GUEST COLUMN | by Jeremy Friedman
The last decade has seen the rise of the Student Information System (SIS) and the Learning Management System (LMS) to help improve the workload of teachers and to help districts better prepare students. For the next-generation teacher and K-12 students, the Internet is an integral part of their lives that has become seamlessly integrated into their daily routines. Technology is everywhere, and its role in the learning lives of students, inside and outside the classroom will only expand as computers get faster, and data mining stronger for personalized learning and better decision-making. However, there is a major hurdle holding education back — moving and managing data.
Data integration and data standardization are a major obstacle for school districts. As districts collect increasing amounts of information on their students, from attendance records to test scores, they are seeking new ways to store, analyze and view this data to improve the academic performance of students and help teachers improve their processes.
More recently, there has been the emergence of consumer-focused education technology focused solely on the individual teacher. Before these technologies, managing data was more feasible because a majority of the technology was managed and deployed at a school district level.
Now there is a combination of school district technology, as well as self-service tools that are adopted by students, parents, and teachers without direct involvement of school districts. Newer companies are often hesitant to integrate with these local systems because data synchronization is complicated and expensive.
Frequently, experts of the education system argue that the key to improve our education system is to increase the amount of data collected and analyzed. The conventional wisdom is that more data will allow us to identify, personalize, and improve the outcomes of students. I agree that a plethora of data has the potential to help push us in the right direction if we have access to the proper tools and expertise to assess it.
If and when this is accomplished, it can lead to better results and overall performance. However, in order to accomplish this there needs to be an overhaul of both technology and ideologies. If we examine what type of data currently resides across all aspects of a school district, it would include the SIS that stores student records and demographic information, the LMS that helps individualize instruction, library systems, transportation management systems, disciplinary records, etc. Combine this with the fact that most of these programs have overlap with a subset of the other’s functionality and therefore require shared amounts of data from other services. You can see how this becomes an extremely complex data exchange.
It may not be immediately apparent why this data is so important, but the implications can be far reaching. For instance, using an online learning platform like Schoology allows teachers, schools and districts to have a deep understanding regarding the amount of time students spend engaging with their schoolwork, to have an understanding of how certain assignments mesh with particular student learning profiles, and how certain items impact student performance.
The challenge that we are presented with is related to how we inform an instructor, guidance counselor, or guardian about the most productive methods of supporting a student. What if we could tell you that based on a given student’s grade on a test, we can identify exact areas of weakness and based on that student’s learning profile and usage patterns we could identify a few items that might help improve his/her performance? In an ideal scenario, this would be easy; however, the data required to make intelligent decisions exists in a disparate number of non-integrated environments.
The most common method for exchanging data is manual import and export. This process is slow, tedious, and prone to errors. For the past few years, people considered automating the import/export process as innovation; however, adding more moving parts to an antiquated process does not count as innovation. Others have discovered that a much more manageable experience has been to develop custom integrations that connect directly to databases. This can work, but major issues are ignored, the most important being scalability. There is a significant amount of legwork required to do custom integrations, and many cannot be reused.
Additionally, building custom integrations that connect directly to a database makes an incorrect assumption that the underlying data structure will never change. That may have worked in the past, but with new database technologies being developed almost monthly, the process is changing. Where applicable, companies are shifting more and more data from relational databases to non-relational databases. This potentially means a much more complicated and deeper skillset required to simply build and maintain custom integrations.
Luckily, new ideas are coming to the market and solutions are on the horizon. One major initiative for more than a decade has been the Schools Interoperability Framework (SIF). SIF has been a step in the right direction toward a service-oriented architecture model. The platform enables a publisher/subscriber model to enable real-time data transfer of academic information. Some school districts have successfully implemented SIF and it has helped their organizational processes. Unfortunately, it is not easy or cost-effective enough to do at scale. There are too many moving parts, since a school district would be required to have a zone integration server (ZIS), and each application would need to have a SIF Agent (application that connects to the ZIS).
A similar, but more efficient solution occurs with the rise of initiatives like the Shared Learning Collaborative (SLC) or with companies like GetClever and Learnsprout. These companies are developing standard open RESTful APIs that allow data to be shared back and forth using a universal endpoint. If you are thinking that this sounds very much like the SIF implementation I described above, that is because it has the same overall goal of data standardization. Instead of adding another layer, however, they are simplifying the process by abstracting the zone integration server into one centralized API, rather than one server per student information system. While a seemingly small change, this is a major design change because it means a smaller dollar investment for schools, since it is one less item that needs to be purchased or managed. It also means quicker startup time and use of data.
Lastly, a similar trend we are seeing is openness among vendors. My company, Schoology, has opened its application programming interface (API) and developer platform (http://developers.schoology.com) to allow other vendors and users to build custom functionality and to extend Schoology’s platform with integrations into third-party products. The only limitation is that in order for this model to work across the board, vendors need to reciprocate by also providing robust open APIs.
The majority of SIS providers share closed and proprietary API’s and are reluctant to work with other companies. Many of these providers are nervous that an overlap in functionality will eventually render their products obsolete. We are hopeful that this mentality will change in the near-term because the ease of use and interconnectedness of API-based platforms has already led to massive amounts of data sharing among companies in the education technology space. We are seeing this currently in our business with the release of our App Center containing integrations with Blackboard CollaborateTM and TurnitinTM.
One company that should be recognized is Pearson. Pearson announced in May 2012 that they were releasing a PowerSchool API for partners (disclosure: Schoology is a Pearson ISV Partner). We are hopeful that some of the other providers will follow and begin to release open APIs. To date, Infinite Campus has given no indication of any intention to move in that direction, nor have others such as Skyward. In the meantime, vendors with open platforms will begin to leverage API’s, either on their own, or through services like SLC, Clever, and Learnsprout.
There is no easy solution to the data problem, but one thing is for sure — the next few years will be exciting and foster an environment of data sharing and openness. If you are using a product without an open API, encourage them to create and offer one. In the end, we are all in this together with the ultimate goal to improve student outcome. It is just a matter of how and when we get there.
Jeremy Friedman is the Founder & CEO of Schoology.
Student management system is a software application for education establishments to manage student data. MyClassboard.com Fully Online School Management System is a database application that runs on a server, which is the center of your student information system.
Excellent post Jeremy, but I’ve got a burning question: Is it in the SIS company’s best interest to allow write-back capability to the SIS? Seems to me that Pearson might be selling their soul a bit here… Right now they own the undivided attention of teachers, parents and students who must use their grade book and portal for all assignments, grades, and attendance.
Wouldn’t opening up the SIS relegate it to admin-only status? A database in the corner? I suspect this may be IC and Skyward’s hesitation.
That being said, the writing may be on the wall. Having ‘once-and-done’ options for integration like LearnSprout and Clever make the age-old problem of SIS integration a thing of the past.
I work in a district that has partnered with Clever. They talk a big game, but ultimately the challenges remain. Also while free to schools and districts, rates are high for our partners who are bullied by Clever (and districts) to buy in at a high cost. Then at the end of the day the technology is still cumbersome to manage and the data flawed due to challenges on the ground.
It almost makes you miss the days of .CSV files for integration.
With real power comes real responsibility. I haven’t met a tecnology ‘Spiderman’ yet. As a parent, I am floored as to the number of ‘technogists’ that think nothing of how seemless data sharing will affect the privacy of children and the rights of parents. It’s all about data for you, please don’t attempt to cover that up by indicating it’s about education.
Data organization is a crucial step, but in my team’s experience, many schools lack the internal capacity (people, time, money) to translate that data into action. The industry appears to be taking an “if we build it” approach. In our experience, getting savvy data users into schools has a bigger bang-for-the-buck than great data org systems. We need to bridge that gap between data access and data-informed decision making before we can get value out of all the beautiful data management systems.