Edward Tufte's preoccupation with the effective presentation of information is nowhere more evident than in his own publications. Examine any of his three major works - The Visual Display of Quantitative Information, Envisioning Information or Visual Explanations - and you'll find that each volume is an absolute paragon of typesetting and graphic virtue. Not surprising then that his latest monograph, The Cognitive Style of PowerPoint is not available by electronic download. You can order it through his web site, but it is only delivered in hard copy, no doubt so that he can control the exact form in which the paper appears in your hand.
Tufte's ideas on data visualization are soundly grounded in statistics and psychology, but his appreciation of the aesthetics of design and presentation evidence a fine artistic sensibility. When it comes to data presentation, he knows that of which he speaks.
Several hundred million copies of PowerPoint have been sold, and somewhere between 10^10 and 10^11 PowerPoint slides are produced every year. The essence of Tufte's criticisms of PowerPoint target the uniformity of style and formatting that it imposes over potentially diverse subject matter. He analyses PowerPoint as a means of data presentation on the same basis that he analyses the presentation effectiveness of graphs, charts and other more traditional presentation mechanisms, and concludes that it is largely ineffective as a mechanism for relating information to an audience; as he describes it PowerPoint allows speakers to pretend that they are giving a real talk, and audiences to pretend that they are listening.
The typical PowerPoint presentation is a sequential narrative of bullet points, accompanied by cheesy clip art and bright colors. Not only does this read like a children's story book, it is a means of decomposition that is inappropriate for many types of content. Information is assimilated through the formation of relationships amongst pieces of data, a process for which comparison is essential. The linear, piecemeal supply of that data provides little opportunity for the comparison and meaningful juxtaposition that supports visual reasoning. As Tufte puts it Bullet outlines make us stupid.
Tufte argues that PowerPoint, with it's plethora of styles, templates, fonts and animations, is encouraging presenters to spend too much time on the prettifying of their work and distracting them from the meaningful development of content. The frequent consequence is an overproduced presentation with cheerleader logotypes, garish colours and valueless animations, in which content has been forced into an aesthetically pleasing format at the expense of accuracy and completeness.
Tufte also has some particularly uncharitable things to say about the use of PowerPoint's charting facility, giving several examples of how easily it can be used to distort and misrepresent, encouraging false analyses and conclusions. Tufte declares To deal with a product that messes up data with such systematic intensity must require an enormous insulation from statistical integrity and statistical reasoning by Microsoft PowerPoint executives and programmers
There are several specific examples given of the way that PowerPoint can corrupt information and mislead an audience.
The first is the PowerPoint presentation that NASA engineers gave to senior NASA officials during the flight of the ill-fated Columbia space shuttle in January 2003, in which they assess the possible damage to the left wing resulting form the impact of several chunks of debris 81 seconds after liftoff. Tufte includes copies of some of the critical slides in the presentation, each of which has been "content impaired" in typical PowerPoint style. Abbreviations and syntactic shortcuts intended to squeeze the content into the slide format and consistent use of bullet point hierarchies six levels deep obfuscates the message and hinders objective analysis. The viewer is left with an over-optimistic impression of the problem, and an unwarranted degree of confidence that any tile damage caused will not be problematic upon shuttle reentry.
On a lighter note, Tufte illustrates how the style and finesse of a superior oratory can be reduced to an insulting series of information snippets, by presenting the Gettysburg Address as it would've appeared had Abraham Lincoln been using PowerPoint.
A few weeks after reading this monograph, I was called on to deliver a presentation on code reviews. My first reaction, having found Tufte's damning arguments quite convincing, was to consider delivering the presentation without using PowerPoint at all (a little daunting for those of us not very comfortable with public speaking, for whom the slides serve as a comforting intermediary between speaker and audience).
In the end I used PowerPoint to present just the major graphic elements of the presentation. Had I followed Tufte's advice more closely, I would have put those graphics on paper handouts instead, taking advantage of it's higher resolution and also giving audience members something to refer back to later, but I've always found that giving people pieces of paper at the start of a talk distracts them too much from what is being said, as they tend to read ahead through the material, in advance of the presenter.
Whether you accept Tufte's arguments in their totality or, like me, only in part, the monograph is worthwhile reading for anyone who routinely uses PowerPoint for their presentations. It will at least encourage you to think critically about the role that the product will play, and perhaps discourage you from following the read the slides style talk which has more to do with story time in the class room than it does with the effective transferal of information.