Child pages
  • Rethinking Learning Activities - Taking Action on User Research
Skip to end of metadata
Go to start of metadata

Description Instructors have different needs for creating and managing learning activities (papers, problem sets, and online discussion) that depend on their institutional situation and discipline. In the past we addressed this by developing separate tools, forcing users to go to different locations to manage different activity and learn multiple interfaces. The emerging ethos of Sakai 3.0 is to recognize areas of commonality and make the capabilities of these former tools available in contexts that are more natural to users. But who are these users and what are these contexts? What are those areas of common functionality? And how do we address cases when users have very different needs? To develop this understanding, representatives at eight institutions interviewed 30 instructors, students and staff about the learning activities for their courses, regardless of the tool they used. We distilled what we learned into 12 persona, or user archetypes, representing different roles, disciplines, attitudes and motivations. We'll discuss how these personas can inform decision making about the product, describe scenarios these persona would likely have and walk through preliminary design sketches. All attendees welcome and thoseinterested are invited to a separate hands-on design session entitled "Rethinking Learning Activities: A workshop to practice research-based design."


Keli Amann, User Experience Specialist,  Stanford University

Daphne Ogle, Senior Interaction Designer, University of California, Berkeley


  • No labels


  1. I really like the results you've gleaned from the user research. I think it's basically correct.

    But as I look at some of the preliminary sketches of possible interfaces, is it possible to make it more simple and direct; certainly for instructors, but in particular for students?

    So on the "Denise adds a new assignment" stuff, for example, for sake of argument, what would happen if we throw out the notion of "assignment" entirely, and instead focused on simple, direct, student-centered, action-verbs; "tasks" like:

    • read
    • watch
    • listen
    • view
    • write
    • answer
    • etc.

    ... and rather than have big, real-estate intensive, buttons like "add reading", you simply use attractive icons?

    And I'm going to repeatedly say this so that it's as widely heard as possible: OAE needs a notion of student-centered learning outcomes in its core assessment functionality.

    So in this user story, imagine instead Denise:

    1. adds a title and a quick free-text description
    2. adds a handful of pre-defined (in the course template) learning outcomes that can be drag-and-drop reordered
    3. adds a rubric that allows her to tie assessment directly to the learning outcomes, both for the student and for her subsequent assessment of them
    4. the rest of the assignment definition is just adding tasks in a sequence (read, watch, write); each task has an optional due date, etc. (also, tasks need to be able to shared with groups for group-work)

    ... and all of this happens using direct, in-place, ajaxified goodness (no dialogs, no context-switching, etc.). Simple, elegant, flexible, direct.

    Moreover, all of this is equally clear, meaningful, and direct to students (it's clear what they are to accomplish and how they will be assessed, what steps they must take and when they're due). 

    Then when Denise gets to assessment (or even if there's an interim peer-review step), she can use the rubric to more quickly grade, and to provide more meaningful feedback to students. Further, that data can then later be acted on, aggregated, etc. by the program, the institution, etc.

    PS - I don't mean the term "assignment" in this vision would exactly be banished. It would just be a subtle distinction in the UI of a task/activity that happens to have a due date, and may be assessed.

    1. Reflecting on this a bit more, I see one major problem with how these (learning activity) tools have been designed in the past: they aren't learner-centered. E.g. to do this right, I think you'd want to match this instructor research with student research.*

      But on boiling down the essence of my proposal here, it is a very simple, but very powerful, learner-centered model, that is simply two things:

      1. topics (named, ordered, bundles of activities)
      2. activities

      But for activities entirely throw out how we've thought about this for the past ten years (no "quizzes" and no "assignments," for example) and just ask "what do students need to do?"

      So have an extensible subject-verb-object activity model with some core verbs (view, watch, read, write, reflect) and object types (text, video, image, writing field, etc.). You allow activities to have optional:

      • sub-activities
      • due dates
      • assessment status (it gets assessed in some way, or it doesn't)

      It's clear and simple for students and instructors alike, it's easy to implement (in the sense the model is simple), and it also has a secondary benefit: it's consistent with more widely used nascent standards like ActivityStreams.

      * Anecdote: my institution just deployed CLE site-wide, and my wife has just started using it. She noted that students are confused by a basic thing: where to find their readings. It turns out many looked in "assignments" (because they thought reading is a task for them to do), and had no clue to look in "resources."