight from a paper  Use a serif font (e.g., Times)  Include lots of unnecessary digits  Include raw computer output  Read the slides verbatim  Cover everything you’ve ever done 20Computer output anova(aov(x factor(d) factor(sex))) Analysis of Variance Table Response: x Df Sum Sq Mean Sq F value Pr(F) factor(d) 4 203.661 50.915 49.3485 2e16 factor(sex) 1 2.199 2.199 2.1317 0.1478 factor(d):factor(sex) 4 1.923 0.481 0.4660 0.7605 Residuals 90 92.858 1.032 Signif. codes: 0 '' 0.001 '' 0.01 '' 0.05 '.' 0.1 ' ' 1 21Ways to annoy me  Spend a lot of time on a useless outline slide  Unlabeled graphs; unreadable or meaningless labels  Use graphs/tables straight from a paper  Use a serif font (e.g., Times)  Include lots of unnecessary digits  Include raw computer output  Read the slides verbatim  Cover everything you’ve ever done 22More ways to annoy me  Skip the motivation  Mumble  Spend a lot of time discussing simulations  Focus on your computer rather than the audience  Use a noisy background  Use a canned Powerpoint background  Include your institution’s logo on every slide  Include other useless stuff on every slide 23Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 24Yet more ways to annoy me  Use gratuitous animations 25Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly 26Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work 27Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs 28Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details 29Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember 30Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos 31Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 32Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 33Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 34Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 35Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 36Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 37Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 38Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 39Yet more ways to annoy me  Use gratuitous animations  Flash your laser pointer around constantly  Make extensive references to others’ work  Write full paragraphs  Spend a lot of time on technical details  Use complicated notation that’s hard to remember  Have lots of typos  Reveal just a bit at a time 40Preparing the presentation  What is the takehome message  What figures do you want to show  What background needs to be covered  Want a logical unfolding of the information  Think carefully about the transitions 41Slides  Be consistent (colors, labels, layout)  Polished, but not overly so  Think carefully about colors (projector vs computer; color blindness)  Use care regarding graphic type (jpeg/png vs pdf/ps)  Design figures specifically for the talk  Add more illustrations  Know your slides (don’t let them surprise you) 42The textfree talk  Can you do it with only pictures  Wordy slides seem more for the speaker than the audience  Consider index cards (but don’t drop them) 43Practice  Practice by giving talks (take every opportunity)  Don’t let practice kill your enthusiasm  Think it through carefully (imagine what the audience is thinking)  Focus on the transitions and the tricky bits  Do you know how long it will take  “10 practice runs for one presentation” (Yikes)  Know what you need  I usually have a good idea for my “opening”  Videotape (Not me) 44Delivery  How to dress (reflects your respect for the audience)  Exude enthusiasm and confidence  Move around  Look at the screen, with the audience  Explain fully or not at all  Don’t apologize  Avoid getting yourself stuck; feel free to stop and summarize  Relax; slow down; occasional silence is okay  You want the audience to care and understand 45Nervousness  Expect to be nervous (always)  Be prepared  Know your audience  Have something to drink on hand  It’s probably not noticeable  Use two hands with the laser pointer  It’s okay to screw up now and then (I think)  It’s easier when you care about what you’re talking about 46The 10 or 15 min talk  Think: what is the goal  Explain the problem, sketch the solution, give a few results  5 – 15 slides  Simplify, simplify, simplify  If possible, choose the topic to fit the time limit 47PDF vs Powerpoint  Powerpoint can have terrible graph/font problems  PDF (almost) always looks as expected – PDF can have font problems, too – technically you need to “embed” any nonstandard fonts – but problems are rare.  Powerpoint makes sharing slides easier  Powerpoint can be easier to create  Powerpoint: easier to have color and black/white versions  Powerpoint’s defaults are terrible 48Q A  Listen  Repeat the question  Be gracious  Admit ignorance  “I’ll have to give that some thought...”  “I’m glad for the suggestion...”  “I’ve thought about that, but haven’t yet found a solution...”  Don’t say, “That’s a great question” for every question 49Final thoughts  You should be able to give a talk on your current work at a moment’s notice  Simulation results are almost always dreadfully boring  Not every piece of work makes a good talk  I always empty my pockets, just in case  I like to have an index card with slide numbers  Know your computer; adjust the screen/energy saver times 50Resources  M Alley (2002) The craft of scientific presentations: Critical steps to succeed and critical errors to avoid. Springer (Lots of fun stories about famous scientists)  S Morgan, B Whitener (2006) Speaking about science: A manual for creating clear presentations. Cambridge University Press (A bit simplistic)  RRH Anholt (2005) Dazzle ’em with style: The art of oral scientific presentation, 2nd edition. Academic Press (I didn’t like it)  PB Medawar (1979) Advice to a young scientist. BasicBooks  RA Day, B Gastel (2006) How to write and publish a scientific paper, 6th edition. Greenwood Press  PJ Feibelman (1993) A PhD is not enough Addison Wesley  Brad Efron (2013) Thirteen Rules. http://wwwstat.stanford.edu/ckirby/brad/other/2013ThirteenRules.pdf www.biostat.wisc.edu/kbroman/presentations 51