Open edX - Assessment Setup Improvements

The following potential improvements have come up as a result of this research (not necessarily tied directly to this area of configuration):

• Implement actual quiz or assessment handling for subsections

  • We currently have timed exams that allow staff to handle subsections as quizzes with a time limit, but no functionality that allows staff to create quizzes without a timer

  • Key functionality requirements of quizzes:

    • Quiz resets - Right now Open edX requires resetting a quiz question-by-question.

    • Quiz submission - Right now Open edX requires each question to be submitted question-by-question

    • Shared variables - Less common, but a request of many educators surveyed was to have variables that carry over between questions, which would require those variables to be associated with something, and I propose the quiz itself.

    • Problem defaults and settings - All of our settings exist at the course-level and problem-level, making quizzes time-consuming and irritating to create.

    • Problem randomization and content library linking - Quizzes in other platforms can commonly select a random assortment of problems from a pool of questions either present in the quiz itself or its associated bank.

• Improve the user experience of past-due graded content

  • In other platforms, closed quizzes can serve as learning resources which do things like display answers or explanations. In Open edX graded subsections, problems just stop being submittable, which is a potentially confusing experience.

• Allow the existence of freeform assessed objects not associated with content

  • In most platforms, there is a way for staff to manually grant learners grades for off-platform tasks, attendance, discussion participation, but in Open edX you can only grant grades associated with content that exists in a graded subsection.

• Augment discussions with grading functionality

  • In academic courses, discussion participation is often used as a graded requirement, but there is no simple way to do this on Open edX. Freeform grades cannot be captured without creating some kind of hidden dummy grade component or XBlock. Even with these, it is difficult to track participation by individual learners. This could be automated (grant grades if the learner makes X posts in Y period), data on discussion participation could be exposed through a reporting frontend, and learner submissions could be aggregated and summarised in some kind of graded component.

• Make grading even more flexible

  • Open edX’s grading flexibility is simultaneously its biggest strength and one of its biggest weaknesses. We allow formative and summative content to be mixed in with learning content, and yet we lack true flexibility in grading, as we cannot assign individual components or units to the grading criteria. Much as “quizzes” should exist as assessment containers for basic usage and configuration, we should make it possible for content to be assigned as graded at other levels.

In general, Open edX’s assignment setup is incredibly light compared to other platforms, which have a much more rigid structure where a quiz is a quiz, and that’s where the graded content lives, and the assessment settings reflect that.

Open edX’s ability to have graded content anywhere flexibly is absolutely a huge boon that we should lean into, but it shouldn’t come at the cost of the existence of properly contained assessments with associated management tools.