Successful Innovation Starts with Finding the Right Points of Integration

As we consider the potential of various innovations in our education system, it is helpful to look at that system through the lens of system architecture. A system’s architecture determines its constituent components and subsystems and defines how they must fit and work together. In any system, components work together through interfaces, and these interfaces can be either modular or interdependent.

Modular interfaces are clean, well-defined, standardized interfaces that make it possible for the teams developing subsystems and components to work at arm’s length. The drawback of modular interfaces is that they restrict designers’ freedom to innovate in how subsystems and components fit together. When components and subsystems must conform to modular interfaces, some design decisions are inevitably made not in the interest of providing the best service to end users but rather for the sake of making sure the component will be compatible with other components on the other side of the modular interface.

In contrast, interdependent interfaces are interfaces that have not yet been clearly defined and standardized. Successfully connecting components across interdependent interfaces requires a great deal of collaboration between the teams developing those components. Although this coordination requires extra effort, interdependent component interfaces can be optimized for meeting the overarching job of the system without having to conform to the requirements of rigid, modular interface.

As we think about how to leverage online learning to bring about a student-centered education system, the architecture and interfaces of our education system affect how innovations are able to develop. Because most schools and districts do not have the internal capabilities (i.e., software development teams, learning psychologists, and curriculum designers) to develop their own online-learning systems, these systems are of necessity being created largely by external providers. Of course when external providers working independently from schools and districts create the components, the components must connect with the education system through modular interfaces.

A few points are worth noting here. First, many online-learning components plug into the school system through the modular interfaces that were created for connecting traditional components. Second, modular interfaces exist at various levels of the subsystems that make up the public school system.

For example, classroom rotations were common in elementary schools long before blended learning. Thus, when elementary schools started integrating online-learning components—such as Dreambox, i-Ready, and ST Math—into their rotations, the scheduling and classroom procedures were already in place to define the modular interfaces for this integration. Basic Station Rotation models simply substituted computer-based learning for other traditional, station-based learning activities such as reviewing flash cards, playing educational board games, or working on projects. The situation is similar for Lab Rotation.

Implementing Station Rotation or Lab Rotation proves to be more challenging when schools want to integrate online learning not only with the rotation schedule, but also with the instruction provided by the teacher. Many schools would like their teachers to be able to use software data to more effectively group students and adapt their lesson plans, and they would like to have the software adjust what it presents to students according to what has been covered during teacher-led instruction. But most teachers are not trained on how to do this type of data analysis and most software packages do not present data in formats that meet the specific needs of the teachers. In other words, modular interfaces between online instruction and teacher-led instruction are not yet defined. As a result, schools that want to capture the benefits of instructional integration must do a lot of heavy lifting to figure out work-around interfaces with the software. Unfortunately, because schools can only control the teacher side of the interface and not the software side, these integration efforts are often less than satisfying.

In contrast to Station Rotation models that plug online learning into the education system at the level of in-class learning activities, A La Carte blended-learning models plug into the education system at the level of an entire course. By integrating across a larger sub-system, A La Carte course providers have the design freedom to optimize interdependent interfaces between online-learning tools, curriculum, assessment, daily schedules, and teacher-led instruction. But although A La Carte providers can control the interfaces between the components of a course, the course as a whole must still conform to the modular interfaces specified by the schools, districts, or states that hire them. These modular interfaces include such things as assigning A through F course grades, awarding course credits, and structuring course schedules on a semester basis.

Blended learning promises to greatly improve student achievement by making learning more personalized to students’ individual needs. But there is still much work to be done to improve the technology and the models until they fulfill that promise at scale. As innovators move this work forward, they need to be conscious of how they connect to the architecture of the school system because the points of connection they choose will determine which components of the system they can control and the degree of freedom they have to innovate on those components. This fact is important because designers of innovative systems are best able to address the needs of students when they are integrated across the performance-defining components of the overall system.

This article was written by Thomas Arnett for the Clayton Christensen Institute.