Solution aims to transform math assessment

In school systems across the country, teachers are using handheld computers and a software solution from Wireless Generation, called mCLASS, to administer the Dynamic Indicators of Basic Early Literacy Skills (DIBELS) to elementary-age students. The technique has helped boost students’ reading scores dramatically. Now, the positive impact this approach has had on reading soon could be replicated in math.

Wireless Generation, along with the Teachers College at Columbia University and the University of Missouri-Columbia, recently received a four-year, $1.5 million grant from the U.S. Department of Education’s Institute of Education Sciences to develop a math-related version of mCLASS.

As with the company’s reading solution, mCLASS Math would allow teachers to administer one-on-one assessments with K-3 students, while recording results and observations on a handheld computer. Teachers use the Palm-based software to guide themselves through the assessment process and record students’ responses. Once an assessment is completed, teachers sync their handheld with their desktop or laptop computer to transfer the data to a secure web site, where they can examine the results almost instantly to inform their instruction.

The math version of mCLASS seeks to enable teachers not only to screen for math proficiency and monitor students’ progress, but also to learn about their students’ thought processes.

The system works by giving students a math problem and instructing them to solve the problem any way they want. During the assessment, students are asked how they came up with their answer.

Although Wireless Generation is not the first company to develop a handheld-based solution for math assessment, the company might be the first to target the K-3 level.

To understand how children learn math, it’s worth examining their thinking processes as early as kindergarten, explained Herbert Ginsburg, a professor at Columbia University Teachers College and co-developer of mCLASS Math.

“We find that little kids, starting in kindergarten and going through third grade, have very interesting ways of thinking about math, and teachers need to learn about that,” Ginsburg said.

Young learners, Ginsburg said, might have certain systemic ways of using strategies that can result in mistakes. “What teachers need to learn is that if a child gets a wrong answer, there may be a reason why, or a strategy behind it,” he said.

Ginsburg’s beliefs led him to connect with Wireless Generation. Starting out with a grant from the National Science Foundation, Ginsburg and Larry Berger, Wireless Generation’s co-founder and CEO, began working on ways to use handheld computers to help teachers interview students about their knowledge of math concepts.

Berger said Wireless Generation’s customers were asking if the company offered an early math assessment similar to its widely successful early reading assessment.

In districts throughout the nation, such as Chicago Public Schools and Florida’s Orange County Public Schools (OCPS), mCLASS Reading is having a profound effect on reading achievement. At OCPS, the solution has enabled teachers to get assessment data back in a third of the time, allowing them to adjust their instruction accordingly and target students’ weaknesses far faster than they could by using paper-based assessments.

For mCLASS Math to have the same effect on math achievement, its assessment tool must effectively pinpoint students’ strengths and weaknesses. The four-year, $1.5 million federal grant is intended to help Wireless Generation develop these assessments and gauge their efficacy.

The primary goal of this research is to look at the adequacy of the assessments from a scientific standpoint and determine whether they are reliable, as well as if their scores are predictive of how students will perform later.

“Just like in reading, … there are many things that you can identify in kindergarten and first grade,” Berger said. “We see that as the highest leverage point in the whole education system, because it’s expensive and difficult to remediate problems that have been allowed to get worse and worse by fourth grade–by then, cognitively, our brains are less flexible, and emotionally, kids have sometimes figured out that they are not good at math. At ages four and five, it’s rare that [students] have emotional barriers to success. That’s why we’ve focused on that [age].”

Ginsburg said his research has revealed that children seem to fall into four groups when it comes to math problems.

The first group of students will get the answer right and will understand the process behind the math problem. A second group arrives at the correct answer, but students in this group can’t always determine how they arrived at the right answer. Students in the third group have a good understanding of the process, but get the answer wrong owing to sloppy mistakes. And students in the fourth group, Ginsburg said, get the answer wrong and don’t seem to understand the process or might need extra help.

“I see this as helping the teachers to understand the kids better–it’s not just to get a score,” Ginsburg said. “Don’t think of your students only as people who get the answer right or wrong; they have concepts and strategies, and that’s what we have to focus on. Once a teacher finds out a student has one concept but not another, then we try to link all this up with instructional suggestions for teachers.”

Leslie Koske, a math consultant at the Region 14 Education Service Center in west Texas, is working with five districts that are piloting mCLASS Math. Those districts already were using Palm computers to administer the Texas Primary Reading Inventory assessment, so teachers were familiar with the hardware and only needed to learn the additional software.

“We know that an earlier intervention, in kindergarten or second grade, is far more helpful than in high school,” Koske said. “With this tool, we can evaluate weaknesses early enough to have a lasting impact on student learning.”

Koske said officials plan to evaluate the math test scores of students on third-grade state assessments. When second-graders who have used mCLASS Math reach third grade, they will have their test scores compared with the previous year’s third graders who did not use mCLASS Math, to determine the product’s effectiveness.

“We do believe it’s going to impact student performance,” she said, adding that mCLASS Math has been well received in the pilot schools.

Above all, students seem to enjoy the assessments, Ginsburg said.

“It’s hard for some people to understand that this isn’t a math test; this is a teacher talking with a student about how he or she solves a problem,” he said. “Kids love that attention and like talking with an adult who takes them seriously.”

He added: “I think we really are doing something interesting and unique.”


Wireless Generation

Teachers College, Columbia University

University of Missouri-Columbia

Laura Ascione

Want to share a great resource? Let us know at