Design research for user experience
Wesley Lindamood, senior interaction designer on the NPR Visuals team
During what has been an, um, interesting debate cycle for this, um, interesting presidential campaign, separating fact from fiction has been more complicated than ever. National Public Radio sought to make sense of the noise coming from either direction with its innovative, live debate fact check. Here, about 20 reporters live-annotated the debate, correcting and adding context to candidate statements. Lindamood shared how his team planned, executed, and adapted their idea. The following strategy is how the team approaches its work.
Lindamood suggests the following order of work:
Define your goal first by asking: What do people want to know? What do you need to give them? In the case of the live fact check, the goal was to provide users with critical fact checks, analysis and context to help interpret and make sense of the debate. The focus is on what, not how, at this stage.
What form should the content take? Do a competitive analysis. That is, look at other, similar work being done by other outlets, research how people currently get that information, etc. Lindamood said he uses a collage technique, pulling everything he found during his competitive analysis to look at the commonalities of the different projects.
During this phase, there is much team brainstorming. His team meets regularly have in-depth conversations around designing the project. This is the “how will we do this to meet our goals” layer. For the fact checks, how could and would they verify claims, identify new issues, call out general campaign themes, present annotations alongside a full transcript?
How does all the reporting fit together to meet the goal? At this level, the team brings its research, reporting and “actuals” to the design team. The question is how will we present this? For the fact check, the question became “How do we differentiate between the transcript of debate and the fact check?” Discussion of design structure included adding visual differentiation through color, attribution methods, etc. Had to think about an embeddable widget that could be shared. It’s about the presentation and trying things at this point.
Action at this point turns to building out and testing a model of what has been decided. Things like how does the design of the project behave, does the interface design work. For example, the skeleton layer testing revealed that fact checks could be added at any point in the transcript, but not in chronological order. The team built a notification system so the user could easily identify what had been read or not read
Action at this layer turns to the “skin” of the project: colors, fonts, etc. Lindamood pointed out the risk is jumping in too early to the surface layer and not starting with substance. You must build user experience first, not design something to be shiny. Move beyond your surface assumptions.
What are some of the implications for The Shorthorn? How can we use this process for both digital and print design of stories and engagements?