Saturday, July 05, 2008

When PowerPoint is wrong...

Last week I spied, tucked on a shelf following a recent office shuffle, a copy of The Cognitive style of PowerPoint: Pitching Out Corrupts Within, by Edward Tufte - a critique of Microsoft's infamous presentation app as an effective method of communication.

It had been bought by a colleague, who has since moved-on, who was a great fan of Tufte and skeptical of the use of PowerPoint as an authoring tool; though I'm not sure there was necessarily a direct causal link. Since I'm spending a lot of my time (ie most of it) elbows deep in slides, images and Articulate Presenter, it seemed an opportunity ripe for seizing.

On 1 February 2003 Space Shuttle Columbia burnt up on reentry.It actually makes for pretty shocking reading. The central strand of evidence is the example set at NASA, where PowerPoint usurped all the traditional methods of inter-scientist communication and ended up being the de facto reporting tool - of one of the premier scientific organisations in the world. Tufte gives us reasons that show that it was patently absurd that this should happen - the inability of PowerPoint to accurately present scientific formulae is only the beginning, and other flaws in its suitability to the task at hand meant that it could be argued that Bill's beloved slideshow generator played a part in the chain of events that led to the 2003 space shuttle disaster.

The deeply hierarchical structure of the default layouts in PowerPoint comes in for particular criticism, the root cause of the tendency of users to bullet EVERYTHING, principally it seems, because they can. As a result of the layouts chosen and the implicit level of importance that the bulleting hierarchy imposes, argues Tufte, crucial aspects of evidence that pointed to the risks posed by a seemingly minor incident on launch were missed by those taking the decisions to proceed with re-entry. In the event, damage sustained to the heat shielding on one of the wings caused the shuttle to burn up as it tore through the atmosphere.

With this point firmly made, Tufte moves on to the information density of PowerPoint. If you've read Tufte at all (and I wholeheartedly recommend that you do if you haven't - try this or this), you'll know that he loves the elegant display of data. His books are full of slick maps of Japanese train timetables that merge the beauty of ukiyo-e with the simple brilliance of underground maps and neatly tabulated rows of data that at a glance allow the reader to compare multiple 1000s of pairs of data with nary a flicker of the eye. Tufte's point is that, with careful planning (and not a shortage of artistic flair in many cases) it is possible to fit extraordinary levels of comprehensible detail in to quite small spaces. Even straight forward words can achieve this effectively. Examples of text density he cites include comparing characters on a page (Guinness Book of Records, 4,600; NYT website, 4,100; BBC News website, 3,400; PowerPoint, 98-250) or characters per square inch (for the above, 162; 43; 36; 1-3 - though this will obviously become mere fractions once you project it on the wall).

Pravda - hardly a barrel of laughsIn another list of examples, PowerPoint rather damning comes only second to worst in amount of detail conveyed to Pravda, the old Soviet corporate adver-zine. Enough, surely, to have all but the most die-hard Microsoft fans scrabbling for the uninstall button?

Of course, Tufte is not alone in these criticisms. The "death-by-PowerPoint" meme has been around the block a number of times, and having myself witnessed meetings and training courses where the facilitator or trainer has read every word on screen, or plumbed the darkest depths of the standard Office clip-art files for the same image as I saw in the last course. Where Tufte stands out is at least he has some measured criticisms that make it easier to avoid the same mistakes - though I'll let you go to the source for these.

And so to the, "Yes, but..."?

These criticisms are entirely fair. Scientists at a major organisation like NASA should not have been encouraged to make all their reports in PowerPoint. The established method for, well, centuries, has been the written account or report. Simply because it is established doesn' mean it always has to be this way, but as Tufte forcefully argues, they are far more suited than a slideshow to the large quantities of evidence that scientific enquiry demands be provided (after all, peer-review can't happen without it).

It's surprising that even in a an organisation as august as NASA, the modern urge to boil everything down to soundbite, or elevator pitch had taken a firm grip to the point that everyone was scurrying around clutching sheaves of slide handouts. Surely with something as serious as spacecraft re-entry with a crew of talented, dedicated people on-board warrants careful consideration of all the facts - after all, for once it is rocket science.

I think that what was missed is PowerPoint is PRESENTATION software. It is designed to assist you in presenting something. It shouldn't be the thing itself. Which would appear to be the problem at NASA and is all too often the problem that other users fall prey to - the same problem that gets characterised as "death-by-PowerPoint". The presentation itself became the focus instead simply of the medium.

PowerPoint in the context of presentations is a quick, functional way to get images, graphs and key messages on a screen, to support your main message. It is not a place to write lengthy reports - it simply isn't designed to cope.

However, if you step back from the all-too-easy-to-adopt position of the "death-by-PowerPoint" crowd, and evaluate Microsoft's bete-noire as a simple, screen orientated page layout tool that even the most ardent technophobe can get familiar with, suddenly you can begin to evaluate it in a new light, and it this use that I will look at in my next piece: "...and when PowerPoint is right."

Post a Comment