Make Online Education Great (For The First Time)

Volume VII, #5

It’s human nature to think things were better in the Good Old Days. Even near-death experiences of yesteryear can be gilded with nostalgia. Like the time at dinner when my mom asked me to pass the mayonnaise. Inspired by the funky 70s-era combination she wore as she headed out for a night of disco dancing, I took it upon myself to slide the giant jar of Miracle Whip across the length of the table like a curling rock. I missed the mark and the mayonnaise toppled onto her bizarre sweater and ultra-suede skirt. To my 8-year-old eyes, her outfit wasn’t any worse. But with a look I’d never seen before, she let out a scream, chased me up the stairs in her heels, tripped, fell and cut her knee, which probably saved my life.

Today we joke about the mayonnaise and for my birthday last month, mom sent me a limerick:

In a story that’s somewhat bizarre
There’s the tale of the infamous jar
Ryan hurled down the mayo
And ruined my day-o
If forgotten I’ll show you the scar

The new Administration shares this tendency to idealize the Good Old Days. Our President wants to Make America Great Again, and I admire our new Secretary of Education's sincere desire to Make American Education Great Again, although it's not clear which era of American education she's referring to. Still, all the nostalgia in the world won't lead to calls to Make Online Education G‎reat Again.

Online education has never been great. We were reminded of this last week when the National Bureau of Economic Research released a working paper by Caroline Hoxby, a Stanford economist, on poor returns from enrollment in online programs. Despite many methodological flaws, no critics disagreed with the fundamental conclusion that many online programs are producing a poor return on investment for students.

And if online education was ever great, when was that exactly? Back when we called it distance learning and it resembled a correspondence course on a screen? Or when we witnessed the birth of WebCT, the first widely adopted Learning Management System? Perhaps when Michael Milken’s UNext spent $180M to build million-dollar simulation-based business courses (designed by Roger Schank, future Chief Learning Officer of Trump University) with universities like Columbia, Stanford and Chicago, only to disappear several years later? Or when University of Phoenix Online and other purpose-built online universities enrolled hundreds of thousands of students in online degree programs consisting of text-based courses that translated the onground environment so literally it was almost robotic: read material, participate in discussion, submit weekly assignment. Or was it more recently, when the New York Times declared 2012 to be “The Year of the MOOC,” because somehow making these same online courses available for free and without any associated credentials would somehow improve the situation?


Those who make a living from online education might point back five years ago, before we began to see a slowdown in enrollment growth. From 2000 until about 2012, rapid growth in online enrollment was a result of enhanced accessibility to degree completion. The (mostly for-profit) providers that provided the greatest accessibility (open enrollment, weekly starts) boomed. But there were already signs that the pool of students valuing accessibility was drying up. According to a Parthenon report in 2012, nearly 50% of the target demographic for online degree completion programs had already been in contact with a (for-profit) postsecondary institution about completing a degree online.

Since then, online market share has shifted to brand-name universities like USC, UNC and Georgetown (all partnered with 2U), Arizona State, Penn State World Campus, University of Maryland University College, and the state universities partnered with Academic Partnerships, as well as to more legitimate-sounding brands like Southern New Hampshire and Western Governors. Increasingly, prospective students recognize that all university brands and degrees aren’t created equal and aim higher than generic online degrees from open enrollment institutions. Students have been rewarded by higher quality programs. Arguably, online education has never been better.

Nonetheless, overall online enrollment has flattened, particularly at the undergraduate level. Meanwhile, according to Eduventures, OPMs have increased the number of online programs by 110% in the past five years. This means fewer students per program, which is deadly for a delivery modality that only produces a surplus at scale.

For online education to take the next step, it will need to demonstrate to students that online credentials – even those from recognizable brands – will help them get a better job. The two keys to this are persistence and integrity, and last week University Ventures announced two new investments in companies with encouraging solutions.

Motimatic combines behavioral science with state-of-the-art digital advertising techniques to improve student engagement and completion. Students receive tailored messaging encouraging them to persist in place of the online advertising they would normally encounter on Facebook, Instagram, Pinterest and elsewhere. The first randomized controlled trial of Motimatic’s impact, completed last fall with a group of over 3,000 students, found that the number of students remaining in the Motimatic group after one year was 9% higher than in the control group.

Examity is the best available and fastest growing solution for integrity i.e., ensuring that the student of record is the individual who has achieved learning outcomes. Examity provides 24x7 online proctoring of summative assessments in three ways: live, recorded and automated. Universities and faculty select the appropriate level of security to ensure the identity of the student and the integrity of the test. Solutions like Examity will be key to convincing employers that online credentials are not only equal, but actually better indicators of competency and job fit than traditional programs without comparable authentication.

We’re excited to see solutions like Motimatic and Examity improve online education’s value proposition for students.


I met K. Ryan Burke a few years ago in Washington, DC, at an event organized by the Department of Education. At the time, she was a staffer at the National Economic Council helping to lead policy development on workforce training and was previously at McKinsey, where I began my career. So last month I was fascinated to learn she’d gone back to school. In her excellent blog, From The White House To Coding Bootcamp, Ryan spells out her reasons for returning to formal postsecondary education. Then in a new piece last week, she does a good a job as I’ve read explaining the Whats Whys and Hows of effective learning. Here’s what she says in If You Want To Go Quickly, Go Alone. If You Want To Go Far, Go Together.

There are a couple of labs (that’s how we refer to our assignments) that are notoriously hard. Yesterday morning, I started one of those labs and with some hesitation paired with my classmate Holt… I started as the driver with Holt guiding me through the steps he wanted to take to fix the errors... The whole time I’ve been in bootcamp, I had been under the impression that I was slowly mastering an approach called Test Driven Development (TDD). As it turns out, I have been using about 10 percent of the power of TDD. I know from seeing Holt operate. He would spend literally 10 seconds looking at lines of code that I had previously ignored and knew exactly what we needed to do next… Had I not worked with Holt, there’s a decent possibility I would have gone through all 12 weeks of this program without truly realizing the power of TDD. Not only that, but it was a lot more motivating to work together. We spent six hours coding with moments of joy, frustration, and tears (not really, but I came close). When the 53rd test lit up green, we jumped out of our seats cheering with euphoria (the way that Alabama football fans do when they win a big game).

The other thing they tell you when you start bootcamp is that you will learn just as much or more by teaching others... On Monday, my classmate (a former professional poker player turned investor who next year is planning to only eat things he spearfishes) asked me about a lab that I recently completed. As I looked over his code, I realized that I had really not internalized what was actually happening in the problem (that tends to happen when you are learning so fast that you only have minutes to spend on huge concepts). So I went back through it trying to explain piece by piece and going back to the lesson together as we needed. No kidding, those 10 minutes with him were way more beneficial than when I did the lab the first time.

So over the next eight weeks I am going to give this whole pairing, questioning, and teaching thing a try. I’m pretty sure I’ll get farther that way in the end and very confident I will be happier and saner in the process.

Beyond critical solutions like Motimatic and Examity, making online education great for the first time will require new technologies to replicate the intensity of Ryan’s bootcamp experience. Online students need to learn from and teach each other in an environment so immersive it will become commonplace for students to spend six hours (coding and) learning together. Integrating synchronous class discussions (as 2U does) is a start. But there’s so much more room for new thinking and innovation in online education.

As I was writing this, my 11-year-old son Leo came into my office along with 6-year-old Zev. They started bickering, as they do. Then, in response to something dumb out of Leo’s mouth, Zev responded with the following: “I wish your idea had a face so I could smack it.”

In my family, as in online education, the Good Old Days are now.


Three articles that tell us where the puck is going in higher education

1. Elementary, My Dear Watson IBM announcement of Esaki, a “cognitive advisor” that utilizes Watson artificial intelligence to recommend jobs based on candidate skills, interests and personality. The application uses two Watson API’s to power the application. The AlchemyLanguage API gathers concepts, skills and keywords from both resumes and job descriptions to provide the best job matches for a candidate. The Conversation API enables candidates to have a conversation with Watson as if they were talking to a recruiter. Just like a recruiter would ask questions about the candidate’s interests, skills, and experience, Watson does the same. In turn, the candidate can ask about company culture or specific roles, empowering the user find jobs that they are the best fit for. This means Eric can simply submit his resume, answer a few questions, and sit back, relax, and let Watson recommend roles tailored for his skill set and interests. Read more 2. “I’ve Worked in the Private Sector. They Expect Results!” SHRM write up of new Burning Glass report on internship postings showing that employers expect specific skills from candidates for internships. A marketing internship will typically demand skills in social media and marketing research, while a sales internship will ask for business development or sales management. Read more 3. Singapore Leads the Way on Employer Engagement Singapore Straits Times article on new programs at Singapore Institute of Technology and SIM University that allow students to alternate days or semesters of learning and work in programs in information security, software engineering, engineering, finance, business analytics and hospitality, by Sandra Davie. If, as expected, 40 per cent of each cohort takes the university route by 2020, employers would not hire people simply because they had degrees. In a more competitive climate, they also want to be assured that their fresh hire is a young talent who can integrate well into their organisation, enjoys the nature of work and is passionate or deeply interested in the industry they are in.
Read more