Tip of the week: ask yourself these two simple questions

You're about to give a presentation to a scientific audience.  Before you do, ask yourself:

1) What do I hope to achieve by giving a presentation?
2) What can I realistically achieve?

I typically proceed my own "how to do a scientific presentation" presentation by tossing these questions out to the audience.   I usually get blank looks, as if no one ever thought about such questions.  Too many scientists simply "grab their slides," without establishing goals for their presentation. 

Eventually someone in the audience will say, "I want to explain my science."  Fair enough, but this gives rise to other questions.  To whom do you want to explain your science?  Scientists in your specific subdiscipline?   Scientists in related disciplines?   Scientists in different disciplines who may have methods and technologies relevant to your work?   A more general scientific audience?  Have you altered your slides and your spoken narrative to target the audience you care about? 

A second related question involves the depth of your explanations.   Realistically, nobody is going to remember your slides 24 hours after the talk.  Even experts in your sub-discipline will have trouble as you rush through dozens of detailed slides in a futile effort to communicate a comprehensive and detailed review of all your work. 

Slide presentations are not the place to "prove" results or give overly detailed reviews.  A slide presentation is an opportunity to inspire your audience, motivate new collaborations, get other scientists to follow up with questions and perhaps get them to read your journal articles.  Presentations are a great place to motivate thoughts about how your work relates to other ongoing research.  These are realistic expectations for what you can achieve.  Detailed "proof" is best left to journal articles and efforts to replicate results. 

Tip of the week: Engage the audience and you need to "break the ice!"

Audiences are passive.  Typically they defer questions until the end of a one hour scientific presentation.  By that time the questions have been forgotten, assuming the speaker has left any time.

Why such passivity?  The desire to be polite?  Fear of asking a question that reveals ignorance?  They'd rather be doing their email? 

In any event, passivity is a prescription for getting nothing from a talk.  People don't learn effectively in a passive mode, especially when dozens of complicated slides are whizzing by.  You as a speaker get nothing if there is no audience reaction.  It's lose-lose.  You might as well pass out your journal article and let them read that.

The solution?  The speaker needs to "break the ice."  Tell the audience they can ask questions as they occur, rather than wait until the end.  Come with a set of questions that you will toss out to the audience.  Give one of your friends advanced warning that you're going to call on him/her if nobody responds.  Have another friend set the lower bar by asking a very basic question.  Once the "ice is broken," there will be a lot more interaction.

Tip of the week: An 11th presentation pitfall

This week I posted a commentary on the Nature Jobs Blog.  The commentary identified the 10 most significant presentation pitfalls, taking off from talk show host David Letterman's famous countdowns of the top 10 in a category.  The post can be found at: 

 http://blogs.nature.com/naturejobs/2016/02/10/a-david-letterman-like-countdown-to-the-10-biggest-pitfalls-in-scientific-presentations/#more-8957

Actually it was hard to limit the number of pitfalls to 10, so here is number 11:

Creating separate audio and visual narratives
Too often the speaker's words are not related to the visual on the screen.  The audience doesn't know whether to listen to the speaker or try to read the slide. 

Your job as speaker is to explain the slide.  There should be a communication synergy between your oral narrative and the visual.  Use the pointer to show the audience which particular part of the slide you are discussing.  If there are elements of the slide you don't discuss, ask yourself why they are on the slide.  Most of the time you can remove them, helping the audience focus on the key points.  

Tip of the week: A car is not an automobile (in a presentation)

One sure-fire way to lose the audience is to switch terminology in the middle of a presentation.  You need to stick to one name for anything you are trying to represent.  The USA is not the United States of America in a presentation unless you tell the audience it is.  

You may want to invoke abbreviations after you introduce a concept.  For example, suppose you want to use the term EMR for an Electronic Medical Record.  The first time the concept appears on a slide it should look like the following: Electronic Medical Record (EMR).  Now point to the words with your pointer and tell the audience that you will now be referring to Electronic Medical Records as EMRs.  Anytime you switch nomenclature (which you should avoid as much as possible) tell the audience that you are doing that. 

The same holds true for logos or any sort of visuals symbolizing a concept.  Keep your terms and visuals as consistent as possible. 

Tip of the week: preparing a presentation is more science than marketing

I often get the complaint that refining a presentation takes scientists away from their research for the sake of marketing.  This represents a fundamental misunderstanding of what can be achieved through careful preparation of a presentation.  First an example...

Recently a graduate student invited me to his PhD oral exams.  The exam consists of a 20 minute presentation followed by questions from the committee.  I asked the student if he would like me to review his presentation prior to the exam.  He replied that the presentation was not ready, and as it turned out, he worked on it right up to the exam.  The presentation itself was adequate (given the low standards for many biomedical research presentations), but the question and answer session did not go well.

It was obvious that had the student reviewed the presentation with an editor prior to the oral exam, and had that editor not been an expert in the student's area, and had that editor insisted that every slides have meaning, many of the questions would have been anticipated.  A detailed slide-by-slide review by an objective outsider is the best way to gain an understanding of one's own material.  It forces the scientist to think about issues in ways they have not considered.

But the lessons go well beyond a thesis exam.  Biomedical research is very much a bottom-up activity, with little theory to structure research or suggest obvious ways to organize research results.  A presentation is very much a top-down activity, defining the major ideas and then bringing in supporting data to "prove" those ideas.  Presentation preparation is one of the few activities that forces the biomedical researcher to organize findings in a highly structured way.  It forces the researcher to relate findings to scientists in different but related disciplines.  It is a necessary follow-on for the detailed lab work that often leaves scientists thinking that laboratory technique, rather than scientific ideas, are the end points to research.

Tip of the Week: A tip worth repeating

The above slide repeats the first "tip" on this website and it is the most important "tip" I can offer.  "Build up" don't "cut down."  Almost every other "tip" is derived from this idea.  

Too many scientists prepare a presentation by digging into a folder with hundreds of PowerPoint slides and asking the question, "What can I cut?"  The problem is that everything seems essential and the presentation ends up containing far more information than the audience can digest.   Coherence is lost and the talk is little more than a "data dump."

"Build up" don't "cut down."  Identify a concise narrative (symbolically illustrated above by a single slide) and then add detail given the available time and ability of the audience to understand. The September 21, 2015 "tip" gives three alternative ways to develop that narrative.  Once the narrative is in place, try giving a two minute version of the talk without slides.  Then try a five minute version with slides, then a ten minute version, etc.

Not only does the "build up" strategy allow the speaker to maintain coherence, it helps cope with other problems.  Your talk may be scheduled for 45 minutes, but the "build up" strategy will prepare you for a 30 minute version should your time be shortened (no, the answer is not to talk faster when your time is cut).  Also having developed a simple narrative at the beginning, you will have a story and structure that is accessible to a wide audience, even if some of the details are only comprehensible to a few specialists. 

Tip of the week: a theoretical model for developing a presentation

This "tip" draws on several other "tips" in this blog and are referenced in the text.  This "tip" places many of the other "tips" into an overarching framework.  

Below are the first four steps in building an effective presentation. 

The first step is to develop a simple narrative of your presentation.  The September 21, 2015 tip gives three strategies for developing such a narrative.  With the narrative in hand, formulate the overarching question for the presentation in very simple terms, recognizing that you will have an opportunity to define it in greater detail later on.  However it is important that you, and as much of the audience as possible, understands the overarching question that will be answered.  The third step is to develop a "Set-up" slide that motivates the question and then an outline slide that describes the sections of the talk that will allow you to answer the question (See Nov. 17, 2014 tip for a discussion of the set-up slide).  You've now got the first two slides of your presentation.

What next?  As shown by the red arrows and text in the figure below, you essentially repeat the process for each of the sections you've identified in the outline slide.  What is the narrative that gets you from the first section to the second, second to the third, etc.?  Once you've defined that narrative, "build-up" each section for the time you have and the level of audience sophistication (see the December 7, 2015 tip for a discussion of "building up").  Start with a two minute version of a section and then add more material if you have additional time.

The above figure is my theoretical model of how to develop a presentation.  Like all theoretical models of a qualitative process, it is probably impossible to follow precisely.  I find it to be iterative, e.g.  I may do steps 1-4 and then realize I need to reconsider my narrative.   Throughout the process the narrative remains the guide for what to include and, more importantly, what not to include in your presentation. 

Once you have mastered the model you can jazz it up to fit your style and perhaps the peculiarities of individual presentations.  One "jazzed up" approach is shown by the green figures below.   I like to start a presentation with a highly visual slide that motivates the audience (see November 24, 2014 tip).  I may also answer the overarching question right after the "set-up" slide, knowing that I will have an opportunity to give a more detailed answer later.  Remember, surprises don't work well in a presentation, so let the audience know about your findings right away (see November 15, 2015 tip about surprises). 

Tip of the week: surprises don't work in slide presentations

Presentation creators have a tendency to want to build toward a surprising climax where unexpected results are revealed at the end.  Don't to this, it doesn't work!

Slide presentations are inherently difficult to listen to.  Some members of the audience leave early, some zone in and out, and some just miss things, no matter how good the presentation.  So weaving a complicated narrative with unexpected twists and turns usually doesn't work (think about how many detective movies you've gone to, where despite months of script editing, you can't follow the plot).

There's an old saying about slide presentations, "Start out by telling the audience what you are going to tell them, tell them, and then tell them what you told them."

One of my favorite techniques is to present a visual roadmap diagram early in the talk that allows you to describe all the conceptual points concisely (see October 15, 2015 "tip"), clearing the way for the discussion of the individual measurements, methods, and uncertainties.

If the audience understands the entire story at the beginning, they'll get much more out of the entire talk.  

Tip of the week: Do we have the right incentives?

Taking a break from the usual "technical" tip, I'd like to ask a question about incentives in the system.  Is there a strong incentive to communicate effectively?  I ask this because it seems:

1.  Many senior researchers seem to be too busy flying around giving talks to take the time to prepare a good talk.  Is giving many lousy talks better than giving a few great ones?
2.  Often researchers say that they don't have the 3 or 4 hours to improve a talk.  This ignores the argument that each member of the audience may waste an hour listening to a talk they don't understand.  The system doesn't seem to optimize the time of the entire community.
3.  And perhaps most troubling, I've noticed that a few researchers really don't care if they are understood.  Their primary goal is to record their talk on their CV.

I hope this is not the case, but what I see troubles me.  Another indication of this is that many researchers use their full hour and don't leave time for questions.  What is the point of giving a talk if there is no interaction?  A journal article is probably easier to understand than a slide presentation.  

Tip of the week: don't rush when time is short

This week I attended a conference that featured a 90 minute session.  The first two speakers took 35 minutes each, leaving the last speaker with only 20 minutes for a planned 30 minute talk.

First of all it is important to stick to the allotted times.  The overruns meant that there was no time for questions.  Questions and challenges are a critical part of a presentation.  You want the audience to react.  You want to understand what they are not understanding. 

The last speaker chose to give his 30 minute presentation in 20 minutes by talking faster.  The slides became a hopeless whirlwind.

A major theme of this blog is to build a presentation "up" and not to "cut down" (see August 18, 2014 tip).  The first step to developing a 30 minute presentation is to develop a two minute presentation and then build up with a 5, 10, 15 minute presentation.  If you have developed your presentation properly, you will know how to cut back if the time is less than you planned.   Talking faster only confuses the audience.  Resist the childlike urge to say everything you know.

More often than not, the time you have for a presentation will be less than originally planned.  Go into every presentation with a battle plan for cutting content, not speaking faster.

Tip of the week: 1) It can help to summarize first and 2) an idea for an innovative outline slide

I've been auditing a course on tumor immunology.  The  format centers around individual students presenting results from particular journal article.

The presentations have been hard to follow because students are scanning figures from journal articles and using them as presentation slides.  As discussed on this blog (see October 1, 2015 "tip") journal article figures generally are too complicated to be effective overhead slides.  The axes are too small to read, more information is reported than what the speaker discusses, the figures were not designed for projection on a screen, etc.  The talks would be much easier to understand if students would redraw the slides with only the critical information they plan to talk about.  No need to make the audience strain their eyes to figure out what is relevant and what is not.

But on to the main point of this "tip."  One of this week's talks focused on the following system of pathways.

This figure was shown at the end of the talk and as such, it was only at the end that the proceeding slides were placed in the appropriate context.  The first 90% of the talk, where individual pathways were discussed, was difficult to understand.  Had this "system" been shown at the beginning, the discussion of the individual pathways would have been far more meaningful.  In reality, the main thematic points of the talk could have been summarized in 2 minutes using this figure.  After that the details would have been far more comprehensible and meaningful.     

The "bottom line" is that the audience generally doesn't know where you are going unless you tell them.  They don't understand the significance of individual measurements unless those measurements are explained within the context of an overarching biological problem.   They need the big picture summarized up front so they can place details in the appropriate context.   See September 21, 2015 "tip" for techniques to help you do this.

One other note, the above figure could be used as a very innovative outline slide.  After discussing a particular pathway, the speaker could then return to the above figure with the next pathway highlighted in some way (e.g. a big red circle around it).   This would provide the audience a reminder of why the next few slides are important and how they related to the big picture.

Tip of the week: don't convert journal article figures into slides!

It is not uncommon for scientific presentations to include graphics that have been prepared for journal articles.  Don't do this!  A graphic for a journal article serves a very different purpose than  an effective visual for a slide presentation.

Figures in biomedical research journals typically contain multiple graphs, abbreviations, and a great deal of complexity.  Axes are in small font and often written in a vertical format.  This can be acceptable, even desirable, in a journal article because there is the opportunity for the author to explain each graph and the reader can take the time to reread the article and study the figures in depth.  In a journal article the audience (the reader) sets the pace for learning.

The opposite is true in a slide presentation.  The speaker sets the pace for learning and the audience suffers with inappropriate tempo.  When a new slide appears the audience does not know what part of the slide to focus on.  Those multi-graph slides derived from journal articles add to the confusion.

Every slide in a presentation needs to have one overarching message and the data should support that message.  Extra graphs that the speaker does not discuss only confuse the audience.

Remember, "talk about what you show, show what you talk about."  If you don't talk about it, why is it on the slide?  Take the time to edit those journal article figures so that they are appropriate for a slide presentation.   

Tip of the week: 3 ways to prepare for a presentation

The worst way to prepare a presentation is to start by making slides.  A presentation should be a coherent narrative.  Slides inherently fracture this coherence into individual pieces (each slide).  It is very easy to get caught up in the details of each slide and lose the coherence.  Below are three techniques to avoid this problem.

1.  The Essay
On the September 2, 2014 "tip," I suggested speakers write a one page essay identifying all the points they intend to cover in the talk.  Prose are inherently more coherent than individual slides.  This essay should be geared toward the least informed member of the target audience (not all members of the audience, but the least informed that you care about reaching).  It should not contain data, but should be used to form the narrative.  It should identify are the key points that will be "proved" in the slide presentation.  It should have an introduction, methods, results, conclusions. 

2.  The "Chalk Talk"
A lot of biologists are not comfortable with writing an essay that doesn't include figures and schematics.  If this is the case for you, I'd recommend you prepare a "Chalk Talk" before drawing slides.  A "Chalk Talk" involves going to the chalk board (or poster board on an easel), giving a talk, and drawing figures as they are needed.  Since you need to draw the figures in real time (don't draw them in advance) you will be forced to include only the essential details.  Your oral narrative will be simple and straightforward and you will cover all the essential points.  You can then begin to think about visuals that support that simple oral narrative

3.  The "Index Card" Method
Method 3 should probably be done after (and in addition) to methods 1 and 2, but might do as a substitute. 

Get some index cards and envision every card as a slide in your presentation.  You may write one sentence on each card.  That sentence should be the main message of the slide.  After you've done all the slides (index cards) see if you can talk your way through the presentation using only the index cards.  Just like in the essay, you don't need to talk about the actual data, but you do need to say (qualitatively) what the data will show.  When you go to draw slides, include only material that helps you make the point labeled on the index card. 

You may want to lay the index cards out on a table and move them around to find the optimal flow of information. 

Tip of the week: An addendum to the March 12, 2015 tip, "Never Believe the Comment, 'Great Presentation.' "

As I wrote last March, it's too easy to get compliments on a bad talk.  People want to be nice.  Make sure you have an honest friend in the audience who can give it to you straight.

My addendum is that even the most experienced and skilled presenter can mess it up.  A few years ago I was working with a very senior biomedical researcher and we were preparing a presentation that represented a major "ask" from a very wealthy donor.  I suggested he "dry run" the presentation.  His response was, "I've given hundreds of presentations so I don't need to do that."

Not surprisingly, it was a disaster.

Every presentation poses problems for audience comprehension.  The presenter needs to identify those challenges before presenting and think about the language that will best help the audience.  This challenge exists no matter how experienced and expert you may be. 

Two common mistakes

Early this week I heard a presentation by an excellent speaker.  The individual spoke loudly, clearly, and had a gift for verbally organizing thoughts and words.  Nevertheless the presentation had some real problems and the speaker communicated far less than intended.  There were two major problems:

1. The slides did not help the speaker.  They were cluttered with a massive amount of irrelevant information that distracted the audience.  The speaker never made reference to this extraneous material on the slides.  Remember, if you don't talk about something on a slide, ask yourself why it is there, and nine out of ten times, you'll remove it.  This speaker would have been far more effective not having any slides and drawing a few figures on a white board.

2.  Information overload.  The speaker attempted to present a fairly detailed account of six recent clinical trials.  This was way too much information for the audience to absorb.  The speaker would have been better off  describing one trial and providing concise summaries of the other five (if the audience really needed to hear about all six).  We all suffer from a child-like impulse to talk about everything we know and there is a tendency to assume that the audience understands and retains everything we say.  Reject these urges!  The audience cannot absorb great detail.  There is a tendency for speakers to develop talks for themselves rather than the audience.  The speaker needs to help the audience every step of the way.