10 Tips for Data Visualization by David Steier, William D. Eggers, Joe Leinbach, Anesa Diaz-Uda.
From the post:
After disaster strikes or government initiatives fail, in hindsight, we see all too often that warning signs were overlooked by decision-makers. Or sophisticated technology was installed, but nobody took the time to learn to use it. It’s often labeled “user error,” or “problem between keyboard and chair.” In analytics, the problem is especially acute — the most sophisticated analytics models in the world are futile unless decision-makers understand and act on the results appropriately.
This problem often arises because designers haven’t truly considered how those using the fancy dashboards, maps or policy visualizations will interact with the analytics. They may become enamored of the model’s power and try to fit every piece of data into it. However, in offering more options and parameters to control the model’s operation, and filling up every pixel of screen real estate, designers can fail to recognize that most government decision-makers are inundated with inputs, pressed for time and can only focus on essentials. As Yale professor and information design guru Edward Tufte wrote, “Clutter and confusion are not attributes of information; they are failures of design.”
In many cases, users can’t answer basic questions like “What should I pay attention to?” and “Now that I’ve seen this, what should I do?” If the answers aren’t readily apparent, the interface and analytics aren’t solving a problem — rather, they might be creating a bigger one. As one federal executive said recently, “No tweet stops bleeding.Unless something has actually changed, it’s just information. What pieces of data are actually going to help us make a better decision?”
Agencies should consider a more user-centric and outcome-centric approach to analytics design to visualize policy problems and guide executives toward better, faster, more informed decisions.
The good news is that government leaders are getting serious about making sense of their data, and constant advances in graphic, mobile and Web technology make it possible to translate “big data” into meaningful, impactful visual interfaces. Using visualization tools to present advanced analytics can help policymakers more easily understand a topic, create an instant connection to unseen layers of data, and provide a sense of scale in dealing with large, complex data sets.
Read the post to pick up the ten tips. And re-read them about every three months or so.
I am curious about the point under letting users lead that reads:
If enough users believe an interface is unsatisfactory, the designer is well advised to accept their judgment.
That runs counter to my belief that an interface exists solely to assist the user in some task. What other purpose would an interface have? Or should I say what other legitimate purpose would an interface have? 😉 (Outside of schools where interfaces are designed to educate or challenge the user. If I am on deadline with a project, the last think I want is an interface that attempts to educate or challenge me.)