Question bank hosted on a USER level rather than on a Course Level

guswib

Hi,

I just started working with question banks which should be accessible for different course. Clearly, it would be annoying to rewrite a question bank. In this process, I noticed that a question bank can only be hosted in a course shell, in contrast to for example files which can be hosted on a user level. This is a great limitation, as my course are not persistent and thus, it is possible that a question bank disappears. Further, although it is possible for course shells to share a question bank, students have complained that quizzes have shown up blank. So, the sharing does not seem to work flawlessly.

So, I would like to see a way to store question bank on a user level so that they become more persistent, and potentially more likely to be able to be used in different course shells without problems.

Anyone who have found similar issues, and/or have a good solution?

Best

Gustav

jeremyfw

Hi @guswib ,

You've put your finger on a real structural limitation in Canvas, and you're not imagining things. Question banks in Canvas are course-scoped, not user-scoped, which means they're tightly coupled to a course's lifecycle. When courses are concluded, archived, copied, or eventually cleaned up by the institution, those banks can break or disappear. The current "sharing" mechanism also relies on references back to the original course, which explains why students sometimes see blank quizzes even though everything looked fine during setup.

One way instructors work around this (without fighting Canvas's architecture) is to treat Canvas as the delivery layer, not the long-term storage layer, for assessment content, by using an external question-bank tool.

Full disclosure: I built Canvas Question Bank Platform specifically because I kept running into this problem myself. As a department head in an international school context, where teacher turnover is high and courses are routinely concluded when instructors move on, I watched years of assessment work disappear whenever course owners left. Shared quizzes would break, banks would vanish, and the next teacher would start from scratch.

What this kind of approach does is simple but powerful: it lets you store your questions outside of Canvas, at the user level, organized into reusable collections. When you need those questions in a course, you export them as a QTI package and import them into that specific course. Each import creates a local copy in the target course rather than a shared reference, which is exactly why it's more resilient.

That approach gives you some real benefits:

• Your questions persist regardless of what happens to any individual Canvas course
• Each course gets a clean copy of the questions, rather than a fragile shared reference
• You can update questions in one place and re-export them to new courses as needed
• If a course shell disappears or is reset, your source content is completely unaffected
• The time you invest in building good questions becomes a durable, reusable resource

What it doesn't do is magically create a true "user-level question bank" inside Canvas itself, Canvas simply doesn't support that natively today. Instead, it gives you a stable external home for your content that you intentionally publish into Canvas when needed.

A typical workflow looks like this:

1. Build or import questions once (from Excel, existing Canvas quizzes, or manually)
2. Organize them into collections outside Canvas
3. Export to QTI and import into whichever Canvas course needs them
4. Repeat for future courses, without any shared dependencies that can break

It's not so much a hack as it is a way to stop depending on course shells for long-term storage. Given the issues you're describing, non-persistent courses, disappearing banks, and student-facing failures, this pattern is often the most reliable way to protect your effort until Canvas offers true user-level question banks.

Cheers,
J