All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Beautiful EvidencePaper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
All 4 clothbound books, autographed by the author $150
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer:
Visual and Statistical Thinking $2
The Cognitive Style of Powerpoint $2
Seeing Around + Feynman Diagrams $2
Data Analysis for Politics and Policy $2catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
Houston TX, January 29
Austin TX, January 31
Dallas TX, February 2
In the past, I've had good results presenting a some scientific results at a conference as a 4x6' poster. I talk people through the poster's figures, which are laid out from left (intro/methods) to right (results/conclusions). I provide a reduced-size handout of the poster, which most visitors to my poster choose to take away. This format seems to work well for communicating what I've done.
So, how can I reproduce this successful experience when I am giving a lecture to a larger, seated audience? The entire argument is visible to the poster viewer, who knows where each figure fits in as I discuss it. The lecture viewer, however, must be shown individual figures in sequence. How can I let them scan through my entire arguement and know where we are within it? You've recommended providing a handout, such as my reduced-size poster. In addition, what about a second screen to project the talk's overall organization? How to do so in a way that could be read at the by back of the room - perhaps a flowchart? Alternatively, perhaps the top of each slide would list the section ("Methods", "Results", etc) that it falls under, similar to the way a book lists the chapter title at the top of pages. The slides' section titles would match those in the handout.
Any thoughts on these ideas, or alternatives, would be appreciated.
-- Kathleen Edwards (email)
Give everyone, always, a paper handout: a preprint, a reprint, an 11 by 17 piece of paper folded in half. You want to leave traces by providing high-resolution materials that live beyond the moment. Posters, which are usually read at a distance, tend to be low resolution (like slides). Real science is high-resolution; use high-resolution methods of presentation.
There is much more in the several threads on presentations on this board, as well as in The Cognitive Style of PowerPoint.
-- Edward Tufte
First question first...The posters that Kate refers to are very common at scientific conferences. There are "poster sessions", where rooms full of these posters are put up for perusal and authors are commonly there to discuss. So the question is, do the posters offer an effective way to display and explain the work? I think they do, when put together well. If posters have low probability of effectiveness, like PowerPoint, then we should discuss that.
The second question, then, is that if posters are effective to small audiences, why aren't they to large audiences? (Is that a fair summary?) I can't see why the same poster could not be used for the larger audience. Kate, have you tried that? Have you had trouble just using the poster with the large audience?
-- Kent Karnofski (email)
Posters work well for ~10 or less people, who typically stand 3-6' away and can step closer to see detail in the figures. For a larger audience, however, the ones in the back would be too far away to read the poster. When I present the same material as a seminar, the audience is ~50 people and they want to be able to sit and listen. When I speak, how can I incorporate a poster's advantages, namely 1) high detail in figures; and 2) the viewer's choice of path through the argument, for example backtracking to an earlier figure that was confusing the first time around, or comparing two figures? The suggestion of handouts is excellent but I think there might be some speaking/slidemaking tricks as well. My goal is that the audience retains my argument, even if the supporting detail is forgotten. At my upcoming conference, I'll try to see how the good speakers do it.
-- Kate Edwards (email)
Kent Karnofski asked, "do the posters offer an effective way to display and explain the work?" at scientific conferences. Yes, when they're done well. Posters are good old high-resolution paper, they allow people to absorb material at their own pace, and their major drawback -- their transient nature -- is easily overcome with handouts.
The major advantage of posters, however, is not how the display information, but that they facilitate social interaction at conferences. A presenter is very close to the audience (typically only a handful of people at a time), which is much more conducive to audience members asking questions, volunteering ideas, etc. Poster sessions are lively places, with many warm greetings, casual conversation, and jokes in addition to the exchange of technical information.
Slide presentations are monologues (sometimes brilliant ones, to be sure); poster presentations are conversations.
-- Zen Faulkes (email)
I think Kindly Contributor Zen Faulkes is exactly right.
-- Edward Tufte
I have been surveying poster designs across multiple disciplines for some time and have yet to see anything which lays the groundwork for research poster style in particular. I have done some literature review, reading Visual Literacy texts (Dondis, et. al.), Data Graphics books (Bertin, Tufte), Grid design works (Bosshard, Muller-Brockman.
These texts are all quite informative, but I have this sense that there is something about the balance of textual information and graphical representations of data and information which is not quite the same as those two elements separated. This seems especially pertinent as technology accelerates (often without contemplation), and research ideas get forgotten/abandoned under the demands of the economy, many good ideas good benefit from better presentation.
The (sub-)genre of poster design runs from the tacking up of Powerpoint presentations (which I agree introduces too much cognitive dissonance), to decently laid out posters, to what I believe Edward has termed "duck" posters; all of them seemingly without a grounded theory of aesthetic/design for their particular genre.
Could Mr. Tufte or any other forum member recommend any other documents that either focus on the information design and layout of Academic Research Posters?
I have also found the following texts: Scientist's Guide to Poster Presentations, by Peter J. Gosling; Preparing Scientific Illustrations: A Guide to Better Posters, Presentations, and Publications, by Mary Helen Briscoe; Displaying Your Findings: A Practical Guide for Presenting Figures, Posters, and Presentations, by Adelheid A. M. Nicol, et.al.
-- Sean M. Lanksbury (email), January 23, 2004 (moved to this thread)
-- Edward Tufte
Could Mr. Tufte or any other forum member recommend any other documents that either focus on the information design and layout of Academic Research Posters?The Society for Neuroscience in famous for its huge poster sessions. Row after row after row of posters, changing twice a day. Their guidelines for effective posters can be found here. Very basic stuff. Unfortunately, I've seen many posters at the SFN meetings that didn't follow these guidelines, and I wished they had!
-- Zen Faulkes (email)
A few years ago at the Ecological Society of America conference, I jotted down some notes on poster design while I was bored out of my skull in an oral session. Probably there are similar points in the cited materials, but I'll throw out some of the ideas I had that day.
Put the right amount of text on the poster, maybe around 1000 words, which can be read in the few minutes that people are likely to spend on your poster. Resist the urge to copy and paste whole sections from published manuscripts! Abstracts duplicate information that is later to come in the poster, so they are unnecessary. I think of posters as expanded abstracts with figures and tables.
The number one problem that I saw on the ESA posters (probably rampant elsewhere as well) was the use of unfriendly text blocks. Fourteen inch wide single spaced paragraphs in sixteen point sans-serif type are so annoying to read that I'd bet people don't read them. Also, don't typeset your important research in Comic sans or other silly fonts.
I think the benefits of using multiple panels are totally underappreciated! Yeah, poster plotters are cool beans and you look all official carrying around a big poster tube, but consider... Multiple panels ease layout restraints and can be arranged in captivating and useful ways. As opposed to a 5' by 3' poster, a collection of smaller panels can be easier to make, print, transport, and replace. Depending on your printer they can be cheaper and higher resolution too. If you have photographs, you can bring real enlargements for a stunning addition to your presentation! For extra spiff, mount panels on foam core.
I find studying displays at good museums a good way to learn poster presentation skills. I put one more vote in for the superiority of poster to oral sessions.
-- Anthony Darrouzet-Nardi (email)
Let me get to the point first, and then fill in some details.
During the past few years, Leon Liegel and I have developed a web site called "Creating Effective Poster Presentations" at www.ncsu.edu/project/posters/
First, we think readers of this thread will find our site useful. Our view is that posters are a visual communication tool, and that an effective poster serves two purposes: (1) to engage colleagues in conversation and (2) to get your main point(s) across to as many people as possible. Handouts during the presentation can provide a "higher bandwidth" information channel, but the poster itself should be straightforward and relatively simple. I find myself agreeing with many of the points made on this thread -- yes, posters are useful when done well. Anthony Darrouzet-Nardi provided an excellent summary -- the concept of the poster as an illustrated abstract is very much in line with what Leon and I suggest.
Second, we'd love to hear your ideas about improving our site, both the content and the presentation. Be brutal, if necessary. We've reveived good reviews from visitors, but it still needs more work -- for example, we should provide more depth behind our simple prescriptions. I'm also wondering if we've been pushing a bit too much toward simplicity. Perhaps the graphics should be denser (though we're quite sure that the text should not be).
Now, the promised details.
After looking at a lot of ineffective posters at scientific meetings, I decided to do a formal survey at one meeting in which 142 posters were presented. The survey focused on overall appearance and the ease with which key pieces of information could be found (e.g., objective, results, summary). I designed a survey instrument that could be completed in about a minute. Details are available at www.ncsu.edu/project/posters/documents/PosterFailure.pdf
Here's what I found: Many of the posters were cluttered or sloppy (33%) and had fonts that were too small to read comfortably (22%), especially in the figures (44%). Within the one-minute review period, I was unable to find the research objective (38%) and main points (42%) on many posters. These problems diminish the communication value of a poster for all but the most determined readers, but are very easily fixed once authors are made aware of them.
I think the rapid survey approach I took was a reasonable measure. Although one minute doesn't seem like much time, posters are displayed in sessions that sometimes contain hundreds of other posters. These sessions sometimes last only a couple of hours and are mixed with lunch breaks and socials. You have to get readers interested quickly, and nothing will turn off readers quicker than not being able to find things easily -- after all, there are many more posters to look at. Once you have a reader's attention, you can supplement with high bandwidth handouts and conversation to provide more depth.
At about the same time, I was using posters as a teaching technique in my biological modeling classes and wanted to provide accessible guidance for the students (see Hess, George R. and Elizabeth N. Brooks. 1998. The class poster conference as a teaching tool. Journal of Natural Resources and Life Sciences Education 27: 155-158). This led to the development of some brief checklists and, utlimately, to the website. I hooked up with Leon Liegel along the way -- he contacted me after reading the Hess & Brooks paper. Leon had been judging poster sessions for Sigma Xi, a scientific research society (www.sigmaxi.org) for years and wanted to provide guidelines to improve the quality of the posters.
And so our web site was born. It's been through a couple of overhauls and is in it's third edition. We sense it's time for another overhaul, but could really use some help. We'd like to confirm that we're headed in a reasonable direction before putting a lot more work into the site. And we'd like to hear what those of you who have also spent time thinking about posters and communication think is lacking (you wouldn't be reading this thread if you weren't already thinking pretty deeply about this stuff).
NC State University
[link updated February 2005]
-- George Hess (email)
An alternative to PowerPoint called Beamer solves one of the problems I mentioned earlier: how to show where an individual slide falls within your talk. Each Beamer slide can have a header containing the talk outline; the subheading that the slide falls under is highlighted. Here is an example of a talk made with Beamer, http://latex-beamer.sourceforge.net/beamerexample1.pdf
Versions that I saw at my recent conference were heavier on the figures and lighter on the text than this example, which made for a better presentation. As a viewer of Beamer-created slides, I found the following drawbacks: 1) hard to read the small heading with the outline, or to realize what it is, 2) template shares some of PowerPoint's problems, like distracting colors, bullet points, etc.
The software can be obtained from, http://latex-beamer.sourceforge.net/ My past experience with Latex was that it requires a steep learning curve, but it is excellent for typesetting equations.
-- Kate Edwards (email)
Scaling up poster techniques to a large-room audience
As previously stated, the key benefit of using a poster comes from the small-audience interaction - the conversation. The format allows for comparisons from one dataset (e.g. graph, picture, table, text) to another. The poster isn't the research, it's the way to discuss it in a small-group setting.
Whether you mean to discuss the data with audience participation, or just present it to this group, you should present data that needs comparison all at the same time. In addition to the handout, you could add that second (or third) screen you mention. Rather than just an outline on the extra screen, which likely serves as more of a distraction than a benefit, I suggest that any data that needs to be compared should be shown on these closely-placed multiple screens, just as you would place several pictures near each other on your desk to compare and sicuss them. When comparisons do not need to be made, project your information onto only one screen, with the other(s) blank.
To expand upon Anthony Darrouzet-Nardi's suggestions for printing and displaying the panels, I add that if you have broken down your 4x6' panel into several subsets for ease of printing, travel, and display, these subsets could be printed as individual 11x17" pages, for highest resolution examination, following a single (11x17) page of the entire layout that can show the relationship of the information.
-- Bruce Murden (email)
"An effective poster is neither a page-by-page printout of a journal paper nor a slide show, but balances figures and text."
Many posters are viewed in between or during conference sessions so the conversation aspect may be limited. Posters should be easy to read and graphically apealling. All to often I glaze past a poster that has dense text and few pictures. I agree with the "extended abstract" concept as a good way to think of a poster.
Remember that many viewers may not be in your field of study. If the they are really interested in a high level of detail they can read the paper, so have copies available. They should be able to contact you for more info, proposals, etc. I like to have buisness cards clipped to the poster that people can walk away with. The card could also list a website with a pdf of the poster and paper as well.
Question, is an email address sufficient contact information or should it include phone, address zip etc?
-- Antonio Llanos (email)
A wonderful, plain-language guideline for poster presentations is here: http://www.onlinebachelordegreeprograms.com/resources/designing-scientific-posters/
I've had excellent results using a single folded 11x17 2-sided handout (per Mr. Tufte's prescription), and then creating slides that show the entire handout and visually 'zoom in' on the various parts of the handout I am currently speaking about.
[link updated October 2012]
-- leMel (email)
This past weekend, 1000 of Australia's "best and brightest" were invited to the nation's capital, Canberra, to take part in a summit looking to Australia's future, c.2020.
This was the newly elected Prime Minister's concept, in order to reach out to the community, and seek solutions to the problems confronting the nation. I will add here that the new PM, Kevin Rudd, is a former diplomat with a special interest in China, and is likely to be the only elected head of state (outside of China and Taiwan) to speak fluent Mandarin.
Ahead of the summitt, discussion papers in various areas were released as both Powerpoint and pdf files. Let me just say I would not have used any of the slides in a formal presentation but as dashboards for preliminary reading, they were quite good. You can see examples here by following this link: http://www.australia2020.gov.au/topics/index.cfm
I have blogged my concerns about what could have happened with regard to abuse of Powerpoint here http://lesposen.wordpress.com/2008/04/10/australia-2020-summit and include a screen shot of one slide.
But today's newspaper reports, offering a summing up and opinion of the summit (which is available for viewing via streaming video in Adobe Flash until May 2008) special mention was made of the absence of Powerpoint. Indeed, what was highlighted was the use of "special facilitators" (apparently from McKinsey and Co.) who used butcher's paper, Post-it notes, marker pens, and coloured dots to collate materials "on the fly" and collect votes about priorities.
These were then transcribed into laptops and summaries provided to the Prime Minister who will now create action statements by the end of the year.
The summit has attracted huge media and public awareness and it has been interesting to see the means by which 1000 individuals each with their own agendas, have come together and had their ideas and visions collected by such "old fashioned" methods.
Googling "Australia 2020 summit" should provide a variety of interesting opinions and insights. I can't recall any other country ever attempting such an endeavour.
-- Les Posen (email)