Conversation Platform
Spring 2019-NYU
In Fall 2018, NYU IT and the Provost’s office collaborated on the development of a new “Conversation” tool for the university’s Learning Management System (LMS). The goal of this project was to modernize the online class discussion board, a historically dated tool that affords few opportunities for online communication. This new, reimagined tool would provide a clean experience reminiscent of other modern applications that faculty and students use outside of academia. In addition, the tool would provide faculty with the ability to facilitate different types of student interactions through the use of different interaction models.
Project Overview
Need
Develop a new discussion and conversation tool for the Learning Management System that would provide a clean, modern, streamlined user experience. Provide faculty with the ability to select between different modes of conversation, or “Interaction models,” when creating new topics (e.g., responding to an initial post versus a more free-form discussion) and deliver different user experiences based on their selection.
Solution
A new “Conversation” tool was designed, tested and rolled out to pilot participants in Spring 2019.
Role
Product Owner, UX/UI Designer
Task
Gather stakeholder requirements and manage product backlog. Design tool workflows and interface, from low-fidelity wireframes to high-fidelity clickable prototypes. Conduct usability testing and gather feedback. Iterate and work with developers to implement a minimum viable product for launch.
Tools
Sketch, InVision, JIRA
Approach
Empathize
Requirements Gathering
Kyle met early and often with stakeholders for the project to define a clear set of requirements for the tool. Most critical was the idea of “interaction models,” which would allow for the creation of discussion topics with differing functionality based on the pedagogical need (for example, the need for students to post responses to an initial prompt, or for students to brainstorm, upvote, and come to consensus around their own ideas).
By providing a set of differing configuration presets tied to different interaction models, the tool could present different experiences depending on the selected model:
InVision Freehand document showing how different elements of the interface could be rearranged to accommodate different types of topics.
Tool Requirements
- Interaction Models - Faculty would be able to select from a variety of conversation modes that they would like to facilitate for a given topic; depending on the selection, the resulting topic would have a specific layout / combination of features
- Customization - When setting up their topics, faculty would be able to customize the features available within them (e.g., upvoting, moderation, etc.)
- Inline Posting / Editing - To ensure a streamlined and modern UX, posting/editing within the tool would be done inline, with auto-saving and dynamic loading to minimize disruption
- Timeline - To make it easier to navigate long threads, a persistent timeline would be available within topics. This would allow you to jump to specific posts. When grading, it would allow you to jump to other posts by the same student.
- Granular Notifications - To make it easier to handle notifications for different topics, implement the familiar subscription / muting behavior found in modern apps such as Slack, Google Chat, etc.
User Research / Validation
Kyle, along with the university’s Usability Lab, conducted usability testing with a total of ten faculty members using initial low-fidelity mockups. The purpose of this testing was to validate the initial “interaction model” proposal, to determine a) if the conceptual model made sense to them, and b) whether they would be likely to use said tool.
Based on test participant feedback, 7 out of 10 faculty were likely to use the tool, while 5 of these were very likely to use it.
User feedback: Likelihood to use the tool
When asked how easy or difficult to use the interaction models within the tool would be, 8 out of 10 faculty felt they would be easy to use.
User feedback: Ease or difficulty of use
“The new design looks more fluid. It’s helpful because it will help me to model the conversation for students.”
Scope Definition
Once the tool’s requirements had been validated, Kyle created a prioritized product backlog and scoped out the necessary requirements for a minimum viable product. He worked closely with the project’s developers to ensure that these features were viable, and relayed this backlog back to the stakeholders for approval before beginning initial design/development sprints.
Design
Iterative Designs
With the scope and requirements for a minimum viable product defined, Kyle proceeded to create a series of mid- and high-fidelity designs for the tool, iterating based on end-user and stakeholder feedback.
Interaction Model Selection UI
Early Iteration
Final Design
“React” Topic UI
Early Iteration
Final Design
Deploy
Production Rollout
The resulting tool, designed to be fully responsive, was released to as a pilot tool within the LMS in Spring 2019.