This paper describes the fundamentals to focus on before we build a visual representation of information. The paper addresses the fact that the process of creating a good visual representation of data is not random but rather precise and scientific. Each step of the visual representation selection and design is presented concisely, yet proficiently. The author has taken pains to illustrate his points with wonderful graphical examples, perhaps emphasizing on how effective visualization can be, if they are planned and constructed correctly. The seven common relationships of quantitative data and utilization scenarios for basic 2-D representational visualizations like tables, points, bars, line graphs and boxes are well presented. Sticking to light and natural colour schemes rather than varied, dark colours that can overwhelm readers and the importance of appropriate positioning of different elements of the visualization are some good points made by the author.
I disagree with the authors on a couple of points.Firstly, though the author defines the difference between nominal, ordinal and interval scales well,one example of his regarding the months of the year being an interval scale is a bit dicey. In my opinion, a time scale could also be viewed as an ordinal scale (a precise ordering among the months without a quantitative unit being attached to the months).It is a matter of perception,after all.Secondly, I disagree with the author's view on novel ,unconventional visualizations. The author starts the paper with an office story detailing essentially not to try something fancy as it may lead to misinformation. This may be true in many cases, but it must be acknowledged that dramatic, novel representations may sometimes be needed to deliver a strong message. (Example: Florence Nightingale’s rose diagram).
0 comments:
Post a Comment