Mike Brown | TechChange | The Institute for Technology and Social Change

Traditional education in the US has long been required to accommodate those with disabilities through statutes like the Individuals with Disabilities Education Act (IDEA), but online learning has lagged behind. The excitement for improving online access to Ivy League classrooms should extend beyond just connectivity to intentional instructional design. But what standards and guidelines exist?

We recently took on the challenge of bringing all of our coursework into compliance with Section 508, a set of regulations which sets technical standards to promote equal access to (among other things) web content and multimedia for populations with disabilities.

Online learning can improve access to information by people with disabilities. Compliance with Section 508 of the Rehabilitation Act isn’t just the right thing to do, it also makes good business sense. But we still have a long way to go.

According to Wikipedia: “Section 508 (29 U.S.C. § 794d), agencies must give disabled employees and members of the public access to information that is comparable to the access available to others.” These accommodations include things like closed captioning and audio descriptions for multimedia, machine readable design to allow screen readers easy access to and navigation through content, and other methods of ensuring that everyone has the ability to benefit fully from our courses. Some may say that regulations such as these impose a great cost, and likely help few. But we think differently, for several reasons.

This is an endeavor we want to be doing anyway, because it’s the right thing to do.

A lot of the courses we do are about inclusion: using technology for democracy, better health, and conversation across traditionally disparate groups. We are proud to have students from around the world in each of our courses. Leaving behind those who have difficulty accessing technology would undermine our mission.

It is not difficult if done from the beginning.

Though including closed captioning or audio description tracks obviously involves more than the bare minimum amount of effort, if included from the beginning, it becomes part of the content generation process, and overhead is low.

It makes us more competitive.

Federal agencies and contractors are required to conform to the 508 standards if compliance is possible. This includes procurement. So a compliant product must be chosen over a non-compliant one.

It naturally follows from good design and coding principles as well as web standards.

Good code and good design have a common theme: they are clean. Clean design and code is also easier for assistive technology like screen readers to read. It’s simpler to do things like make text larger if there is plenty of space to do so, avoid using colors to denote meaning since it can’t clash with your color scheme, and leave room for captioning and transcripts in empty space that doesn’t distract other users with unnecessary detail.

Thinking about the challenges of accessing our content helps us make the experience better for all of our students.

Thinking about how to minimize the impact of compliance on your media forces you to think about how you present material, and to present it in different ways that will complement each other. Not only because different people learn different ways, but also because reinforcement through a different mode is still repetition, the most effective form of learning.

Thankfully, the 508 Standards are fairly straightforward. Indeed, they involve a careful analysis of the problem and what solutions work, which is a long and arduous process we are ill equipped to duplicate.

What isn’t straightforward ways to test your product and underlying platforms for actual usability. The next couple of posts on accessibility will talk about some of the more troublesome edge cases in 508, our process to make all of our content as accessible as possible, and how future standards and technologies can continue to make learning inclusive.

What processes do your organization use to expand access to your services?

Big data is already giving us better TV shows. Could it also help build a better education system?

This week our team went to 1776 Reboot: Education Meetup, where we heard from leading experts from Coursera talk about the future of online MOOCs, as well as entrepreneurs from TechStars about applying an accelerator approach to learning. But one of the real stars of the night was Richard Culatta of the Department of Education, who declared that we now have more data about what kids watch on Netflix than how they learn in school.

So what?

When Netflix rolled out House of Cards as all 13 episodes developed on metrics learned over the years from their watchers, Kevin Spacey stated in a Business Insider article:

“It’s a real opportunity for the film and television industry to learn the lesson the music industry didn’t learn. Give the audience what they want, when they want it, in the form they want it in, at a reasonable price, and they’ll buy it.”

In our last post, Four Reasons Why Universities Aren’t Ready to Move Online, we looked at how universities need to invest more heavily in producing compelling online content — not just videotaping professors lecturing. The dirty secret behind online all of the education platforms that are generating the creative chaos around online education is that they are not providing an online education at all, but rather educational content in a structured format. If that’s the case, what can online education learn from the current revolution in content distribution?

In criticizing current approaches to online learning, we often refer to the “Netflix” approach to online education — passive consumption of videos instead of interactive back-and-forth learning. But there’s no doubt that there is a market for passive consumption of educational videos, ranging from the current gold standard of Lynda.com to simply looking up a how-to screencast on YouTube.

Piecemeal Content (Amazon). Amazon is a retail company, that wants to also sell digital content. Think of this as purchasing and streaming an episode of Ken Burns Civil War. But are customers willing to buy educational content when there has been hesitation to do so for TV (hence the existence of PBS).

Free Prosumer (YouTube). YouTube is a Google product that wants to build general user data. The problem here for users is discovery and quality control — it’s hard to find quality, and it’s hard to find programs of study as opposed to small snippets of knowledge.
Freemium Model (Hulu). Hulu is an ad-supported subscription video service, that wants to build interest in existing broadcast content. It’s also perhaps the closest to the existing Coursera and EdX model. While the education is free, they are looking at “freemium” educational models where they can charge for certificates of completion or credit.
Premium Distribution Model (HBO). This could be TED talks right now, although those are free — TED controls the vertical by organizing the conferences, filming the speakers, and then distributing on their own platform. The content is often superb, but–like HBO–is restricted in it’s theme and format. HBO is a cable company that happens to be online.
Content Buffet + Original Content (Netflix). Netflix provides on-demand Internet video streaming. Most interestingly, they then used big data from how their users watch other media to figure out how best to deliver its own content. There’s not a perfect comparison yet, but there are some indicators of what’s to come.

Khan Academy started off with simple YouTube videos of basic skills. Since then, they have aggregated them into groups with clear skill progression and then allows students to practice with post video problems, which give them an enormous amount of feedback on how well students are learning. This data not only helps students learn through application of knowledge to problems with instant response, but easily enables Khan to present additional problems to support or remediate weak areas. And all that data can show Khan how to make an even better experience by combining it with theories on effective learning.

Big Data companies like Hortonworks are trumpeting data-driven education, while new startups like Clever and learnsprout are helping developers get in touch with new sources of data on achievement and teaching. Even government is getting in on the game. Led by the Department of Education’s Richard Culatta and the administration’s general open data philosophy, every metric available is being drafted to the use of improving education across the country. But the metric gathering potential of online learning is even greater.

Brace yourselves: We may be about to see some of the best educational content of all time built on metrics that traditional educators could only dream about. And moreover, since this isn’t just about producing content one time for one show, but for topics that will require constant updating and modification for improvements.

What will your courses look like when your professors are actually producing quality content and A/B testing the heck out of it? When they improve not semester to semester, but week to week?

This post was co-authored with Mike Brown.

Co-authored by Mike Brown.

The future of higher education may be online, but the present is still a mess.

The New Yorker recently published a thorough exploration of MOOCs and higher education. Coincidentally, this piece came out as the same week that my alma mater announced it had failed to fill about a third of its incoming freshman class. Whether a temporary enrollment misfire or permanent disruption of the education system, both the struggles of terrestrial universities and the potential for an online future raise important concerns about how higher education will survive.

Although perhaps not the author’s intention, the article revealed five key differences between traditional teaching institutions moving traditional courses online and courses designed to be online from the very beginning:

No experience in producing online content. The main video editor for Nagy’s course is a graduate assistant who recently defended her dissertation in Greek history, not a Web editor by vocation. Good educational content requires audio, video, graphics, and subject matter to work in unison. Universities are buying platforms like marble mansions and filling them with cardboard content. But live teaching is hard, which is why good lecturers are hard to come by. The same applies to other modes of delivery, and with MOOCs, the potential efficacy lost from skimping on the experience will scale with the course, growing linearly, while the cost of getting it right from the beginning is fixed, getting cheaper per person as number of students scale.

No clear teaching or evaluation model. This is still the “let a thousand flowers bloom” stage of online learning, but that has to end eventually. While it was good to see the back-and-forth on the socratic method, without methods of evaluating work, it seems premature to congratulate education on cracking this nut. Multiple-choice quizzes to test reading comprehension will never replace essays, and machines are a long way off from being able to grade 31,000 essays accurately. But besides peer grading, which is successfully used by Coursera and the Comprehensive Test Ban Treaty Organization’s “Around the Globe and Around the Clock: The Science and Technology of the CTBT,” we don’t have better ways of evaluating student progress in depth, as well as breadth. New models and tools are needed for these subjects.

No clear business model. It initially seemed unnecessary to take a trip and cameraman to Greece as part of the budget for this course, but if students are willing pay for that authentic experience, then why not? It may well be that including such edutainment content as shots from the real places, much like the history channel used to do, will benefit students greatly; what’s most important is that they track how it changes how students engage, and perform further experiments to validate these theories.

No access to social networks. Perhaps the most telling part of the article was the admonition of universities not just as delivering elite education, but connecting elites with one another into lifelong networks. Emphasis on admonition. Not only is more data available to mine when students interact in social network type settings, but students and teachers benefit from the collaborative and iterative experience inherent in group-based contemporaneous learning.

Traditional universities are, in the words of the article, standing in front of an avalanche. They are understandably attached to their current model, which they have developed over centuries, but it leaves them vulnerable to the scale-free model of online learning. The prospect of a global audience and substantial cost savings from online coursework is attractive. However, they are poorly positioned to benefit from either without revolutionizing their entire approach. Universities, in this new age, are facing the classic Shumpeterian forces of creative destruction. Much like the railroads, which once dominated transport, innovation is placing pressure on their model, and if they remain attached to a model displaced by innovation, they will be destroyed by it.

Are there more ways that universities are failing to keep up with the times? Are products of e-learning startups falling too far from the educational tree? Join the conversation in the comments.