No, no, no … you’ve gotta say it like, “data interoperability!”
Over the past several days, I’ve had the opportunity to spend time with district leaders from around the country as we learn together about a fantastically riveting topic: data interoperability! (see, I said it right)
So what is data interoperability?
In a nutshell, data interoperability relates to the ability of systems that create, store, or consume data to seamlessly talk to one another. Once the systems can speak to one another, data can be pulled – generally in real time – for analysis, regardless of what system the data originally resides in.
Specifically, we’re looking at implementing the Ed-Fi Standard from the Ed-Fi Alliance, a non-profit organization supported by the Michael and Susan Dell Foundation. This standard – which is freely available – will allow the district to develop a secure, internally-maintained operational data store and data warehouse that would allow the district to collect and utilize data regardless of the system it was created in, without a tremendous investment of time or money.
So what? Why is this important?
By any definition, we are data-rich; schools possess a wealth of data relating to students, teachers, programs, facilities, geographic distributions, and beyond. This includes the data most would associate with schools, such as assignment grades, course grades, test scores, student grade level, course enrollments, activities, and so forth. Beyond the most obvious data, we also maintain data like bus number and transit times, library resource checkouts, staff member years of experience and degrees attained, technology support call frequency, time spent using specific educational software applications, and many more.
While we have access to this data, we – as with many school districts – don’t possess the tools, structures, or funding to maximize our use of data. So back to the question of importance; if we possess all of this data, it is critical that we leverage that data in ways that support constant improvement across all areas.
Let’s take a look at an example:
Let’s say that student attendance has been identified as a priority for improvement. In order to try to improve student improvement, the district launches four new programs at the start of a new school year.
- One program includes school staff visits to each family enrolled in the school, providing an opportunity for families to learn more about the school, the importance of attendance, and how they can be involved.
- Visit data collected in a Google Sheet
- The second program is a rapid notification program where we would leverage our mass notification system to send text messages to parents and students within a few minutes of an unexcused absence being reported, using a system that would allow bi-directional communication for parents/students to follow up via text with information or questions.
- Absence data collected in PowerSchool, notification data (and call answered results) in our mass notification system
- The third program is a public information campaign, and includes strategically-placed billboards promoting the benefits of regular attendance at school.
- Billboard locations and durations maintained in ArcGIS
- The fourth program involves reaching out to absent students to remind them that they can access course content and assignments using Canvas; this is done regardless of the reason for the absence
- Data regarding whether students access course resources is maintained
One challenge in terms of determining the effectiveness of these programs, though, is the fact that data is stored in different platforms that don’t talk seamlessly to one another. Further, the variables we’re hoping to influence – attendance rates, course completion, assessment scores, positive behaviors – are also stored in different systems.
An effective operational data store / data warehouse would break down these barriers. With a single piece of (free) software and no coding or APIs, the district could develop a dynamic dashboard that could show cross-platform reports such as:
- a geographic heat map color coded by attendance rates and/or change in attendance rates, with billboard locations identified
- An overview of district attendance rates that could be drilled down to building, geographic region, or even an individual classroom, which would be filterable by whether or not students’ families had received a visit
- A chart showing course achievement metrics (assignment completion, course grades, etc.) by students who did and did not, respectively, access their course Canvas page during or immediately following an absence
- A similar chart showing the same type of information for students who engaged – or whose parents engaged – in bi-directional communication following the text message absence notification
You can probably think of other reports or visualizations that could be generated using the data available; the real power of data interoperability is the ability to do this without coding, development of manual reports, or other activities that we can’t rely upon in order to meaningfully analyze the effectiveness of the hundreds of programs that are underway in the district at any given time.
Where can I learn more?
Keep an eye on the blog; as we progress, I’ll be sharing examples of our work – and the results of our work – on the Technology & Innovation blog. To learn more about data interoperability in general or to see some examples of what other districts have been able to do, I’d encourage you to check out the Ed-Fi Alliance page or the data interoperability-focused Project Unicorn.