Skip to main content

The Study Hall

      Previous Page

      Now Open: Dataset on Canvas Network

      Like we’ve always said: Good things come to those who’re open. So when Harvard and MIT released a dataset based on their MOOC offerings, we decided to give and get good things by releasing our own dataset for Canvas Network.

      Now Open: Dataset on Canvas Network

      It’s been three years since our very first courses, including David Wiley’s aptly themed “Introduction to Openness in Education,” went live on Canvas Network. Since then, we’ve partnered with educators, institutions, and organizations around the world to deliver more than 500 open courses to 665,000 learners—nearly half of whom reside outside of North America.

      All that open learning generated a lot of data. So in the spirit of openness and for the good of education, today we’re announcing the release of our dataset via the Canvas Network page on Dataverse.

      This dataset provides a snapshot of more than 230 Canvas Network courses offered between Jan. 2014 and Sept. 2015. It aggregates and anonymizes person-course data following a schema and methodology similar to that used by Harvard and MIT, which, if you really want to nerd out, is based on k-anonymity. Because the data is necessarily limited by de-identification techniques, a more robust dataset (based on the Canvas Data star schema) will also be available to academic researchers who request access and agree to our terms of use.

      To see the data, request additional access, or learn more about the project (a.k.a. our good things in the form of open course data) visit

      Keep learning,
      Jared Stein
      VP, Research and Education