In the first part of our interview, Daniel Schunk, professor of public and behavioral economics, talks about what schools should consider when creating a digital learning environment. He describes the characteristics of high-quality learning software that educators will be eager to work with, and explains why digitalization provides an opportunity to combat inequity in education.

Sabine Gysi: Let’s start with a thought experiment: About a year from now, at the beginning of the next school year, a large truck will stop in front of your elementary school to unload 1000 laptops, which are to be used to promote student learning. The school is not yet fully digitalized. What needs to happen between now and then? Starting today, what steps have to be taken, and by whom, if digitalization is to be successful?

Daniel Schunk: By the following year, the necessary IT infrastructure will be in place. That’s a good start! But it will take a great deal more than that – and will probably cost much more than the laptops themselves.

The first point I would make is that you have to know what you can do with the laptops. That means having a plan for using them, as well as the necessary software and teacher handbooks. I’m mentioning this first because it is very important, especially in Germany.

There are many different approaches and programs to choose from, of course. But we still lack evidence-based criteria that will help teachers select high-quality programs, and hence there is a lack of transparency. I argue that ministries of education should work closely with teachers to develop criteria and promote transparency, so that teachers can identify the programs that are effective, in the sense of actually promoting learning.

Why is that so important? I recently read a study that confirmed what I have long suspected: Many of the learning programs on the market today have low levels of effectiveness. This is simply because private companies have little incentive to ensure that their software does in fact promote learning. It’s fairly easy to sell people a good feeling, leading them to believe that they are learning something when they use these tools. But educators have every reason to demand much more; they obviously want software to be truly effective at promoting learning.

“We still lack evidence-based criteria that will help teachers select high-quality programs, and hence there is a lack of transparency.”

Given the current trend toward digitalization, some ed-tech companies are simply focusing on being the first to make a lot of money by selling to schools and parents. But is their software actually good? Someone, such as a central educational authority, should make that determination. Practice-oriented research is needed if software is to continue to improve.

The second point: Teachers need to be trained over the coming year. The goal is to help them overcome their reluctance and to show them what new technology can do. It’s not enough simply to press the “on” button, download the software and start using it. A great deal can be accomplished with good software, but it needs to be integrated appropriately into the respective educational context. The ultimate goal is to support teachers. And software is far more effective when teachers really know how to use it.

The third point is this: It is important to provide comprehensive information and encourage parental involvement, since some parents have concerns about ed-tech and are resistant to its use. Unfortunately, many people still see digitalization not as an opportunity, but as a threat. In part, this is because media discussions of digitalization in education are often emotional rather than substantive.

The fourth point is particularly significant in Germany, although it is important elsewhere as well: Reliable, nationwide high-speed wireless internet coverage is essential – both within and outside schools.

Fifth, legal issues need to be clarified. Schools – principals and teachers – understandably require a certain level of legal certainty if they own the laptops. School personnel mustn’t get into trouble if – despite built-in parental controls – children download something they shouldn’t.

“Many of the learning programs on the market today have low levels of effectiveness.”

And my last point is this: Many schools have reported that they underestimated the resources needed to maintain this technology. It would be useful to designate someone who is responsible for IT issues throughout the school district; this would make life easier for teachers.

SG: It sounds like the school in our example should get to work! As schools are introducing digital learning, do you find that they are collaborating with one another?

DS: Yes, there is some communication and coordination, but for the most part it takes place at an informal level, requiring considerable commitment on the part of individual teachers. In many areas, the structures are not yet in place that would make cooperation truly efficient and effective. Of course, schools would benefit greatly from working together, not only on content, but also in matters of infrastructure, including computer support.

But it’s also important to allow schools substantial freedom. At the moment, different schools are choosing different approaches – which leads to a great deal of heterogeneity. Schools will learn more from sharing their experiences than from a system in which everything is centrally coordinated and then simply passed on to the schools in a one-size-fits-all way.

SG: Schools are becoming increasingly heterogeneous, and more and more children are already behind by the time they start school. Will the personalized learning systems of the future help to level the playing field, by providing individualized support?

DS: I think that’s entirely possible. Digitalization provides a valuable opportunity to combat inequity in education. However, we are not as close to that goal as you might think. We are just beginning to understand how digital tools can facilitate personalized learning and individualized support. This will benefit everyone – high-achieving students and especially students who are struggling. It will make the education system more equitable.

Some approaches, such as the New Classrooms initiative in the United States, are already producing results. In Germany, too, we are conducting a new study that is showing signs of similar effects. In the field of mathematics, we created a digitalized learning environment that was in place for a period of five weeks, and there are already indications that it is having the desired impact. The bottom line is that initial studies have shown that this approach works.

Certain important requirements must be met, however, as we know from research on effective learning in the non-digital realm. For example, software must be adaptive and meet the needs of individual students.

SG: What features of adaptive learning software are particularly important?

DS: The real question is this: What kind of adaptivity actually promotes learning? To answer that question, we have launched a new research project focusing on older students at vocational schools.

Adaptivity obviously means assigning tasks to children that are appropriate for their level, neither too easy nor too hard. That means individualizing, to some extent, the level of difficulty and even certain aspects of the curriculum. But does it always mean pushing a child to the limits of his or her ability? Or are there times when you don’t want to do that? Is that precisely what high-achieving students need, while weaker students may benefit from the inclusion of playful elements that guarantee them a sense of achievement?

More research is needed; it is not yet clear how the learning systems that are currently available can be implemented more broadly. But many people believe that this is possible, and that it will happen in the future – and I share that opinion.

SG: Does the success of such learning systems depend in large part on teachers – and on whether they are able to reinforce the system’s positive effects?

DS: Yes, I think that’s true. In all of our studies we have observed very positive responses from teachers, both in Germany and in Switzerland. In the case of truly high-quality software, even teachers who are initially reluctant quickly recognize its benefits. They realize that the software isn’t going to replace them, but simply offer support.

Let’s assume that teachers use the software in twenty percent of their lessons. This frees up time that they can devote to what is really important to them, such as providing coaching and support during the learning process and responding to the needs of individual students, for example those who are struggling. While the teacher is offering individualized support, the software is testing the limits of high-achieving students. Advanced students can focus intently on a specific task, while the teacher has time to explain something to a weaker student. When software is well made, it quickly becomes apparent to teachers that the concept works.

“Schools will learn more from sharing their experiences than from a system in which everything is centrally coordinated and then simply passed on to the schools in a one-size-fits-all way.”

SG: You have explained why learning software should be adaptive. Now could you tell us what other characteristics high-quality software should have?

DS: In order to promote the learning of both low- and high-achieving students, software should allow for individualized, targeted and direct feedback. We know from the psychology of learning that feedback needs to happen quickly, and it needs to be individualized. Most teachers would like to provide rapid feedback to every child, in the interest of that child’s growth and development. But that’s obviously impossible when they have a large class to attend to.

This is precisely where high-quality software can be extraordinarily helpful. Top-quality software also communicates feedback to teachers, so that they can quickly identify gaps in a student’s learning and adjust their instructional methods accordingly.

Software needs to be easy and convenient to use; it should be adapted to the curriculum and lend itself to modular use. To cite one example: Programs designed for practicing arithmetic skills are wonderful, but not if teachers can’t integrate them into the curriculum.

Over the long term, it should be the norm for traditional learning methods and computer-assisted methods to complement one another, in keeping with the “blended learning” approach. And this will be reflected in the curriculum.

SG: That makes sense. However, you mentioned earlier that manufacturers lack incentives to produce high-quality software. Why is that?

DS: Established textbook publishing companies have long recognized that it would be sensible for them to add digital learning programs to the list of products they offer. They are looking closely at this issue and having plans drawn up, internally or externally. The ideal, of course, would be for them to develop targeted, modular concepts geared to the curricula – and this is already happening.

Startups are also playing a role in this area; this is a very good thing, and should be encouraged. In this new field, some competition and heterogeneity is healthy. However, startups are under a great deal of financial pressure. Obviously, one of their goals is to help children learn. But I know, from my own experience, that if you truly want to evaluate the long-term impact of a software program, it has to be in use for a school year, or at least part of the year. Then you will be able to assess that impact by comparing the treatment group with a control group.

“I hope that in the future we will see more active cooperation among software developers, teachers and researchers.”

Researchers need to put in the necessary effort, but so far, too little is being done. So the first goal of a startup may be to get its software on the market, and this is entirely understandable. And as is always the case with new products, many are not yet fully developed.

I hope that in the future we will see more active cooperation among software developers, teachers and researchers. That will make it possible to identify the features that characterize high-quality software, and programs that are truly effective at promoting learning will ultimately conquer the market. I am confident that this will happen. I am grateful to the many companies that are working in this area, since researchers and school systems lack the necessary resources to develop such software on a wider scale.

SG: Am I correct in assuming that this kind of partnership might also make it more complicated to develop new products?

DS: Yes, since researchers will have two questions: First, does the software have positive impacts? Does it promote learning in a direct way? Second, are there potential negative side effects? Just think about the pressure that is put on students to succeed. Poorly designed and implemented software might make them more self-centered and perhaps more nervous, or it might have other negative effects. These are the kinds of issues that a careful, evidence-based approach can prevent. When software is well made and properly used, everyone benefits.

An environment characterized by productive competition to create superior designs is therefore helpful. The startups I am familiar with in Germany would benefit from a more positive environment. I also believe that open-source approaches are important.

All of these initiatives, at a variety of levels, should be encouraged. It’s great when startups have the courage to enter this new field. In such a new and rapidly expanding area, it is important, first of all, to explore as many widely diverse approaches as possible. This is the path that leads to true innovation.

In the second part of our interview, Daniel Schunk talks about important and often forgotten aspects of digitalization and the changing role of teachers.


Daniel Schunk is a professor of Behavioral and Public Economics at Johannes Gutenberg University Mainz (JGU), Germany. His work examines how preferences, skills and personality develop over the life cycle and what this means for the design of educational systems and for public spending. He is a founder of the research priority program “Interdisciplinary Public Policy” at JGU. His work in economics and education regularly incorporates perspectives from neuroscience and genetics.


Keep up to date with the BOLD newsletter