Data visualization at the 2018 Library Technology Conference

I attended the session A Bigger Boat: Data Visualization Lessons from the Movies at the 2018 Library Technology Conference. Why should we care about data visualization? If you have ever had to put together a presentation for your funders, created an annual report, or tried to summarize results from a survey, you’ve probably had to think about how to display the data visually. Well-displayed information can be powerful and help answer questions quickly and easily.

The presenters defined the perceptual processing types Preattentive and Attentive. Preattentive processing is very fast and is the first type of processing that occurs, then our brain processes what is important. Attentive processing is much slower, and it is a sequential process that results in understanding and remembering. For example, try to count how many times the number 5 appears here:

9 8 7 3 4 9 7 8 5 0 0 2 8 3 4 7 5 2 3 3 8 5 4 8 9 6 5

This is too complex to process preattentively, and in order to count all the 5’s, we have to use attentive processing. Now count all the 5’s in this string.

9 8 7 3 4 9 7 8 5 0 0 2 8 3 4 7 5 2 3 3 8 5 4 8 9 6 5

You should have been able to see the 5’s easily and immediately, because I used bold font to make them stand out. This is an example of a preattentive visual attribute, color.

Preattentive attributes include form, color, spatial position, and motion. For form, you can use length, width, orientation, shape, size, and enclosure attributes to show data that is not like the other. Think of a bar chart, which is actually really effective in showing quantitative differences between data points using length. Preattentive attributes enables us to emphasize the most important information visually. Most of the concepts the presenters shared were from the book Show Me the Numbers by Stephen Few. There’s also a bibliography that was shared for more exploration on your own.

The presenters also touched on the Gestalt Principles. Often, even if there is no pattern, we're going to see one. These principles are organized into the following five categories:  Proximity, Similarity, Enclosure, Closure, Continuity, and Connection. Examples were given on how to apply each when displaying data. View the slides to see the examples.

My main takeaway from this session is that you don’t have to necessarily use fancy graphics to be effective in presenting your data visually. 3-D charts and graphs are not great, as our human perception of depth is generally weak. And if we understand more about how perception works, we can present our data more effectively.