Neil Oliver

I appreciate the language and tone used throughout Healey’s introductory chapter from ‘Data Visualization for Social Science’. Each discussion topic is well defined and supported with references while not hindering the reader by the use of overtly technical language.

The separation of aesthetic (bad taste), substantive (bad data) and perceptual issues (bad perception) is extremely useful. The topic of both substantive and perceptual issues in particular have personally been my biggest considerations when entering the world of data visualization. While creating beautiful work and finding the hidden story within data is of course a key goal, I am always concerned that my choices in visualization method may in fact tell the wrong story, and the consequences this may have. The continuation of this topic in the honesty and good judgment section shows that there is not a ‘one rule fits all’ approach, but that simple considerations such as the inclusion of a zero baseline can sometimes help to keep us honest.

The advice of ‘simplify, simplify’ encourages me to think backwards in my design process. So often I have found myself considering what additions I can make improve my visualizations to make them more appealing or to make them stand out; however in Healey argues that I should be considering what my graph can do without and what may be hindering the users interpretation and understand of the data. I feel that this would certainly be an interesting and useful focus point in relation to the classes latest time-based designs.

Many of the points that Healey make could be considered simply common sense; however, there are so many individual and solid questions and considerations such as why use ‘an area to represent a length?’ that the paper would be useful in reflecting on almost any visualization creation process. Personally, I would love to see these questions extracted into a checklist form to be used more readily when creating my own future designs. Healey does however still challenge my own assumptions and practices. The example that stood out to me is that of color. I often use color as a way to distinguish between different values and I would have (previously) said that this is a more effective method than a monochrome approach. The examples by Ware (2008) clearly show how wrong this assumption can be and also made me reflect on how a monochrome design may also negate some of issues surrounding the many different forms of color blindness; potentially making my designs more accessible.

Overall, the paper was personally a lot of ‘food for thought’ and something that I hope to refer back to when creating future designs.