What would online higher education look like if it was designed using some of the principles behind Google, Yahoo, and Facebook?
Two years ago, leaders at the University of Phoenix decided that its software for students was outdated. So it hired tech-industry heavyweights from Yahoo and elsewhere, installed a team of more than 100 people here in San Francisco, and gave them free rein to rebuild the college’s online-learning environment from scratch.
The team created a social network that borrows heavily from Facebook. It developed a data platform that collects and analyzes billions of clicks, messages, and interactions among students and their instructors. And it started profiling students’ online behavior to personalize how they are taught.
That kind of data-driven approach is standard in Silicon Valley, but efforts to apply it to how college students learn are in the early stages. “The gap between the consumer Internet and online education, whatever school it is, is pretty wide,” says Satish Menon, who left Yahoo two years ago to become senior vice president for product strategy and development at the Apollo Group, owner of the University of Phoenix. “We want to close that gap.”
The Phoenix platform, which is being rolled out this year, is just one of many efforts to create software with a vastly better online-learning experience by building a community that responds more intelligently to student needs. Publishers, start-ups, course-management companies, and other colleges are working on similar ideas.
Such efforts face vexing questions: Can online communities that work well for socializing be adapted for sustained learning? Can course content be systematized so software can intervene when a student is struggling to learn a concept? Will students resist giving away too much personal information?
Built-In Audience
Phoenix has advantages in coming up with answers. Its software has a built-in audience of about 440,000 students. It is less restricted by previous versions of its own software than is a vendor like Blackboard, and less restricted by faculty governance than most nonprofit colleges would be.
And it has a great deal of money to spend.
Mr. Menon and a few others showed off limited glimpses of the company’s new platform for the first time in an interview here last month. Elements of the software have already been introduced, mainly to business students; the company plans to roll out more pieces of it this year.
When students log in, they see recommended tasks for that day and a personalized discussion feed that resembles one pioneered by Facebook. They can see who else is online and chat with other students and instructors.
One goal is to better help students find the right people among Phoenix’s vast network who are online and could help them learn, says Michael White, Apollo’s chief technology officer. “My faculty member’s not online, but 700 faculty members who teach the same thing are online, so it’s really the power of the network,” he says.
When students go to a page for a course, they see course content divided into fine-grained subsections. Dividing the course material into small pieces helps Phoenix determine exactly what students are learning, Mr. Menon says. “The recommended learning activities will be different for each student based on who they are, what they’re struggling with, and whether they need help,” he says. “It’s a very simple way to show where the adaptiveness comes into the classroom.”
Both of these directions—connecting students with a far-reaching network of potential tutors and showing students personalized material—erode the traditional idea of a class, a community based on an instructor and a cohort of students. Will that basic outline of a class eventually stop making sense?
“Probably,” Mr. White says. “I don’t think we’re there yet, ... but I think we’ll see in the very near future that that definition changes.”
A class, Mr. Menon says, is just “one instructional strategy.” Phoenix’s new platform aims to be “pedagogy neutral,” he says: Students can choose to learn through a class, through interactions with peers, or through some other means. “As we go through this, we can figure out—data will tell us—what types of concepts are taught well with what type of instructional strategy.”
