The Center for Data Innovation spoke with Jack Lynch, CEO of educational software company Renaissance Learning. Lynch discussed the benefits of data-driven adaptive learning systems and his vision for the future of “learnalytics.”
Travis Korte: In addition to supplying educational technology, Renaissance bills itself as an analytics company. Can you elaborate on that?
Jack Lynch: The best way to describe us is a cloud-based assessment and learning analytics company. We principally use computer-adaptive testing to identify where students fall on a particular learning progression. That’s the way in which students acquire knowledge, from when they enter to when they exit the K-12 school system. As you know, everyone falls into a grade based on their age, but students learn at different paces. So we use computer-adaptive testing. If you get an incorrect answer to the first question, we’ll adapt the next question to figure out what you do know. So we identify the point where you fall on the learning progression, and we give that to teachers who are able to figure out what to do next for that particular student. It’s a little like a global positioning system to plot where students are on their learning path.
That’s in essence what we do. We produce a rich set of data that helps educators know about school district, or group of students, or individual students. Then teachers, administrators, and principals can use data to make data-driven decisions.
TK : Talk about your adaptive learning tests. What are some of the benefits of adaptive learning versus giving everyone a standard set of questions?
JL: The typical test that I had when I was going to school was a sixth form test. The problem with that is that it doesn’t really help an educator determine what a student knows. It helps determine what a student doesn’t know, which is somewhat helpful. But you could give a test to a student and they could get zero out of 20 questions right. That helps a teacher understand what the student doesn’t know, but not what they do know.
TK: You’ve been in the testing business for a long time, so you’ve seen most of the trend toward data-driven education. Do you think there was a precipitating event there, or was the trend gradual?
JL: It has been evolutionary. The standards movement really came into focus under the Bush administration with No Child Left Behind and then testing became more prominent. But it was a relatively blunt instrument back then. Now it’s become fairly sophisticated to help educators not only know where they are in relationship to standards mastery, but it also provides rich data on student learning that allows a lot of educators to pinpoint missing skills and gaps they need to fix. The phase we’re in is using data to drive instructional decisions. The next phase we’re beginning to segue to is using assessment data to drive recommendations. So go one step further and make a recommendation a teacher can use for a group of students or an individual student. That’s probably the most interesting thing that’s part of a new era in education. The parallel for that can be found in medicine, where we’re beginning to use genomics to understand the genetic makeup of an individual and tailor treatment to that exact makeup. The parallel is what I like to call “learnalytics,” which is to get rich understanding of the learning DNA, if you will, of a student, and use the data we know about that student to make specific changes to their education.
TK: Educational tests seem like they could be useful for academic research on a variety of topics. Have you worked with academics to let them use STAR (a tool which measures student skills in math or reading) data, for example, for research?
JL: We do a lot of research on STAR internally and we have a lot of products as well. It’s important for customers to have that evidence base. But the research you’re referring to is, for example, in math: what are most students in a particular socioeconomic category having the most difficulty with? And using big data to analyze that. But I think what you’re going to see, which is something that we’re uniquely positioned to intercept, is understanding the cause and effect relationship between an intervention that a teacher will employ for a student that is behind grade level and the outcome of that intervention. One of the things our teachers do with STAR for students that are behind is they will use an intervention and they will put that intervention into STAR and track how well it’s working. But for them it’s more of an art than a science. But for them they’re just using their own experience and what works for a particular student. What we’ll be able to do, not today but in the future, is provide an evidence basis between cause and effect between intervention and outcomes. A teacher who finds a similarly situated student can then look at what interventions work best for a particular situation.
TK: Some school districts have had trouble setting up cloud-based systems, in part because parents underestimate the security the cloud. How do you evangelize the cloud to educators who might not know a lot about it?
JL: For us it’s been an issue in the past but it’s no longer an issue. They know that everything we provide to them is from the cloud. They know it’s locked down, hermetically sealed, secure. That it is only available to them, to students and their parents. I think educators know that about us, so it hasn’t been an issue. It’s frankly more of an issue in the press. I think for companies like us who are very clear about the use of the data, which is for the school districts, educators don’t have to worry. We take data privacy very seriously.