An Unmet Need in the Market

School districts have seen a rise in the population of English learners other than Spanish speakers, creating a need for a new class of products capable of meeting shifting demands.

A variety of factors are driving those changes, including geopolitical forces and new patterns in global migration. Those forces are bringing students with new language needs into school systems that, in many cases, don’t have the staffing or resources to accommodate them.

While Spanish-speakers still make up the largest population of English-language learners in U.S. public schools, the most commonly spoken home languages include Arabic, Chinese, Vietnamese, Portuguese, Russian, Haitian, Hmong, and Urdu, in that order, according to data from the National Center for Education Statistics.

Those students’ needs are often overlooked as traditional ELL programs do not account for them in ways that are effective.

In response, districts are looking for better and more relevant training for teachers and setting their sights on potential innovations in technology to meet myriad language needs.

A Lack of Language-Specific Content

Survey data collected by the EdWeek Research Center this fall for EdWeek Market Brief shows that when it comes to the quality of resources to support English learners, district and school leaders and classroom educators have many frustrations — particularly with the lack of high-quality professional development.

But a substantial portion of the survey respondents, 36 percent, also indicated they are disappointed that materials do not address the needs of students in languages other than Spanish.

The nationally representative online survey was conducted from Sept. 26 through Oct. 8 of 1,135 educators, including 236 district leaders, 168 school leaders, and 731 teachers.

There’s a huge need and a market for curriculum development and assessments in languages other than Spanish — because those tools do not currently exist at scale, said Jason Greenberg Motamedi, senior researcher in Education Northwest‘s Center for Research, Evaluation, and Analysis.

Motamedi recalled recently working with a district in Oregon that wanted to create a dual-language program following an influx of Afghan students who speak Dari. But the materials necessary to meet the need were just not available.

It wasn’t realistic for the district to build its own content for that language group, Motamedi said. The district ultimately ended up giving up on the idea — and pivoting to find resources for another student population, those who speak Mandarin — as those materials were slightly more accessible.

School districts not only lack access to resources for languages that are less commonly used in the U.S.; they also struggle to find materials for those that are relatively well-established, such as Arabic, Russian, and Korean, said Jennifer Johnson, a senior consultant for multilingual learners at Education Northwest.

As a result, classroom teachers are sometimes forced create their own materials, often going home and translating the content word for word on their own, Motamedi said. The resulting resources often fail to capture many of the nuances of those languages, such as regional dialects or slightly different versions of the same language, said Motamedi.

While curriculum companies can’t provide content in every spoken language in existence, the first step is understanding the trends of geopolitical situations and the migration of students across country borders, to anticipate where areas of high need might be, Motamedi and Johnson said.

From there, it’s also key to create materials that are age-appropriate — something teachers are “really hungry for,” they said.

Often, available language materials are catered to younger students in a way that doesn’t meet the cognitive, critical thinking, and engagement needs of those who are older and picking up the language for the first time.

It’s also key for companies serving this space to be proactive in marketing to make sure that districts are aware that materials like this exist out there, Motamedi and Johnson said.

“Vendors and curriculum makers have a really important role in thinking about what language learning looks like,” Motamedi said. “Not just in general, but throughout things like math curriculum, or science and social studies.”

Data from EdWeek Market Brief’s survey also revealed a particular need for resources for non-Spanish speakers in school districts with higher poverty rates.

In addition, 49 percent of respondents who expressed disappointment with ELL resources for non-Spanish-speakers came from urban districts, compared to 42 percent of those from suburban districts, and 27 percent of those from rural districts.

Better PD and Specialization

In addition to the lack of high-quality content, educators across the country are largely unable to meet multilingual learners where they’re at because of lack of training and support.

Respondents to the EdWeek Research Center survey indicated that they’re dissatisfied with the level of professional development for teachers they’re receiving in working with multilingual learners.

According to the results, 40 percent of respondents say they’ve been disappointed by the overall inadequacy of professional development to support ELLs; 36 percent say they’re concerned about the lack of PD for teachers who aren’t ELL specialists, and 25 percent say that accompanying PD is low-quality.

When it comes to helping non-Spanish speaking English learners, strategies such as “technology can only take you so far,” said Sidharth Oberoi, vice president of international product strategy at ed-tech company Instructure. “The main thing is thinking through the ways that administrative faculty can allow for more time for their educators to pursue professional development opportunities.”

It’s also important to free up more of teachers’ time and attention to intervene with ELL students who are struggling academically, said Amelia Kelly, chief technology officer at SoapBox Labs, a company that specializes in children’s speech recognition and voice technology. Vendors that can help alleviate the burden of time-consuming tasks will be key to providing educators more space to work with English-learning populations.

Often, those teaching English as a second language are not given many tools to support their work, said Jamie Martin, founder and CEO of The English Quest, which provides curriculum on spoken English to students in India, through courses and an app.

Classroom educators “don’t have a dedicated program very often, or a program that’s empathetic to the fact that you’re teaching children [who speak] a language that you don’t speak, so it’s an extremely challenging context,” he said.

This need to provide teachers with stronger skills to help students’ language development will grow in coming years, as global disruption sends students across borders, Martin said. Teachers who aren’t specialists in working with ELLs will need easy access to materials that offer specific strategies to help those students.

Overcoming Fears Through Technology

One way to reduce teachers’ workload while expanding language-learning opportunities is by leveraging emerging artificial intelligence technology. Some education companies have already begun using AI as a low-cost, one-to-one method of serving multilingual learners.

Khan Academy’s AI-powered tutor, Khanmigo, is a chatbot that serves students with its language capabilities. Students can interact with the platform in any language and get responses back — whether in that language or in English.

As of now though, Khanmigo only supports Spanish and Portuguese officially, because the technology has been vetted and offers a moderation system to detect if students are indicating violence or inappropriate content in those languages.

Other languages can still be used — but the company is still developing out the safety guardrails before claiming that they officially support those additional languages.

“Students would tell us that sometimes they don’t quite understand what the teacher is saying, but they’re too embarrassed to ask,” said Kristen DiCerbo, chief learning officer at education company Khan Academy. “But they can ask Khanmigo in their home language, and it gives them an explanation they can understand.”

SoapBox Labs is another education company using voice-enabled AI to serve English-language learners. The organization’s speech recognition and analysis system is specifically designed for children’s voices and recognizes a range of dialects, accents, and languages, with data from about 193 countries informing its model.

Their technology works with literacy and math products and allow students to interact with those exercises using their voice, rather than navigating a keyboard and trackpad.

“If you’re learning a language, the best way to do it is to have a kid speaking out loud as much as possible without being afraid of making a mistake,” said Amelia Kelly, the company’s chief technology officer.

SoapBox also uses AI for measuring oral reading fluency. For students reading a passage aloud, the technology can automatically transcribe what the student said and compare it to the passage they were supposed to read. This system saves teachers time in having to go back, listen to the recording, and manually mark the words that were missed or mispronounced.

Despite great strides that can be made with AI, the necessary guardrails must be in place to protect students and their data, Kelly said.

Privacy laws have only gotten more stringent over the years, which is a positive outcome, she said. But it’s important that vendors give customers the option of whether they want to share their information, and then only work with what they’re given.

As students’ voices come in through their system, SoapBox gives customers the option of deleting that data immediately or allowing the company to keep it for improving their models. If the customer does allow for the data to be shared, everything is anonymized and can be requested to be deleted at any time.

Building for the Future

The challenge for many education companies in trying to support non-Spanish-speaking English learners is determining the level of investment to make in what they may see as a niche market.

The first step is to reframe the problem as a solvable one, said Khan Academy’s DiCerbo.

AI provides a method of language support so that companies don’t have to develop for each language separately, she said. But it’s also important to build platforms with localization in mind, so that systems are more flexible, and features don’t have to be hard coded from scratch for each new translation.

That means seeing the problem not just as an issue of language translation but recognizing from the beginning that localization involves things like cultural context to reflect students’ communities, as well as including diverse images to reflect the students themselves, DiCerbo said.

Education providers must also be in tune with what happens more broadly on the world stage.

“As a developer, you can’t just think about what the current trends are,” DiCerbo said. “If you’re just designing for the current immigration, you’re going to miss what the next wave might be.”

Takeaway: Districts are deeply concerned about the quality of PD associated with products designed to serve ELLs — including for teachers who are not language-specialists.

They’re also worried about the lack of scaffolding of products for students of different language ability, and about the shortage of resources to help students who speak languages other than Spanish.

Companies that commit to providing supports for classroom educators in those area — using tech or other means — are likely to stand out, at a time when school systems are trying to meet an increasingly diverse array of student language needs — not just those of Spanish-speakers.