Designing an analytics dashboard that helps teachers support language learning
As a Product Designer at STORYWORLD, I was responsible for developing the analytics dashboard from start to finish - turning a teacher's napkin sketch into design ideas, iterating on visual and interaction design, prototyping, testing, and launching. Along the way, I worked with with software engineers, educational researchers, and bilingual teachers for critique and feedback.
Don't just show data. Focus on their utility and usability. Provide actionable insights to help teachers answer specific questions they have and better their lessons, instead of overwhelming them with meaningless student analytics.
Teaching students to read and speak in a new language is hard, especially when students come from diverse backgrounds and have different learning needs. My goal was to account not only for the language teachers’ goals, but also the nuanced ways they interact with technology, data, and students from different backgrounds.
Needfinding with teachers
We interviewed a number of bilingual teachers and researchers. To get authentic feedback, we framed the discussion in such a way that focused on the educator’s needs rather confining her input to the visual context of my design. The open-ended session allowed her to introduce concepts which became so critical in the early development of the teacher dashboard.
After the discussions, I discovered that they needed insights into the following questions:
• Who is the falling behind in class and needs support?
• What particularly words and narratives is my class struggling with?
• How can I use this information to plan my lessons?
Real learning contexts
Framing our design in the context of user personas helped us to dig deep into the most compelling and realistic use cases. We came up with four different types of learners based on performance, background, and needs/challenges in language learning: the high achiever, the average learner, the slow reader, the low achiever.
By seeing how the data we present might correlate with real-life scenarios, we were able to visualize how the analytic dashboard will be useful to teachers.
In developing each persona, we asked the following questions:
• What performance results are expected to look like?
• In what cases will the data have impact on the teachers’ decisions?
When running into design roadblocks, I asked myself how my solution is addressing these three following questions:
• Are we measuring and reporting the right things?
• What is it that teachers should be focusing on?
• Is it a simple dashboard?
Features <> Use Cases
Navigation Flow Maps
I considered the navigation flow in terms of the STORYWORLD platform as a whole, not just the dashboard itself. For example, teachers must be able to access other STORYWORLD like the “Home Library” and manage their user account on top of their class account. After I introduced the “Account Navigation” bar, I found that while many layers of navigation bars offer clear organizational hierarchy, they can diminish accessibility. So I eliminated superfluous navigation items within the dashboard.
After the first few iterations, we decided to show the mock up to a Bilingual educator to see if her reactions validated our direction. My biggest concern was that the default dashboard view would appear too visually overloaded. We received positive feedback in terms of visual design, but found that we didn’t have a clear criteria for selecting which data is important to teachers and worth visualizing.
My interview with an instructional designer confirmed this doubt. She raised the question of user-persona. Are we designing for teachers or educational specialists? Teachers are more likely to be more questions and problem-driven, as they want to know how learning results can be converted into actionable lesson plans. Specialists, on other hand, might be more interested in the data points and patterns they reflect.
Questions First, Data Second
Understanding what each data point is actually measuring was critical to our process. By identifying how the data can help facilitate the teachers’ questions, we are able to make better decisions about what data matters and needs visualization.
We asked ourselves: what questions do these metrics help teachers answer? What questions might they be asking in the first place?
Usability and Utility of Data
Instead of presenting just facts and figures, our design needs to offer immediate actionable insights that help teachers solve their problems. For example, instead of displaying the number of words each student clicked, we show top ten words and phrases the students are having trouble learning. This way, the teacher has a better idea of what vocabulary to review in class or use to design their lessons.
After the implementation of the prototype, we continued to test and refine the original design of the dashboard. We received very positive feedback from teachers and our partner research institutes. In a pilot study in three first grade classrooms, the researchers will examine whether the prototype functions as planned, is easy to use, and provides information teachers can understand and use to inform their language and literacy instruction.