I recently spoke to a class of computer science students studying big data at Georgia State University. They were just diving into data visualization and many of them had no previous training in data visualization best practices or concepts. We covered not only what data visualization is and good and bad examples of it, but also ways to manipulate charts and graphs to get the best visualization outcome.
Seeing is believing
One of my favorite examples of this is an award winning visualization Vaccines and Infectious Diseases http://www.tynandebold.com created by Dov Friedman and Tynan DeBold. It uses a heat map to effectively translate time, location, and intensity into a story about the impact of vaccination.
People need more than facts to make decisions or understand concepts. Data Visualization adds structure, simplicity and a visual story that humanizes information. The structure of a data visualization should organize and add context to large sets of data as well as reveal patterns, trends and relationships. The ability to reduce noise and distraction by filtering what is displayed so the key findings are shown creates a simplified and more focused message. Data is impersonal and people respond best to data visualizations with a human element according to recent research by Harvard (https://www.seas.harvard.edu/news/2015/11/making-visualizations-more-memorable). Visual design can create a memorable and easier to comprehend experience.
Great data visualization design is critical to getting (and keeping) people’s attention
Where is it useful? Everywhere.
Data visualization began as a tool primarily used by journalists, scientists, and large corporations. The ever increasing power of computers and database systems puts the ability to create them in reach of almost everyone. Examples include:
- Showing datasets of brain scans at different stages of life
- vaccine adoption and its impact over time and by location
- reporting dashboards that show real time alarms and trends in enterprise applications
- tiny applications used by fitness wearable devices
- visualization of in depth reporting by journalists.
The audience may vary and should always be considered but people at all education levels are so immersed in data being visualized on their devices that they expect the same high quality and simplicity of use in all their experiences.
Data visualization vs. Scientific Visualization
Scientific Visualization involves the creation of visual models of data sets. An example of this would be weather patterns over time. Data visualization goes further and adds human interpretation. It is possible to do a bit of both especially for more serious uses like [stet] journalism and scientific reports. The broader the audience though the more likely the additional context possible with data visualization will be useful.
People are subjective
Communicating with an audience externally requires more context around information. People are busy, distracted, and want the big picture. You cannot control how people will interpret data anyway – but you can frame the story, provide sources and give them tools to explore.
What makes a great data visualization?
We’ve already seen one example – the Vaccines and Infectious Diseases data visualization I shared earlier in this post. There has been an increased interest in what makes effective data visualizations, and we can break them down into several key best practices. Five keys to a great data visualization are:
- Easy to “get”
- Nice to look at
- Fits the context
I’ll leave you today with a few more examples of great data visualization. Where do you struggle with data visualization?
A World of Languages – This infographic breaks lots of rules, but is a very human approach to visualizing a concept.
Cisco’s Annual Security Report is full of great charts and visualizations: www.cisco.com/go/asr2016
The award winning Creative Routines infographic: http://www.infowetrust.com/creative-routines/