EdTech and the accessibility paradox


Dec 2, 2015

Summer Cox is an exceptional student education coordinator at Henry County Public Schools in Georgia. For a number of years, her district has pursued personalized learning. That’s given Cox a front seat to a new movement in education, which calls for recreating classrooms in a manner that supports learning for each child.

Cox has a unique outlook on where personalized learning is headed: she helps oversee the district’s special education programs and strives to ensure that students with disabilities are included in the overall vision for structural and instructional reform. In some ways, personalized learning is catching up to what special education advocates have long believed. As Cox explained:

It’s my opinion as a special educator that not all students are the same. Therefore, we should not present them with the same learning experiences. … Different students need the ability to access their learning differently, and we should help facilitate that as teachers.

But in her role, Cox has also seen where school districts like hers struggle to ensure that software programs actually align to student and teacher needs, especially when it comes to providing content appropriate for different learners. “Our teachers need the ability to modify the content if needed to fit individual learning needs or the needs of small groups of students, if they don’t fit into the ‘packaged’ curriculum that is provided with the software,” Cox said.

As I’ve written about before, it’s vital that as more and more schools adopt blended and personalized learning, they do so with an eye toward including the needs of students with disabilities into their designs. Having staff like Cox dedicated to this cause helps. But districts also need to make sure that the “packaged curriculum” intended to personalize instruction is accessible and effective for all students and allows teachers the degree of flexibility that Cox described.

These efforts, however, may be thwarted by how EdTech companies go about building software tools and by the current compliance-driven attitude that many curriculum providers take toward meeting the requirements of Section 508 of the Rehabilitation Act of 1973, which ensures software and website accessibility to people with disabilities. Anthony Kim, founder and CEO of Education Elements, a company based in the San Francisco Bay Area that helps districts implement personalized learning, sees this ambitious rule playing out with mixed success in the EdTech industry: “Making products 508 compliant is very complex and costly,” he said. “Frankly, most companies develop a product and 508 compliance is an afterthought. They take what they build and then have to retool and scale back visuals to become some percent compliant. I rarely see any product that’s 100 percent compliant.”

One popular research-based approach to mitigate this failure is to ensure that technologies support Universal Design for Learning (UDL), a framework to help educators and publishers consider how best to design learning experiences and materials for those with and without disabilities. The underlying principle of this approach is that schools should only be acquiring materials and pursuing teaching methods that have been carefully designed to work for every student. Increasingly, policymakers are suggesting or outright requiring that EdTech products support UDL requirements.

Even with UDL gaining steam on the demand side, the question remains, however, of whether the EdTech market can supply tools that meet the needs of all learners or optimize for different students in different circumstances on an equal basis.

A first cut at this would mean honestly assessing which tools are actually best suited to which circumstances and students, rather than treating compliance with rules like 508 as binary. For example, Anthony Kim suggests that a common rubric might help that could show degrees and coverage areas of 508 compliance that companies could indicate they meet.

But even this degree of transparency may not ensure that more or better tools would emerge to meet the needs of students with disabilities, especially given that this population represents only a fraction of the overall customer base that schools serve. Indeed, even with some EdTech tools starting to emerge to serve specifically students with disabilities, the market for these tools remains small.

Another approach beyond UDL or increasing transparency is to overtly reward the development of curriculum tools that meet the needs of students with disabilities. Jeff Katzman, founder and CEO of the Core Learning Exchange (Core-LX), wants to do just that. He asked himself, “What makes more sense … to design learning options optimized for specific disabilities and abilities or make generic content technically accessible for use keyboard access, alt text, and text to speech programs like JAWS?” Katzman believes that the central problem in today’s EdTech market is that there are limited economic incentives for commercial publishers to create specialized content. “They generally opt to create a product that serves the broadest audience, thereby creating a product that serves more students, but none very well,” he said. “The question is, how do we incentivize curriculum developers to develop content optimized to serve specific needs?”

His company, which functions as a crowdsourcing platform for educational content and lessons, offers one solution to this market failure: giving additional royalties for historically underserved pockets of the content market, such as content for students with disabilities or English language learners. As EdSurge recently reported, there are a growing number of marketplaces where educators can access teacher- and district-created content. Core L-X facilitates this same sort of exchange, but weighs the value of different types of content differently based on high-need content areas like those Cox cited.

Historically, the curriculum market and education system as a whole have struggled to serve students with disabilities. Making materials accessible under the current law may not be enough to personalize learning. As the EdTech content market continues to grow, we need to be clear about whether we are successfully designing content for all students, and whether we might be able to design market conditions that optimize for different students in equal measure, regardless of what proportion of the overall population they represent. These considerations will be vital to ensuring that the growing enthusiasm to personalize learning translates to better outcomes for all students.

Julia Freeland Fisher

Julia is the director of education research at the Clayton Christensen Institute. She leads a team that educates policymakers and community leaders on the power of disruptive innovation in the K-12 and higher education spheres.

  • Thank you for adding to the dialogue around inclusive curriculum design practices. The argument that addressing accessibility needs constrains innovation and options has been around for many years and is nonsense. Learners are infinitely variable and those with sensory, physical, cognitive and learning disabilities routinely represent that variability in all education settings. Designing curricular content and delivery systems that can accommodate these students results in greater, not less, flexibility and responsiveness, which is what personalization is all about. Retrofitting rigid, inflexible products is expensive, and generally results in inefficient, inelegant and impractical solutions — witness any ugly wheelchair ramps bolted to the sides of municipal buildings.If the range of learner needs i considered from the outset, including those with disabilities, solutions emerge that end up being better for everyone.

    EPUB3 an HTML5 increasingly offer expanded technical opportunities for implementing multiple ways of representing information, for giving learners a variety of ways to demonstrate mastery and for keeping them engaged. While I agree that not very product can be expected to meet the needs of every student in every circumstance, they can be designed to support the personalized needs of learners by supporting interoperability. Creating a braille version of a social studies textbook is way out of scope for most digital curriculum materials developers, for example, but having a well structured source format does enable rendering devices like the iPad to speak text aloud or transform the content into a braille ready file and immediately out put it to a physical braille device.

    As for the extent to which a products meets an accessibility specification, the Section 508-aligned Voluntary Product Accessibility Template (VPAT) approach is design for that (see http://centerononlinelearning.org/resources/vpat/) as do the WCAG A, AA, AAA designations. Making this information readily available would go a long way towards creating an informed consumer base.

    Adding additional royalties for accessible product design may be commercially attractive, but it reinforces a concept of inequality and formally associates equitable access with inequitable costs, and perpetuates the notion that students with disabilities ultimately need to pay for the opportunities that non-disabled students routinely receive at little or no cost. Crowd sourcing enhancements like text equivalents for images (see http://diagramcenter.org/development/poet.html) is a partial solution (and also helps avoid the spectre of compulsory licensing that often arises in conjunction with the term “market failure”).

    While, ultimately, the market will decide, K-12 consumers are likely to follow their postsecondary counterparts and increase requirements for accessible curriculum materials and delivery systems in order to avoid any liability under civil rights (and, in their domain, special education statutes). As those expectations increase, designing products for learner variability at the outset seems like the best market strategy.

    • Julia Freeland Fisher

      Skip, thanks for the comment and the wealth helpful examples. CAST’s work is a vital piece of this conversation. I think the key point I was hoping to make here is less reiterating a broken claim that accessibility constrains innovation — I agree that is the wrong message entirely. Rather, as we are trying to think through a system that truly personalizes, I think we need to be wary of every resource optimizing for every student, not in terms of students with disabilities but in terms of all students. The risk is that in the interest of efficiency rather than personalization, we digitize against the false promise of teaching to the middle of the class. And my related concern is that students with disabilities may be left out of this new “movement” if personalized learning EdTech tools continue to optimize for certain students and not others. We’re already seeing this happen in some blended learning tools, as Kim noted.

      Your points about interoperability are right on and I think that the K-12 system has a long way to go on this front – it’s amazing how much although we may have the technological capabilities to pull this off, the market has not moved quickly enough in this direction. The question is how can we design a market that does that, taking into account where accessibility regulations continue to fall short. More transparency and better resources, as you mention, may be a big push that more advocates for personalized learning should get behind.

  • Let’s change the term from “market’ place to simply a place where resources can be found. Two categories of resources would be OER resources and Non-OER resources. OER resources by definition don’t exist in a market.