LabXchange 2018-06-14

Owned by Ian Saunder
Last updated: Feb 08, 2021 by Nimisha Asthagiri (Deactivated)


Desired Outcomes

  • Review edX and LabXchange user stories for:
    • Library Catalog
    • Pathways & Pathway Clusters
    • Classes & Grading
  • Versioning use cases - Ian Saunder's findings from interview with educators:
    • Editing video unlikely to happen much, so we don't need to support editing (maybe edX will need to support editing of transcripts or cropping video but not LabXchange)
    • Mostly educators would want to edit text/html content or problems
    • Don't need to support uploading video directly, just referencing external videos like on YouTube, Vimeo.
      • LabXchange is the publisher organization, creating content. Need to work with any content coming from partner sites, like Khan Academy. Don't want to host external content.
    • Only Text needs to be editable. Other content does not need to editable. Copyright infringement is therefore low since it's just text.
    • How do we track content modifications?
      • We don't want the catalog to be cluttered with many minor variations of the same content. (i.e. we don't want to show all the "forks" of any piece of content)
      • Preferred approach: in the Library/catalog, only show the original, but when clicking on the original, show the derived variations of the content. So we need to track lineage.
        • Not all derivatives are published back. Some can remain private (by default).
    • When the original gets updated, how do those updates get pushed out? Do content re-users have a choice?
      • Strongly opposed to automatic update of content.
      • Also want to make sure that "an update is available" notifications are easily dismissed or hidden. Don't make them intrusive.
  • Accounts
    • Users may register or may not register. 
    • Teachers have separate accounts.
      • What is the verification process for getting an "educator" account?
    • Concern: students may look up the past problems created by their teacher, to memorize them without learning, in hopes those same problems will show up in a future exam.
      • Problem sets that are uploaded by educators should be discoverable by educators only.
    • Other concern: instructors mass-uploading copyrighted problems from a textbook.
  • Rating content
    • May want to be able to give a star rating to any piece of content. Could be political issue or demotivating if we support that though.
    • Best approach is probably an only-positive option, allowing users to mark their favorite content
    • Alternatives: advanced search options and sorting by usage
  • Tagging
    • Do we allow free-form tags from users? TBD
    • If so, are those tags public?
    • Namespacing tagging taxonomies and allowing different ACLs for each namespace may allow us to support different requirements for tagging.
  • Grading
    • Do we track multiple attempts on a problem?
    • Do we want to let a user know if they've tried a problem before, or do we always reset it?
  • Content re-use
    • If a user sees a unit like "Into to PCR" in one pathway, then later takes another pathway and sees the same unit in that pathway, it should not remember its state. The user should work through the unit in the new pathway as though seeing it for the first time.
      • Dave: we may want to separate context from pathway, so pathway references a context, but they're not always 1:1? TBD.
    • Nimisha: a possible simplification: Any time you see a Unit in a different place, it always has new state. The only time when state is shared is when it's explicitly configured and referenced by content authors.  It seems shared state is the less common case by far.
      • As a future enhancement, if shared state is a desirable default for a specific block, the course author can override the default for that block.
  • New: For LabXchange, we'll only need LabXchange admins to create clusters, not users.
  • Units versus Blocks
    • Can we say only Units are discoverable, reusable, etc?
    • In the UI, it can seem that you're dragging a single block into a pathway, but in the backend we can wrap a single block into a very lightweight Unit.
    • Units can then be the granularity for access control, licensing, discoverability, etc.
  • Adaptive use cases (Pearson's terminology)
    • Block level: scaffolding and adaptive hints ("adaptive content")
    • Unit level: mastery of a learning objective via practive problems or for benchmark tests ("adaptive assessments")
    • Sequence level: personalized pathways ("adaptive sequences")

Notes