Welcome to the future of education where your textbook—and therefore your teacher—knows exactly how much of your homework you did, how, and when. The idea is that this data will improve outcomes, helping teachers better understand their students habits. And with CourseSmart teachers can do just that, as The New York Times' David Streitfeld explains today. Through what is called "the engagement index" a professor can track the a student's study habits, by looking at what pages of the book were opened, when, and how they took notes. "It’s Big Brother, sort of, but with a good intent," says Tracy Hurley, the dean of Texas A&M's school of business, which has started testing the technology along with eight other schools.

While helping professors understand why certain students are struggling is a good thing, as we've learned repeatedly, it's difficult to predict how people will use data. Take this example of how the CourseSmart data lead to suspicion of a student who had good grades:

Adrian Guardia, a Texas A&M instructor in management, took notice the other day of a student who was apparently doing well. His quiz grades were solid, and so was what CourseSmart calls his “engagement index.” But Mr. Guardia also saw something else: that the student had opened his textbook only once.

Guardia told Streitfeld this was a cause for concern: "Are you really learning if you only open the book the night before the test? I knew I had to reach out to him to discuss his studying habits.” Of course, students, especially at or above college level, don't all learn the exact same way. By the traditional outcome measurements (i.e. grades) and even the newfangled CourseSmart "engagement index" (an example of which you can see here), this student was doing fine, but the new granular data allowed a teacher to be skeptical about that student's work.  

The introduction of new subjective judgments already, understandably, has students worried. Hillary Torres with good grades, but "low engagement" scores, told Streitfeld: "If he looks and sees, 'Hillary is not really reading as much as I thought,' does that give him a negative image of me? His opinion really matters. Maybe I need to change my study habits." But unless instructors want to change how they evaluate students, why should she change the methods that are working for her? 

Even worse, another good student had to sit down with her professor to fit her study habits to some Silicon Valley company's textbook. In the Times comments, Charlotte writes: "I never studied the way the program thought I should. I took notes on paper, read only what I thought was important, and used outside resources to study." All of these would not be captured by the CourseSmart metrics. "By studying the way I thought was best for me I did well in the class, and finished with an A. I ended up with a very low engagement score which prompted my professor to set up multiple meetings with me over the semester about my supposedly poor study habits, and how he could help me. I didn't need help. It was all a complete waste of time."

These aren't the best case scenarios of course. In an ideal world, the aggregate data would help textbook authors to create better learning materials and give educators insight into how much these very expensive supplemental materials are actually helping the learning experience. But since when has data ever been used just for good?