My very favorite thing about OpenEdX is the
quality of the assessment tools. They're robust, they handle a wide variety of tasks, and they're extensible through a variety of methods. I like that you can randomize them, that you can pull them in from content libraries, that you can twist them in various ways with python code. I have things that I think could be improved here, but assessments are edX's greatest strength in my mind.
I think the
clean look of the platform is an advantage. It's fairly easy to make new things look like they fit in.
The fact that edX
permits javascript and css is a HUGE benefit for me as a course designer. It lets me add small interactives to my courses easily, and has allowed me to create lots of tools that other project leads here at HarvardX have used. That kind of customization is not present in most other systems, at least to the same extent. My own talk at OpenEdX is going to be about the pop-up question tool I wrote (for asking questions in the middle of videos), and without this permissive setup I'd never have been able to write that.
(That reminds me: another thing I'd like to see in OpenEdX is course-wide js and css, as per the
Course-Wide Custom Javascript proposal.)
Inline discussion boards are pretty handy. They're so handy, in fact, that when I run a course where we initially say "Let's not use these", there's enough demand from the learners that we add them to every page.
The fact that we can do
A/B Testing is great. I'd love to see it be more flexible and more content-aware, but having it at all is so important for research.
The
documentation for instructors and for learners is outstanding.
There are other things that I like that are more on the philosophical side (the fact that it's an open platform, for example), but those aren't so much instructor-facing parts.
P.S.: I just remembered another item that's a must-add:
more granular permissions. Right now, everyone who can edit course material can also see student grades. Sometimes I don't want to train all my content developers in human subjects research guidelines. It would be great to split that role up.