On December 8th, I attended a terrific one-day course on presenting data and information. Before the course began, I did the socially responsible thing by modern standards and updated my Facebook status to notify my ever-concerned network of friends as to my whereabouts for the day. One of them commented on my status; “you really need a course for that?”. I laughed. It does sound funny, but considering how many people utterly fail to present information clearly, I would have to say that, yes, such a course is a good use of time and money for anyone in business that needs to communicate effectively with others. 

Evidently I’m not alone in this belief because the event was sold out. I was expecting maybe, maybe, 100 or so people to turn out for this thing. Boy was I wrong. The event was held in the Grand Ballroom of the Hyatt on Union Square in San Francisco (which I realized about a half an hour into the course, to my horror, was the same ballroom where my high school senior prom was hosted over 20 years ago! It took everything I had to fight off the distracting mental images of big eighties hair and bad cummerbunds.) In my estimation, the ballroom held about 500-600 people and not a single chair was empty. In fact, they brought in more chairs for late comers and a few people arriving later still resolved to standing in the back. (Hopefully, for the Hyatt’s sake, no SF fire marshals are reading this). 

So what could all of this interest in presenting data be about? Why all the fuss? Well, it didn’t hurt that each attendee received all four of the widely acclaimed and very popular books that Edward Tufte has written on the topic. The real draw though, I think, was that the course was presented by Tufte himself. In case you aren’t geeky enough to know who he is, here’s a quickie wiki:  

Edward Rolf Tufte (born 1942) is an American statistician and Professor Emeritus of statistics, information design, interface design and political economy at Yale University. He has been described by The New York Times as “the da Vinci of Data”. Full wikipedia entry >

In other words, when it comes to analytical design, Tufte is the man and he did deliver a superior presentation, as one would hope. Here is a general list of the topics covered throughout the day:

• fundamental strategies of analytical design
• evaluating evidence used in presentations
• statistical data: tables, graphics, and semi-graphics
• business, scientific, research, and financial presentations
• complexity and clarity
• effective presentations: on paper and in person
• interface design
• use of PowerPoint, video, overheads, and handouts
• multi-media, internet, and websites
• credibility of presentations
• animation and scientific visualizations

And below are the pertinent points that I captured in my notes. I hope you find the info as useful and thought provoking as I did.

Local optimizing leads to global “pessimizing”.

As we all know, stakeholders often ask for additions/enhancements to focused areas of a design without considering the big picture. If gone unchecked this can lead to clutter or create a fracture in the greater meaning of the piece.

Never design to the lowest common denominator.

I found this point particularly interesting. There is an inclination to underestimate the intelligence of your audience. Tufte says don’t do it - don’t dumb down your presentation out of fear that your audience won’t get it. Give them the credit they deserve and present the whole story as it needs to be told, presuming that if you are doing your job correctly by presenting information clearly, they are smart enough to understand it.

Don’t read bullets from presentations.

Okay, duh, but aren’t you surprised how many people still do this? And aren’t you even more surprised when you catch yourself doing it? 

Provide a “super graphic” as a supplement to your presentations.

Super graphics are usually somewhat complex, interesting, involved images intended to present a ton of info clearly. Tufte loves these because audience members get involved with them on their own terms which is better than the presenter dictating how they experience and absorb the information. Tufte claims the presenter should be flattered when audience members are fascinated with their image - even if they seem to not be listening to what the presenter is saying for a time - it’s a good thing. It means they are engaged.

Design visuals so it’s easy to make comparisons of data.

There are some tips on how to do this below, under “grand principles of analytical design”.

On PowerPoint and Chart Junk:

Tufte is not a big fan of Microsoft’s de-facto standard tool for delivering presentations. PowerPoint, in his opinion, is overproduced and self important. His argument centers around the idea that much of the time, it seems to be more about the tool than the content of the presentation. For example, instead of referring to presentations by their subject, too often we just call them “PowerPoints”. Tufte made a further point by calling attention to the fact that too frequently the charts generated by PowerPoint are technically incorrect and compromise the integrity of the data. In his campaign against misusing (or using at all) the charting tools in PowerPoint, Tufte coined the term “chart junk”. He prefers instead to create technically correct and visually elegant charts and diagrams by using other professional drawing tools and either importing them into to PowerPoint slides, or better yet, where appropriate, printing them and distributing them as handouts.

Tufte offered some advice for creating (non-fiction) presentations:

• Find good examples and copy them.
• Create templates / style sheets for each type of info object such as tables, text, bullets, etc. then stick to them to create some consistency throughout the presentation.
• Don’t over design or decorate - just make the information clear.
• Find or create a super graphic - provide as a handout/leave behind. It lives on and gives them something to  refer back to. Let’s them explore a lot of info on their own terms.
• Prepare presentation using the conceptual model of “elite journalism” - because, after all, you’re reporting.
• Don’t make it look “pitchy”.
• Show that you have integrity and that they can believe you. The best way is to provide detail. Back up claims with examples, site sources, etc.
• Ok to use some industry jargon. Avoiding it can appear as though you are talking down to them or patronizing them. Using it makes the gesture that you are including them as an insider.
• Ultimately, convey 2 things: 1) What your story is. 2) Why they should believe you.
• When presenting info, do whatever it takes to make it clear. e.g. for project management presentations, Tufte recommends wall charts. In that context, you need to make comparisons over tasks and time. Wall charts do this way better than a series of slides stacked in time. All data is presented in a flat manner, making comparisons easier - slides require audience to remember too much to make the comparisons in their minds.
• Move away from legends and instead use smart labeling right on the data elements to communicate their purpose and meaning.
• Get off “flat land” where possible (2D - screen, printed on paper, whiteboard etc.). Build 3D physical models, where it makes sense.

Tufte listed his Grand Principles of Analytical Design from his book, Beautiful Evidence:

1. Show comparisons, contrasts, differences
2. Show causality, mechanism, structure, explanation
3. Show multivariate analysis; 3 or more dimensions
4. Integration of evidence; whatever it takes
5. Documentation: Document everything and tell people where it came from - site source - will help build credibility - even if they don’t read it all. Presentations that don’t site sources are likely “cherry picking” - i.e. pulling out nuggets of info that privilege the narrator while concealing big picture details that could tell another story
6. Content counts most of all

One final grand principle that is not mentioned in the book:

• Design so comparison data is presented adjacent in space rather than, say, stacked in time like in a slide show or in a series of graphics revealed one at a time via a rollover interaction, scrolling, etc.

Additional design tips:

• It’s important to remember that principled design is about supporting intellectual, cognitive tasks
• Principles of analytic design are indifferent to the mode of delivery - i.e. technology employed. 
• In general, a “flat” interface helps to eliminate the need for a user manual. Especially in comparison to a deeply hierarchical menu system. Think iPhone vs. most most handhelds.
• Recommended we research “small multiples” in the books and the chapter in Beautiful Evidence on “Sparklines” - p. 46