March 18, 2018

Disengage: Measuring Engagement with Etexts Could Cripple What It Means to Assess | LJ Insider

Fanged book drawing

Open source textbooks a “threat” to Texas education? by

I’m a writer, and a geek. So if CourseSmart had wanted to track students’ use of its etextbooks to improve the texts themselves, I could totally sympathize. But it seems to me that CourseSmart wants to use those analytics to fix, not the book, but the reader, and that has the potential to disturb privacy advocates and put students off etextbooks altogether.

As we recently reported, CourseSmart Analytics will track how much time students spend reading, how many pages they view, and how many notes and highlights they make, and this will get turned into an “engagement score” for each student.

The idea, according to Sean Devine, chief executive of CourseSmart, is that faculty members can reach out to students showing low engagement. But with a metric that does all the work for you like that, it would be all too easy for engagement to become, not just a red flag, but a requirement, part of a student’s grade. And that’s problematic.

Students shouldn’t be evaluated on how (or even whether) they engage with a given textbook, but with how well they learn the material. What if they read unusually fast—or slowly? What if they prefer handwritten notes? (And given the porousness of CourseSmart’s boundaries around issues of student privacy, who would want to risk taking notes within the software that are in any way critical of the text, the course, or the teacher?) What if they prefer other sources altogether?

Yes, we need to teach study skills—but at some point we also need to step back and allow students to use the methods that work for them, and not penalize them for what path they take as long as they arrive at the right destination.

Even if it doesn’t become part of the formal grading system, the analytics are likely to unconsciously influence the professor’s subjective assessment of the students’ finished exams and assignments. And we know from studies of racial and gender bias that unconscious prejudices can be perniciously hard to root out. I would rather see academia go the other way altogether, toward grading papers anonymously, like blind orchestral auditions, than give professors yet more information that they then have to try and disregard.

CourseSmart Analytics has an opt-out, but that’s not sufficient. Opt-outs can be tricky to find, and you have to know it’s a) necessary and b) an option. It’s not news that opt-in is the way to go, if you’re really serious about privacy. And it’s not just librarians that care about this stuff. Repeated kerfuffles on Facebook, among others, have shown that users, including the current undergraduate generation, may merrily choose to share what they had for lunch and how they feel about their latest breakup, but they want to be asked.

In this case, there’s more at stake: students would have good reason to fear that opting out would be penalized, unofficially if not officially—that a desire for privacy could become the cause of heightened scrutiny, if not just outright forbidden by their teacher, department, or school as a whole.

The other factor here is that etextbooks are, if not in their infancy, still in a rather fractious toddlerhood. However inevitable their adoption and replacement of print texts may seem, it hasn’t really been happening like that on the ground, at least not yet.

For those who want to see students make the switch to etextbooks in greater numbers—a group which presumably includes, at a minimum, CourseSmart—it just doesn’t make sense to teach students by example that etextbooks are going to present them with hurdles and hassles that print textbooks don’t.

Even the most diligent student is likely to want the option of having an off day—or even an off week or month—and catching up later without it being anyone else’s business. Part of the point of being in college is that students get far greater freedom to manage long term projects—and learn from it, including learning from screwing it up. If they can skip the headache of being spied on and second-guessed by buying a second-hand print textbook, why wouldn’t they?

And if this is their first, formative impression of etextbooks, are they really going to distinguish between this and the next etext, even if the latter doesn’t come with analytics? Or are they just going to figure the whole thing is more trouble than it’s worth?

If companies want to aggregate anonymous data on usage to improve their second editions, I’m all for it. But when it comes to CourseSmart Analytics, or any other non-private form of backseat-studying, I think it’s time to rethink whether what we need is really more engagement. A generation already criticized for not taking the initiative in part because of their helicopter parents doesn’t need to add helicopter textbooks and professors. They need us to give them some space—the digital version of a room of one’s own.

Meredith Schwartz About Meredith Schwartz

Meredith Schwartz ( is Executive Editor of Library Journal.