Presenting Data and Information. by Edward Tufte San Fransisco, December 9. http://www.edwardtufte.com/tufte/courses author: Joost Geurts Overall impression: The people participating in the course (+/- 500-1000) had no common background, therefore the course was kept rather general and covered topics such as design, financial data and oral presentation. Because of the size of the audience questions could only be asked during the breaks for which you had to queue up. Tufte is an extrovert person likes to talk and present and doesn't have a problem with his ego. Quote: "Newton, unlike myself, was a shy and modest person." Although the course is quite general I found it nevertheless interesting and would recommend it. Tufte 'breaths' presentation and information visualization and because of that you learn 'obvious' things to him, which were not so obvious for you, or you'd never paid much attention to it. The material provided for this course were his three books: 1) The visual display of quantitative information 2) Envisioning information 3) Visual explanation: images and quantities, evidence and narrative 4) essay: The cognitive style of power point. During the course he was referring to pages in the books when explaining things. He didn't use slides and only used the projector to show a few images which were not in the book. note: most of what you'll find below here are statements Tufte made, personal comments are indicated. Principles of Design ----------------- Presenting information is not only visualizing data. You are presenting data with a reason. Make sure a presentation, in addition to being descriptive also shows causality. It should explain what we see and what you are trying to show. Interpreting a data visualization should not involve unnecessary decoding. This means no legenda of colors or symbols. If they are important integrate them in the image. Visualization allows compression. If images are on the same plane you can more easily compare them then when they are stacked. The most import aspect of every visualization is that it is content driven. The quality of content should be accurate and of relevance. Content should never be scarified for design. Principles of design are much like principles of analytical thinking (what, why, when). Document everything for the purpose of credibility. Who are you, when did you make it, what sources did you use, what are you trying to show. Relate everything to the universal grid. Financial Data -------------- Showing detail improves credibility. The more measurements you show the more likely your analysis will be correct instead of a possible anomaly in the data. Context influences interpretation. (I guess I don't have to explain this one) Most of the visualizations used for financial data are only descriptive. Annotate figures to explain what is special about them. Find designs that work and copy. Resolutions of computer screens are way to small for the detail required in financial analysis. Print tables on paper which has a higher resolution, smaller fonts are okay. Make posters and hang them on your wall for colleagues and to discover 'new' information which takes time to incubate. Interface/Web page design ------------------------- The computer screen has a very low resolution compared to printed paper. Therefore make the content space as large as possible. People visit your website for the content, leave fancy designs for posters (which are supposed to be pretty). Minimize the navigation space as much as possible. Don't use icons if text is shorter or more meaningful, and don't use them both. Because computer screens are low resolution devices, information on a web page often needs to be stacked. This is less favorable then presenting them on one plane. The metaphor used for design in corporate websites often is its management structure, which typically is hierarchical. The interface is therefore biased to a hierarchical structure and because we are using a computer the stacking of information is amplified which makes comparing and relating information difficult. Oral Presentation ----------------- jst: This was a subject Tufte really seemed to care about. He gave some examples concerning the Columbia and Challenger space shuttle accidents which could have been prevented if the evidence why they should cancel the launch was presented better. In case of the Columbia the bad presentation was a power point presentation which he analyzed deeply and wrote an essay about. The problem with power point is not so much that it is power point but more the style of presenting it dictates, which is, a title with a list of bullet points. The information presented is nearly always incomplete and needs to be backed up by oral explanation. Speech is an excellent medium to discuss or reason about a topic, but it is, in addition to not very efficient, also not very reliable to convey factual data. The only record the audience has on the facts is what is presented in the power point presentation. In this case not having access to the data and a bad presentation caused a decision to not cancel the launch. Tufte was making an argument not to use power point at all. Paper is far more efficient and reliable in transmitting information. People have a record of what you are saying. They are allowed to access the information in a way they prefer (reading top-bottom, skimming). Afterward you can directly come to the point you want to discuss not wasting time giving people context. Time for some bullets ;) Two main points about giving a presentation: 1) Show causality 2) Show all relevant data other points: *) Show up early *) Never apologize *) Structure: problem, relevance, solution *) Presenting tables: - explain a number in the table - explain the abstraction/contextualize - show your abstraction is correct by explaining another number *) Give everybody a piece of paper. *) Be careful with humor. *) Use gender indeterminate term 'they' *) finish early *) practice Conclusion: On the negative side: The crowd was too large to actual benefit from being there, they could have video taped it all and it would be more or less the same. On the positive side: I enjoyed the course! It was informative, I learned some new stuff. The books are beautiful and I got them autographed. :)