On a recent weekday morning, a six-year-old girl with brown pigtails stared at an iPad perched on the desk in front of her. As she studied the screen, she squinted her eyes, and her brow furrowed into a pair of delicate question marks. A minute ticked by. She was still perplexed. Then suddenly, the iPad emitted a soft, triumphant-sounding ping, and her face lit up.
The girl had successfully solved a mathematical puzzle in the educational software program ST Math. At adjacent desks, her first-grade classmates at Jack L. Weaver Elementary School, in Los Alamitos, Calif., were grappling with their own ST Math challenges. The room was silent, with no hint that the morning recess was just 15 minutes away. “They could do this all day,” the teacher, Kathi Ruziecki, whispered.
ST Math, which uses visual puzzles to teach concepts such as fractions and number lines to elementary school kids, is the creation of the Irvine, Calif.–based nonprofit MIND Research Institute. The organization’s work could help quell some heated debates being waged in school districts and around kitchen tables across the country regarding technology's role in the classroom.
Ever since Weaver Elementary’s principal, Erin Kominsky, signed on to try out the earliest version of the ST Math software 16 years ago, when its interface was only via traditional desktop computers, it has been a cornerstone of the school’s curriculum. Kominsky combines two 35-minute ST Math sessions per week, in addition to homework assignments as appropriate and the option to use the software in their free time, with traditional math instruction and cognitively guided instruction, a teaching style based on listening to children’s mathematical thought processes. The approach is a good example of blended learning, which combines traditional teaching methods with computer-generated ones.
Kominsky said her school has been transformed since incorporating ST Math: “It has changed the playing field with math for us.”
When Kominsky arrived at Weaver Elementary in 1996, it had been closed for 13 years. Even though Los Alamitos residents had recently voted to reopen the school, at first they weren’t eager to enroll their kids. As Kominsky put it, “We had no residents who would be caught dead coming here.” Parents from surrounding cities started driving their kids to Weaver.
Within three years of adopting ST Math, “our school outperformed every school in this district” in math, Kominsky told me while we observed the first-grade class. Local families started getting a lot more interested in Weaver. Now the student body has swelled to 720, with the majority local to Los Alamitos, and there’s a waiting list to get in. This year, 98 percent of Weaver students tested proficient or advanced in mathematics, and the school’s Academic Performance Index, California’s measure of how well a school is doing overall, was the highest in Orange County. Weaver also scored higher than any elementary school in neighboring Los Angeles County. Kominsky credits much of the school’s success—not just the spike in math scores—to MIND Research Institute.
In classrooms across the United States, digital learning has become de rigueur. As of last year, there were 10 million iPads in American schools, according to an estimate by Apple. The U.S. market for educational software is valued at $7.9 billion. But while technologists have been clamoring to enter the education space, some educators have resisted the onslaught, wondering whether more time in front of a screen is a good idea for kids who already watch an average of 28 hours of television a week; whether tech is a good use of limited educational resources; and if it will be used as an excuse to cut costs in other areas, such as by replacing teachers and increasing classroom sizes. Other skeptics fret about privacy of student data, or maintain there’s not enough evidence that technology is effective at helping kids learn.
The Los Angeles Unified School District found itself at the center of the debate after it approved plans last summer to distribute iPads to 600,000 students. When 300 students hacked the security settings of their new devices and began using them for unsanctioned activities, complaints over the expense—$1 billion—segued into questions about whether the district had properly prepared for the rollout. By December, critics were howling about the multimillion-dollar spend on unproved software from education company Pearson, and a survey of LAUSD teachers showed that only 36 percent favored continuing the program.
The survey echoed educators’ attitudes toward technology nationally: A 2012 report by the education nonprofit Project Tomorrow showed only 17 percent of teachers believed tech helped students explore their ideas deeply and 26 percent thought it boosted problem solving. The same study, however, also showed that teachers in training are far more enthusiastic about using technology as a learning tool.
Those in the pro-tech camp like how software can assess students’ progress while they work, a facet known as embedded assessment. It provides teachers with real-time reports signaling which students need more help and allows them to reach all students at their individual ability levels. Given that educators have traditionally relied on tests administered every few weeks or months, most see this as a revelation. “It’s huge,” Damian Bebell, assistant research professor specializing in testing and educational policy at Boston College’s Lynch School of Education, said of embedded assessment. “It’s like the difference between sending an instant message and sending a letter by Pony Express.”
Research into educational technology and software is still in its infancy; companies, meanwhile, are operating at the speed of business. Developers seem to overlook that the stakes in educational software are different from, say, those of business or personal technology.
But the quality of the software that's out there varies widely. There are programs that simply add a point system to ordinary tasks, and video games where the action is periodically interrupted by lessons—a spoonful-of-sugar approach. Researchers say these types fail to keep children engaged because they’re not instilled with intrinsic incentives. For a game to work, it must be motivating in itself.
Given how crowded the educational software market has become, knowing where to find quality software and then figuring out which is most effective is a complicated task. “It’s a digital Wild West out there,” according to Michael H. Levine, executive director of the Joan Ganz Cooney Center at Sesame Workshop, a children’s educational media and research organization. “Curation and discovery are the biggest challenges for educators and parents.”
Research into educational technology and software is still in its infancy; companies, meanwhile, are operating at the speed of business. It’s a lot easier to write some code and package a product than it is to conduct a rigorous and reliable evaluation of such a product. Developers rushing to put out educational apps, games, and other software often seem to overlook that the stakes in education are different from, say, those of business or personal technology. “It’s rare that people stop and ask, ‘To what end?’ and ‘How would I find out if it’s working or not?’ ” said Bebell.
One organization that is asking those questions is MIND Research Institute.
As a Southern California kindergartner in the 1970s, Matthew Peterson was diagnosed with dyslexia. “My challenges were reading, listening, writing,” he says. “The teachers told my parents that I couldn’t learn.” Peterson is now 41, with spiky black hair, and the bookshelves of his Irvine office brim with titles like Understanding Numbers in Elementary School Mathematics and Spark: The Revolutionary New Science of Exercise and the Brain. One wall has a giant whiteboard with mathematical equations scrawled across it.
Despite Peterson’s academic frustrations as a kid, he was enthralled with visual puzzles that required logic and spatial reasoning, but not language skills. Around the time he learned to read—fifth grade—he decided he eventually wanted to study the brain and problem solving. “I told my mom something like ‘I want to go into neuro-engineering,’ ” he says.
That major doesn’t exist, so when Peterson enrolled as an undergraduate at UC Irvine, he cobbled together the rough equivalent: a degree in electrical engineering and biology. While in college, where he also majored in Chinese literature and language, he developed an interest in interactive media, about which he went on to write a book, and he indulged his affinity for brain teasers and mathematical puzzles by creating computer puzzle games. His dream after college was to build software to teach children with dyslexia without using language. He soon realized he couldn’t do that effectively without first understanding how the brain works, so he went back to school to study neuroscience.
As a Ph.D. candidate at UC Berkeley, Peterson began researching the brain’s capacity to think ahead in space and time, known as spatial temporal reasoning. (The "ST" in ST Math stands for "spatial-temporal.") He found that tapping into this ability could help improve critical thinking, reasoning, and problem-solving skills. He also studied how the brain’s neurons process information and how it develops mathematical reasoning abilities.
In 1998, while still a doctoral candidate, Peterson founded MIND Research Institute with the goal of making games designed to boost the intelligence of children who struggled with language, as he had, by strengthening their spatial-temporal reasoning skills.
Improving these kids’ intelligence “was like this Holy Grail,” he says. “If you could actually increase it, then instead of teaching a subject, you would just increase intelligence” and they would be better able to learn the subject, whatever it might be. From the start, Peterson ran tests to see if his games were having the desired result. “People have been trying to improve intelligence forever, and it’s really, really hard,” Peterson says. His initial efforts confirmed this—the puzzles did not raise students’ scores on IQ or other intelligence tests he administered. Yet “brain exercise” programs that have become popular in recent years have shown that it's possible to improve working memory, and after some trial and error, Peterson found he could get the software to improve working memory for a targeted area. “This was a revelation,” he says. “If you are teaching math, then having improved working memory for working with numbers and other mathematical constructs is helpful. We’ve increased what is needed to solve math problems, which is very useful.”
Scores doubled after kids did Peterson’s puzzles. But when he tried to repeat the gain, he couldn’t. The reason? “It didn’t actually work,” he says. In the first instances, the software had engaged students. The takeaway? Simply teaching math wasn’t enough. They actually have to like it.
So he searched for other ways to gauge whether his puzzles could facilitate learning breakthroughs. He turned his mind to standardized tests, which are designed to measure students’ retention of the content they’re taught and improvements in their skills. Because this type of testing is geared to specific subject areas, he would need to focus on just one. Math was the obvious choice. “The one that’s closest to puzzles is mathematics,” he says. (Peterson also knew that math is a subject on which all 50 states administer standardized tests—unlike science and other logic- or reasoning-related subjects. A program that improved math skills would be valued by educational administrators and funders.)
Instead of relying on the puzzles he’d been using, Peterson created ones infused with math lessons. At the time, he was working with low-income students in South Los Angeles. He was ecstatic to find that their standardized test scores doubled after they’d been doing the new puzzles. But when he brought the software to other schools and tried to repeat the gain, he couldn’t. The reason? “It didn’t actually work,” Peterson says.
The discrepancy was that in South L.A., Peterson had spent a lot of time with the students, many of whom had never used computers and were excited by the novelty. The ST Math software had acted as a catalyst to engage the students in math and, with the help of Peterson and the teachers, positively influence the learning culture of the school. In the schools that weren’t able to match the early results, Peterson and his staff didn’t have a strong physical presence and weren’t stirring up enthusiasm for math learning. Peterson’s takeaway: Simply teaching math wasn’t enough. “They actually have to also like it and want to do it,” he says.
This jibes with research on the effect of tech in the classroom—that technology is something for teachers to add to their quiver of tools, not a solution in itself. “Where technology becomes the thing, as opposed to a tool to help an educator do a much better job, it’s inert,” as Levine of the Sesame Workshop put it.
Peterson and his staff threw themselves into more research on how to make the puzzles better and more engaging—how to transform the culture of learning through the software. They discovered that cultivating persistence in students was key; the puzzles needed to be challenging. They also eliminated multiple choice–type problems in favor of those that required students to build solutions. Instead of simply indicating if an answer was correct, Peterson’s new puzzles were embedded with elaborate feedback that was delivered in the form of animations showing why a solution either worked or didn’t work. “In these environments,” Peterson points out, “you can learn a lot by making mistakes.”
He also created JiJi, an animated penguin that’s a positive, motivating presence to guide students along and that teachers, who are an indispensible element in tech learning, use to continuously stoke students’ passion for math. The revamped software worked remarkably well; other schools started to see the same gains Peterson had witnessed in South L.A.
Several years ago, MIND Research Institute’s work got a boost when a study by scientists at UC Irvine revealed that the highest correlating factor between what kids know in preschool and how well they do later in their academic and professional careers is early math skills. It was a better predictor of how kids will do in reading than reading is of reading. The finding shoudn't be viewed as conclusive—another respected study pointed to small motor skills as the main factor—but for Peterson it was a motivator. He felt that the need for ST Math was even greater than he had understood, and redoubled his efforts.
The next step was to get ST Math into more schools. Peterson sought out historically low-performing schools in cities such as Las Vegas, Chicago, and New York, and he created partnerships with corporations and foundations, which provided the funding needed to cover the cost of the software and teacher training and support. Though Peterson had created the program for those with learning challenges, he quickly realized it was well suited to all students.
In 2013, an independent evaluation by the research firm WestEd confirmed MIND Research Institute’s finding that ST Math improves test scores two to three times faster than non–ST Math curricula. Today, 630,000 students are using the software in 2,050 elementary schools throughout the U.S. But Peterson’s goals are more ambitious still. “What we’re after as an organization is to get 100 percent of students to be proficient in math and love math,” he says. “I’m of the belief that all students cognitively have what it takes.” By this he doesn’t just mean fractions and geometry—he wants all students to master calculus, which requires a solid foundation in early math skills. “You can trace it all the way down to early elementary school,” he says. To that end, Peterson plans to expand ST Math to cover all grade levels.
In the first-grade classroom at Weaver Elementary, a student I'll call Macon, who had a glimmering Southern California tan, was breezing through the ST Math puzzles. “He’s a different thinker,” Kominsky, the principal, explained. “He’s not a sit-still-and-listen guy. But he’s in there and he’s getting it. This makes total sense to him because he thinks in pictures.”
The imposition of standardized tests on younger and younger children has resulted in an increase of the teaching method known as "direct instruction"—i.e., a teacher at the front of the room giving a lesson and a roomful of kids sitting and listening. That's upset advocates of child-centered, play-based instruction in early childhood education because it imposes unrealistic expectations on kids that age, especially "different thinkers" like Macon. But if a six-year-old boy can sit still long enough to learn some computational skills, improve his working memory, and have fun doing it, perhaps ST Math can find supporters for its approach among folks on opposite sides of the debates over education reform.
JiJi the penguin scurried across Macon's screen again, signaling he’d solved another puzzle, and when Macon logged out, the program showed he’d successfully completed 13 objectives that morning.
Kominsky then took an informal poll of the first-graders, asking whose favorite subject was math. All but a little girl in a brown-and-white polka-dotted dress raised their hands. Considering that getting kids to love math is one of Peterson’s top goals, this was a sign of success. When Kominsky then inquired how many thought they were good at math, all of their hands shot up, including that of the polka-dotted one.
This line of questioning reminded me of something Peterson had mentioned to me earlier. Last Christmas he went online to see if any students were doing ST Math on the iPads and other devices they’d received as gifts. What he found surprised him: 6,800 students were logged on to ST Math—on Christmas morning.
After spending time at Weaver, I tried out ST Math by doing some practice puzzles I’d found on the MIND Research Institute website. They weren’t as easy as I’d expected—but then, I hadn’t done any math since my high school calculus class. I stuck with the puzzles, as I’d seen the students do, until I’d mastered each one. Then I went back and did a few more, just for fun. Because they were a lot of fun, as I’d been hearing and seeing.
I wondered what other adults thought of ST Math, specifically experts in education. So I called up Eric Hanushek, a senior fellow at Stanford University’s Hoover Institution and the co-author of Endangering Prosperity: A Global View of the American School, to ask what he thought of the software. Hanushek related to me that he had participated last fall in a panel convened by the Business Round Table, a group that seeks out efforts it believes will help improve American competitiveness globally and strengthen the U.S. economy. The panel, composed of education experts, was asked to choose the five most effective educational software programs for elementary and secondary schools students from a crop of roughly 100 entrants.
MIND Research Institute was one of a group of 10 finalists asked to make in-person presentations. Whereas the other finalists trotted out PowerPoint decks or launched into long lectures, Peterson and a colleague showed up with an iPad and simply did ST Math. The panelists, like me and the six-year-olds, were transfixed.
“They’re getting people to have what we call higher-order skills—the ability to make inferences and to figure out patterns that lead you to do other things,” Hanushek said of the software, which was selected as one of the top five. But what surprised him the most was this: “They engaged a group of well-educated, seasoned analysts and policymakers in fifth-grade math.”
There's no question technology like ST Math will become more prevalent in classrooms—it's just a matter of how rapidly the shift will take place, and which versions will survive. If the Project Tomorrow study is any indication, as new teachers are trained, the profession will become more comfortable with it. Still, as Peterson's experience in South L.A. showed, enthusiasm for teaching, and for motivating students, will be as important as ever.