Photos courtesy Asbury Park Board of Education

Students in Linda Smolinski's class learn directly from a computer for part of the hour.

A computer voice guides 12-year-old Amir Accoo to spell the words he hears through his headphones: emergency, bulldozer,  minutes. Accoo spells “minutes” wrong and is asked to try that one again, several times. Later, he clicks on a proofreading button.

“You check what you have wrong out of the spelling words I just did,” Accoo explained as he looked at different spellings of the word until he spotted the correct one and moved the cursor to it, “and you just click on it, like this.”

Accoo is in the sixth grade at Asbury Park Middle School, but because he is so far below grade level when it comes to reading, he goes to a new type of reading class each day instead of a traditional one.

Computer-driven classes are now spreading fast across the country to help bottom students catch up. Already more than 400 schools in New Jersey are using Scholastic Inc.’s  Read 180 program that Accoo is learning from.

Other companies are marketing similar software, too. And, increasingly, it’s not just for the weakest readers, but for all kids, in all subjects and in all grades. Schools are deciding how much time kids should spend in front of a computer in the classroom, without a whole lot of evidence about what works.

Accoo is so excited, he can tell you how many words he’s read each day. On this particular day, it was 341.

Going high-tech to teach reading

Asbury Park Middle School has three of these classrooms with a row of computers along a wall.  The students work independently most of the time without a teacher’s help.

Linda Smolinski has been teaching for 32 years. This year her students are mainly Mexican immigrants and Haitian earthquake refugees. At first she was skeptical about using technology.

“I said ‘No. My kids can’t do that. There’s absolutely no way. My children don’t speak English. How do you expect them to sit at a computer and do something when I’m not even there helping them?’” Smolinski recalled.

She was eventually won over when Scholastic, the maker of the software, showed an increase in her students’ reading ability after just three months. The real test, however, will come later this year, when the state’s standardized tests will show whether or not the program is really working.

Superintendent Denise Lowe was hired away from Central Islip Long Island to help fix Asbury Park’s school system. The district has been a favorite target of Governor Chris Christie because it has one of the  highest per pupil expenditures in the state, but its test scores are abysmal. Some 90 percent of the students at Asbury Park Middle School qualify for free and reduced price lunch. That measure of concentrated poverty qualifies them for extra school funding from the state.

Lowe said she’d had previous success with Scholastic’s Read 180 software in Long Island.

“This was pretty much the first thing I told my director to do that we need to look into Read 180,” Lowe explained. She spent $600,000, which she admits is a big investment.

The downside to technology

Some educators are critical of these fancy high-tech classrooms. Newark wanted to show off its computers, but technical problems caused them to cancel before a reporter’s visit. And there are other problems.

“They’re hearing distracted voices, these unnatural voices and not seeing human lips mouthing the words,” said Sandra Priest Rose, the chairman of the Reading Reform Foundation. Her organization trains teachers to teach reading in poor schools in New York City.

“It’s terribly important to have that human interaction and that spontaneous interaction that you cannot get on a computer,” Priest Rose said.

She added neuroscience research has found that the act of writing helps the brain learn better than tapping on a keyboard.  Although, she will concede the software program can offer some useful data.

Smolinski said the program provides detailed information about the students’ skill gaps every day.

“On one of our first tests, most of them scored high 80s low 90s. Most teachers would say, ‘Fine, I’ll just move on.’ I noticed that every one of them bombed antonyms. That told me that I did not teach antonyms and I didn’t,” Smolinski said.

On the second test, her students again scored well. But the data report showed that they bombed capitalization.

And then there’s the Groupinator. It takes that skill gap analysis and places students in a group with kids who have similar needs.

“It tells me exactly what they need, when they need it. It does all my work for me,” Smolinski said.

Using both old and new techniques

In Asbury Park, part of the $600,000 investment included bright-green bean bag chairs and cushioned rockers, where, after their computer time, kids like Accoo can kick back with an old-fashioned book.

“I get to read my favorite books,” Accoo said. “Right now I’m reading Captain Underpants. The book I just got done reading recently was Frankenstein. And Frankenstein is mostly about this scientist and he creates life.”

Asbury Park’s teachers wish they could doctor more of these classrooms to serve all of their failing students. But right now there’s room for only 120 students a year.

This story was a collaboration between The Hechinger Report and New Jersey Public Radio.

TB

A Rocketship student works at the Learning Lab using adaptive software.

The question keeps coming up: What technology should schools invest their money, time, and effort in? During this fraught time in our economy, the decision to invest in tools like adaptive software and other tech devices is sometimes portrayed as excessive or wasteful.

In Sunday’s New York Times, Matt Richtel and Trip Gabriel wrote about software program companies inflating their effectiveness in schools, and how they “ignore well-regarded independent studies that test their products’ effectiveness.” In the next couple of days, we’ll deconstruct the writers’ sources of information — namely the main source for their claim that the technologies are ineffective, the What Works Clearinghouse.

In the meantime, I spoke to Aylon Samouha, Chief Schools Officer at Rocketship Education, a network of charter schools in the Bay Area that uses software to reinforce basic skills mastery. (You can read more about their hybrid learning program and their competitive scores in this MindShift series). Samouha is in charge of the design and strategy of Rocketship’s hybrid learning model, as well as its teacher and principal training program, among many other things.

Samouha, who lives and breathes educational software and is consumed with finding the best way to integrate technology into the school day, has a very different perspective than what Richtel and Gabriel portray.

First, the facts. In an independent study released in August by SRI International, which conducted a randomized controlled trial using DreamBox Learning, those who used the program for 16 weeks scored 2.3 points higher on the Northwest Evaluation Association math test than those who didn’t — the equivalent to progressing 5.5 points in percentile ranking (for example, from 50 percent to 55.5 percent).

It’s important to note here that not all assessments are created equal, but Samouha believes that the NWEA is an adaptive diagnostic test and a dependable measure because “millions of kids taken it,” he said.

Those students’ gains are not only statistically significant, he said, but the fact that they were achieved in such a short time — 16 weeks — indicates that using the program for an entire school year would show even greater gains.

But Samouha is not a tech evangelist just for the sake of using tech. “It’s true that there’s a lot of time wasted on computers right now. There are wrong ways of doing this. But it’s so clear that we have to figure out how to educate students in the 21^st century in ways that go beyond the traditional classroom model, which was created in the 17^th century in Prussia. The fact that it needs to change is not a question in my mind.”

So how to make this happen? “The more we integrate software with what’s happening in the classroom, the better results we’ll see. That will take work and investment. Schools will have to invest time, content providers will need to invest time and energy into make it more plug-and-play and make it more integrated into the school day.”

But Samouha says the Times article proves the opposite of what the writers end up conveying. “It shouldn’t be a call to stop for investment, but a call to invest more, because we need to get it right,” he said.

What “right” looks like can be debatable, but Samouha believes he has it nailed down — at least in the case of elementary schools. His criteria for successful software are:

• Aligned to common standards, “so we’re all speaking the same language.”
• Data needs to be seamlessly integrated and programs need to be able to talk teach other so teachers can easily make sense of the data.
• Rather than be standalone in their environment, they need to be part of the open Web, so they can be linked to each other. “Right now there are a lot of walls between programs, just like before Steve Jobs agreed to use Microsoft, so people using Macs couldn’t open Microsoft programs,” he said. “Those walls exist now, if you’re trying to quilt together a number of interventions to put the right thing in front of a child at the right time, it makes it much harder when those walls exist.

That’s what Samouha is working towards. But it won’t magically happen. It takes investment from schools, from vendors, from educators. “But we can’t just abandon ship and go back to the traditional school model.”

At the heart of this discussion for the general public and most major media is the question “why.” Why should we bother to invest in all this when we (adults) all turned out just fine with the way we were schooled? Why can’t just teachers keep teaching the way they have been?

Samouha believes that’s a false dichotomy. He’s not arguing that teachers should be replaced by software, but that students will benefit from an array of different learning methods. Software will help them practice drills and basic skills, but teachers will be always be the most important part of the equation.

“People are arguing that it’s either human beings teaching kids or computers,” he said. “But that’s not our answer for anything else. When I want to communicate with my wife, I talk to her in person, I use my cell phone, I e-mail and text, that’s the way of the world. But for some reason, education is exceptionally not like that.”

I also spoke with Samouha about the difference in quality between math and literacy software, and I’ll report back on that in the coming days.

Knewton

As students do more of their work on computers, new technology is able to track their performance in ways it couldn’t before. It isn’t simply a matter of which answers a student gets right or wrong, for example, but how much time they take to answer questions, how and when they hesitate or stall. Taking this data, engineers can build algorithms that are able to examine students’ work and help deliver to them a personalized, or “adaptive” learning solution.

Adaptive learning technologies have long been considered a crucial component in helping students progress at their own level, and until now, it’s only been used here and there in the K-12 setting and with test preparation companies that help students ace their SATs, GMATs, and the like. But one company in this space, Knewton, has made big strides towards making its platform available in schools, not just at home.

This week, Knewton announced that its adaptive learning platform had been selected by four schools — Penn State University, University of Nevada, Las Vegas, the State University of New York, and Mount St. Mary’s University — to help power college readiness courses. These online, self-paced classes are designed to help incoming students who may not be ready for college-level academics. This is an important group to help succeed in school at this crucial stage, since about 25% of students who enter college need some math remediation and about 50% of students who require remediation fail to graduate.

Using adaptive learning technology, these classes can identify the areas in which students need help, deliver content specific to those needs, and deliver it in such a way to help a student build on what she or he knows and how she or he learns best.

The universities that are implementing Knewton’s Math College Readiness Course join Arizona State University, which has been using the technology since earlier this year. An indication, perhaps, of the success of that program is ASU’s announcement that it plans to add the adaptive learning technology to other classes beyond just the math mediation ones. Knewton’s platform will be used in two new, blended learning classes — MAT 117 (College Algebra) and MAT 142 (College Mathematics) — for which the company has partnered with Pearson in order to develop the curriculum.

Ed tech experts, like Diana Rhoten, director of the Knowledge Institutions program and the Digital Media and Learning project at the Social Science Research Council, have been anticipating a product like this for a while.

“The holy grail for any company is not just creating a product that gives instant feedback, but that has a truly adaptive learning engine. And there are few that really do,” Rhoten said in an interview last year. “I don’t use the term adaptive learning loosely, but the market is starting to.” She mentioned Knewton as one of the trailblazers in the field that’s working on engineering the technology to truly adapt to users’ response –  collecting data over time and understanding patterns from the user’s mistakes.

It looks as though more universities are interested in taking advantage of adaptive learning — it might just prove to be a groundbreaking way to leverage technology for more effective learning.