There is a fundamental disconnect in the edtech world: What’s best for profits isn’t always what’s best for learning. It’s not that edtech makers don’t care if their products help students in the real world. The trouble is that efficacy studies require significant funding, research expertise, and time, all of which are in short supply as edtech companies compete in the marketplace. The resulting scarcity of evidence for learning gains leads many districts to choose their edtech based on other factors, such as the price or product features, which exacerbates the problem.
“There has always been this mismatch,” noted educational economist Matthew Steinberg, the managing director of research and evaluation at Accelerate. “In some sense, there is a market disincentive to investing the time and resources to develop rigorous evidence on whether a product works.”
Countering this trend, several edtech partners in the Cooney Center’s Sandbox for Literacy Innovations have won research grants—from Accelerate and other sources—to fund third-party efficacy studies in partnership with school districts and independent researchers. Their stories show a commitment to continuous improvement and offer a roadmap for bridging the edtech disconnect.
Putting Literacy Tools to the Test
Paloma Learning was already deeply informed by research when they won a grant from Accelerate’s Call for Effective Technology (CET) program in 2025. Their app helps parents and guardians reinforce in-school learning at home with their kids through short, daily tutoring sessions aligned to classroom curricula.
“The core insight behind Paloma and all the work we do is that parents are the biggest influence in their kids’ lives and learning outcomes,” said Paloma CEO and co-founder Alejandro Gibes-de-Gac. “There’s no shortage of research that substantiates that, and anybody who’s a parent will tell you that.”
Meanwhile, separate research has shown the power of one-on-one tutoring, especially in the face of post-pandemic learning loss. As a result, a wide array of initiatives are now trying to leverage technology to make one-on-one tutoring available to every student who needs it. Paloma’s solution is using tech to empower parents.
“Parents are the tutors hiding in plain sight,” said Gibes-de-Gac, who launched the app in 2023 to provide families with 15-minute, curriculum-aligned lessons on literacy and math, prompted by a daily text message. Every literacy lesson ends with the creation of a short, decodable story that merges a topic of personal interest chosen by the student with the academic skill of the day.
Like all Sandbox partners, Paloma’s app will be further refined through research, including consultations with experts in the science of reading, guided co-design sessions with kids and their parents, and accessibility insights informed by universal design for learning principles.
Nevertheless, no matter how much care and expertise go into developing an edtech product, the complexities of real-world households and classrooms can still make a big difference in its effectiveness for learning. For example, Gibes-de-Gac cites the so-called “five percent problem,” coined by Laurence Holt, senior advisor at the XQ Institute, a nonprofit dedicated to transforming high schools. As Holt explained, edtech companies often highlight the learning gains made by the small fraction of students in studies who use their products in the manner and dosage prescribed by the developers, while ignoring the impacts (or lack thereof) for the vast majority of students who don’t.
That’s why Gibes-de-Gac considers the study being funded by Accelerate to be so vital. In addition to awarding the grant, Accelerate paired Paloma with the independent research organization, WestEd, to help design the study and crunch the numbers. They will be testing their app in two large, urban districts that serve high percentages of students from lower-income and multilingual households. In total, the research includes more than 2,500 students across 10 schools, drawn from similar demographic backgrounds, split between schools that will use the Paloma app and those that won’t.
The two biggest metrics will be tracking how much and how often families use Paloma’s app and student progress on the districts’ standard literacy assessment from the middle to the end of the current academic year.
“We view it as a learning opportunity. We want to understand what impact the product is making in order to improve it,” said Gibes-de-Gac. “That’s our whole reason for being. We’d have existential angst if we didn’t get the evidence that it was making a difference in student learning.”
Moving from Living Rooms to Classrooms
There are four main research objectives in the randomized controlled trial of the Infinibook, the flagship product of Cali’s Books, another partner in the Cooney Center Sandbox. One of the four objectives will compare literacy improvements among first graders who do and do not use the Infinibook, a digital device that houses paper storybooks with pushbutton interactive elements such as audio narration, story-related questions, and vocabulary support.
The other three research questions all focus on how teachers incorporate the Infinibook in student reading time and how they perceive the benefits and obstacles of using the product in a classroom setting. There’s a good reason for that tilt toward teachers and classroom implementation, according to Anne Blackstock-Bernstein, head of research for Cali’s Books.
“Teachers are the market,” said Blackstock-Bernstein. “If teachers don’t see the value in a piece of edtech, then it’s not going to be a successful tool for education.”
Not every edtech tool is destined for classroom use. Indeed, the original market for the Infinibook was parents. The founder of Cali’s Books, Carinne Meyrignac, came up with the idea when she and her husband felt disconcerted by the amount of screen-based learning in their twins’ kindergarten classroom.
“I was developing a direct-to-consumer product, but my vision was to have it used in schools. That’s where I wanted to have an impact,” said Meyrignac. “I had no idea if it was possible or not, but I knew that to get there, we needed research in the classroom.”
To ensure the most rigorous, unbiased data from teachers, they needed a fully-randomized controlled trial (RCT). They secured a Small Business Innovation Research (SBIR) grant from the National Science Foundation (NSF) to fund the RCT, which was designed and implemented by Blackstock-Bernstein and two independent academic researchers—Bruce Homer, professor of educational psychology at the Graduate Center of the City University of New York, and Jan Plass, professor and chair of digital media and learning sciences at New York University.
Specifically, the research includes 20 first-grade classrooms, randomly assigned to use Infinibooks or not. Teachers will provide students with Infinibooks during small-group and independent learning time for at least 30 minutes a week over three months this spring. In addition to teacher surveys and a brief weekly log of Infinibook use, students will do a battery of literacy tests before and after the intervention.
“Every company that wants to claim there are educational benefits of their product should run a study like this, just like if you want to put a new drug on the market, except we don’t do it,” said Plass. “So, we have all these unvalidated products in the market, and we often find that they don’t do much.”
Building an Evidence-Driven Marketplace
Doing research in schools can be challenging for edtech companies, even those with funding, research backing, and the best intentions. For starters, all studies that involve people need Institutional Review Board (IRB) approval on issues such as ethics, safety, and consent, which is often a months-long process. In addition, districts are often reluctant to saddle busy teachers with the additional responsibilities that a study might require.
For instance, the efficacy study proposed by LitLab, another Sandbox partner receiving a 2025 CET grant from Accelerate, included a teacher survey about the company’s AI-generated storybooks for young readers and the usability of the dashboard tracking student progress. The other data they collected was student progress on the district’s existing literacy assessment.
In partnership with the education R&D firm, SETA-ED, LitLab will compare learning impacts between students who use the product for the recommended 15 to 30 minutes twice a week and students who use it less frequently. They also want to know if literacy outcomes differ based on how teachers use the dashboard or student demographics.
“It’s a 15-minute survey. It’s not meant to be a large investment of time,” said Drew McCann, LitLab’s head of learning. “But it’s been really hard to get schools to agree to any kind of research, even if it’s very low touch.”
Eventually, LitLab enlisted several elementary schools in a district where some teachers were already using their product. They are tracking students in 21 classrooms, some of which are led by teachers who opted in to use LitLab. For gold-standard research, the treatment and control group classrooms would have been randomly assigned, McCann acknowledged. But that would have taken even more time to arrange. As it was, they were barely able to complete the 12-weeks of intervention before this spring’s final literacy assessment.
While companies like Paloma, Cali’s Books, and LitLab are models for pursuing rigorous research on learning impacts despite the costs, closing the evidence gap in edtech will require a broader shift in the market. To that end, Accelerate is working with state and district leaders to develop tools and guidelines for selecting products and interventions based on cost-effectiveness—simply put, the amount of learning gained per dollar spent.
“If we’re going to leverage technology to address what I consider to be a crisis in student learning, then we need this research to inform schools, districts, and families,” Steinberg said, “so we can have a more rational and better functioning edtech market.”
