Discussing the State of Actionable Data in American Education (Part Two)
Editor’s Note: This is part two of a discussion held during a nationwide virtual meeting. If you missed part one, you can read it here.
As part of the continuing series on Knowstory Learning Groups, the Learning Counsel’s Analytics, Infrastructure & Security Learning Group examined the state of data analytics and actionability in a recent nationwide virtual meeting.
The Learning Counsel is a research institute and news media hub headquartered in Sacramento. Its membership includes approximately 215,000 superintendents and assistant superintendents, technology and instructional administrators, curriculum specialists, education publishers and educational leaders from pre-k through higher ed.
One of the benefits for Learning Counsel members is KnowStory, an agnostic sharing and recommendations platform designed to help schools with needed discovery and analysis related to digital content and curriculum. KnowStory Learning Groups explore vital topics and catalyze the curriculum industry. Each group has a sponsored (industry) chairmanship role as well as an education-side chairperson, live meetings online and an in-person working meeting at the Learning Counsel National Gathering each year.
In part one of the discussion, the group talked about the need for interoperability and the ongoing struggle to keep student data secure. The discussion continued, and following are excerpt of that discussion:
Leilani Cauthen is CEO of the Learning Counsel. According to Cauthen, we are still learning how to navigate the balance between humans and machines, especially when it comes to network security. “It's the human side of the equation,” said Cauthen. “That's true of security and the rest of the infrastructure discussion too. How do we make sure we protect kids? What's our role as humans? Or is it really all on the tech side? A lot of people think it's up to the tech people to figure that out and it's not anymore. There's so much going on with throwing an App into the classroom that requires that your front line, all your teachers, also understand their roles and responsibilities. We need to come out with something that addresses human roles and responsibilities with regards to data literacy, analytics, infrastructure and security.”
Gisela Albuquerque Weise is an Education Industry Market Specialist working for Canon USA, Helping K-12 School Districts streamline workflows, boost collaboration and create sustainable, flexible learning environments. She serves as the Industry Chairperson and sponsor of the Analytics, Infrastructure & Security group. “I invite everybody to think about the culture change that goes along with this. We all know that it's not easy to change culture, but I wonder if any of our educators/guests here today could share some successful stories or even failures that could be transformed into success stories down the road.”
Stephen Bosacker, Principal Consultant for Arete Consulting said, “I'm seeing the same thing at the governance level. With charter schools, board members are typically parents, teachers or community members. Their background in areas of governance is weak to start with. They'll get some of the basic measures, like ‘here we're dealing with finances’ or ‘we're not doing well with testing,’ but their ability to understand data and how it relates to how the school is doing and where it needs to go, there’s a big gap there.”
Cauthen added an explanation. “With boards, you may have individuals with one piece of data that's not anchored to any other data or even put into context. That goes back to data illiteracy. When you're illiterate about data and how it fits in contextually with other data, you are prone to making questionable decisions.”
David Kafitz, the Learning Counsel’s VP of School Relations & Consulting, said, “There is a difference in the perception of education data as it's used and received by boards of education/education laypersons versus the actual educator and education leadership. You can give the same data to both groups and those in education are going to view its value and usefulness one way, and those folks that are not in education every day, policy makers, elected boards, that type of thing, will view its usefulness and its resourcefulness and its meaningfulness in an entirely different way. Some of that I blame on the leadership of the school districts not doing their job well enough to educate the folks that they're giving the data to versus just handing it off and trying to simplify it as much as possible to make it useful or make it meaningful in some way.
The states and the federal government also have a role, trying to boil down all this data to something as simple as a letter grade for a school to show its effectiveness and its work. Things are becoming complex and you can't always just boil them down to some simple random point to make it easy to understand. The same thing is true with the coming of the analytics that we're seeing. It's going to be complex and we're going to have to really step up our efforts to make sure people understand what they're looking at.”
“I've done some research out there about different IT systems that could provide some kind of integration,” said Bosacker. “The really good apps and good tools are expensive. The big companies can afford them, and schools are not in that category, by and large. A lot of the companies have been creating their own middleware. The middleware is where you bring in the information from all the different apps and tools and you've got protocols for integrating them into your database so they're accessible. An expert I spoke with suggested that perhaps a number of schools could go together to create this middleware.
“What integration really comes down to in the coding world is data tables,” said Cauthen. “If you imagine a spreadsheet and each of the little fields in the spreadsheet is its own data piece and then those pieces interrelate in a connected way with the other ones in the same row, in the same column and around the program, you've got the primary building block of any data table. Data tables can have a lot of sophistication or very minimal and when you're going to integrate, you can do an open API from one system to another, but universally or nearly universally, there's no system that's going to allow you an open door to scrape data dynamically, constantly. What happens is you have the ability to see the data and scrape the data, as the outside API that wants the data, the outside system, but it's a one-time scrape and then you don't get to have a backend developer access into the other system to do it all the time. You have to really think about that from the perspective of the commercial interest and why they won't let that happen. They don't want any other programmer who's got access to the keys to the kingdom.
What's needed is that shuttle point in the middle, that's a common data table. That means when you build the API, you have a lot of data table work being done to duplicate into the same format of data as is wanted. It's a bit of a programming thing. The most sophisticated programmers are the ones who understand data tables. It's a very expensive piece of work. That's why that software is so expensive, that's why middleware is just extremely costly. Kevin do you want to comment?”
Kevin Lewis, the project manager for IMS Global's Privacy Taskforce, said “That's definitely not something we are working on at the moment. Standardizing. We've looked at a way to make it easier. We have looked at doing that, but it's very difficult. It's something that if we were talking about it, would be a few years from now before we could actually get it going.
“I'm deeply concerned about this,’ said Cauthen. “If schools are really going to get to personalization, they need to have dynamically synching data, not just batch mode sync. That's too slow and if you've got to solve a kid's problem to keep them up in class, you need it now. You can't wait a year to find out your test result. It is ridiculous.
“The only other thing that I can see happening is from discussions I've had with the companies that are doing full-on glorified, adaptive learning products. Those are different, because when you enter the program, it's doing a lot of the teaching, (That always scares teachers when I say that), but the program itself is pre-assessing, it's plotting the student into their own personalized path. It's asking them questions about material that they're learning and then helping them with more questions if they keep failing that, shunting them back if they need to be given material over again to restudy and then forwarding them on, assessing them and then gating them into the higher levels of the program, just like a game does. That, if it's built out across all subjects and all age levels and does actually allow an interception point for teachers, for kids that just are still stuck, automatically pinging the teacher and saying, ‘Please get into this, this kid needs some help.’ That world is the potential to not actually need integration but so far, it's not fully built out.
I think what might happen is we'll have a divergence between those two worlds. Some people will go full on adaptive curriculum wherever they can, and other people will remain in the document reality. And that's a big concern of mine because the two are expensive in different ways. One is very human intensive, the other one is not as human intensive, but it's certainly very expensive to produce that level of adaptive curriculum. Those are the two worlds we see emerging.
I don't think there are a lot of companies out there who get it yet. How complicated this is for you guys in education. In the rest of the industries, let's say healthcare, they have a couple of major systems to interact with. Regular corporate America has their salesforce system, which is like a student information system and then they've got an HR system and then they're using Microsoft Office and that's it. So, the average corporate worker has three logins. Here in education it's almost unlimited. Every year our survey is showing that the average teacher has 75 logins. If you have 75 third graders in your room and they've all got 75 logins, you're never going to get any teaching and learning done because everybody forgot every login. That's why single sign-on is a popular App this year. But it doesn't solve the problem of the sheer complexity of the knowledge base. And there are a lot of companies that aren't even looking at trying to open up APIs. They just want to stand alone in their own little universe and make everybody have another log in.
If I picture the scale of maturity, I see it as, first devices went in and then a ton of various random apps went in and Google®, Google docs. Mostly. There's still some Microsoft Office® hanging around out there, and then the next level becomes a single sign on and document management level. And then after that people start realizing they need more LMS function so they can customize assessments and the LMS companies are far and away trying to do complete dominance of that world. They want to integrate with everyone. But the companies that they don't integrate with are sort of pushed out.
There are more than 7,000 companies in the field with various digital learning objects. Some of them are full collection websites, some of them are one-off math programs, some of them are other types of Apps like parent/teacher communication apps, coloring apps, games. There will always be companies that are left out of that mix. So, there's an inequity in a different way because maybe the new stuff is better than the old stuff and you're stuck with the old stuff. I think it's an interesting time, but what you guys are saying, we really need to communicate with other people about what the issues are and help them sort of feel their way into a utility of "faster" because other people have been down the road.”
“I completely agree with that,” said Weise. “I think that technology vendors are in a good position to partner with schools to help identify and build solutions that meet the needs of the education sector. The R &D departments of technology manufacturers are actively looking for scenarios, actual education workflows, to better understand the industry pain points. And that's one of my roles at Canon, having come from the industry, understanding the workflows and then helping Canon integrate devices and systems with existing LMS, SIS and others. But we all could do more and better, with the help of organizations such as the Learning Counsel, helping us better understand these needs and congregating the right players. Maybe a set of recommendations will be one outcome of this group, like a list of industry priorities that could be pursued and shared with technology vendors.”
If you would like to join, you’ll find discussions about issues like a “Model Architecture” for schools, including the divisions between the major systems and all the little “Applets,” digital curriculum pieces, and sites. In addition, the group brings up important changes about network and hardware infrastructure related to school and district planning – inclusive of security news for both digital and physical hardening of the school environment. Finally, the road to “prescriptive” analytics for full personalization for students has many stages, and the group provides a community forum for discussion and helpful advice.