Lessons from the Canvas Alpha Test: Deepening the Instruction of Deep Reading

By jeremydean | 8 December, 2016

Canvas Panda presents the Hypothesis app
Canvas Panda presents the Hypothesis app

We’re wrapping up the first semester of testing on the new Hypothesis Canvas app in about a dozen classrooms around the country. We’ve had a great group of alpha testers who’ve bravely experimented with this prototype and offered us invaluable feedback that has already informed our development. Here I want so share one of my major takeaways from the alpha testing phase that was a direct result of this kind of authentic collaboration between the team at hypothes.is and practicing classroom teachers and students.

Our original intent with the Canvas alpha app was just to make Hypothes.is annotation client native to the Canvas environment. We had to do this because, generally, if you want to play in the Ed-Tech Playground, you need to play well with the LMSs. We started with Canvas because of a specific problem there: hypothes.is could not be properly activated on PDFs hosted in Canvas because the LMS currently presents PDFs using Box’s viewer.

Quite early on in our hacking–mostly done by my more technical colleague Jon Udell–we got some great feedback from a teacher-technologist, Chris Long. Chris said he wanted to be able to see an individual student’s annotations on a text in isolation. That view is actually something hypothes.is itself didn’t do very well at the time: our activity stream could show all annotations by a user, but it was difficult to focus in on a user’s activity on a particular document. (This will change once we release our new activity page feature in the next couple of weeks.)

Jon, who’d been playing with various prototypes for searching and exporting annotations as part of our Hypothes.is Labs efforts, had experience using the API to pull and display sub-sets of annotation activity. So we decided to make the annotatable documents in Canvas–either PDFs or web pages–“assignments” in the language of the LMS. This allowed students to “submit” their individual annotations on a collaboratively annotated document for an instructor’s viewing.

A teacher’s view of a student’s hypothes.is annotations in SpeedGrader

This feature is currently unique to our Canvas app for two reasons: one, there isn’t such a view for a user’s activity on a document, including both annotations and replies, in the main client (not even in our new activity pages feature); and, two, thanks to the Canvas SpeedGrader app, educators can now offer private feedback on annotations, and grade them if they want to. So a student-user’s annotations are actionable for a professor in a way they are not using the client. Being able to respond privately to student annotations is actually a major request by educators using hypothes.is and a feature idea further validated by the positive responses to our Canvas functionality.

I’ve been thinking throughout the semester that this submission and feedback feature is the key to the future of Hypothes.is in education in several ways that I’ll outline below.

First, from a market perspective, online homework is a big one in ed-tech. Making annotation a homework assignment makes Hypothes.is something users might actually pay for down the line. Of course, there will always be a free version of H. But perhaps if an instructor or institution wants to sync annotation activity with a digital grade book, we offer them a premium version, purchased by students as part of a course or by an institutional site license. In this regard, we might think of hypothes.is as a kind of iClicker for homework, increasing engagement for homework assignments as iClicker claims it does for in-class time.

An annotated iClicker
An annotated iClicker

But making annotation homework isn’t just a business decision. When I taught high school English, colleagues who required students to annotate in traditional paper books would actually have a physical completion check in class. Students opened their books and teachers went around the room inspecting their marginalia at a glance. With Hypothes.is, holding students accountable for their reading and annotation can be done much more efficiently and robustly. The fact that hypothes.is can be used to hold students accountable for having done the reading is a big deal–it’s a frequent complaint of professors that students are simply not doing so. But it can also be used to help students develop their annotation and thus reading skills.

Through annotation homework assignments teachers can also see more precisely how students are highlighting or annotating, and further offer specific feedback on that practice. In this way, Hypothes.is not only continues to encourage the age-old literacy skill of marginalia, but makes it something that can actually be instructed. As a teacher, I might be able to advise a student to be more discrete in their text selection when highlighting or to more explicitly tie textual evidence into their marginal commentary. When I taught English, this type of granular feedback on close reading practice wasn’t given until a rough draft of an essay was submitted–or perhaps in class discussion if we were lucky. Indeed, such a window into reading and thinking habits might even be better a better indicator of a student’s learning than a final essay or other summative assignment–a point NYTimes book critic Sam Anderson makes outside of the context of formal education in one of the best essays I’ve read (and annotated) on the power of social reading and collaborative annotation.

Unhelpful highlighting
Unhelpful highlighting

Being a long-time academic annotator myself, I remember buying used books in college and riffling through the copies on the shelves to find ones that were either unhighlighted or had been, what I’d call, “thoughtfully annotated.” I’d immediately toss books back that were over-annotated–pages and page of endless “highlighting”–in which the reader seemed to have no sense of deliberation in their process. And I’d reject books that were sparsely annotated–maybe some notes on the first few pages and then nothing else, as if the reader had lost interest and never finished the book. In retrospect, these latter copies would have been serviceable for my purposes, but I compulsively wanted my books to be artifacts of my own systematic process of reading and analyzing texts, a resource I could pull on in other assignments.

A helpful annotation rubric
A helpful annotation rubric

Hypothes.is has the potential to provide teachers and students such a systematic sense of the process of reading and analyzing for every member of a class. The benefits of annotation have been discussed in education circles for a long time–here’s great recent piece at KQED. Kelly Hutchison gives her students a handy bookmark that elaborates on the ways that one might annotate a text. But with the added functionality that we’ve discovered in the development of our Canvas app, a teacher can use Hypothes.is to actually can see if students are doing those things and guide them through the process of developing those specific skills.

Also published on Medium.

Share this article