Frank Chien, CEO of LearnSprout, shares insight into the company’s methods around data and data management at schools and districts.
SC: LearnSprout helps schools manage and transfer data through automated APIs. How is this filling a void currently found in school and district data management?
FC: Background – The student information system (SIS) is the system of record for schools and districts. Examples include PowerSchool, Skyward and Infinite Campus. It is where a student record is born. It is where school and student schedules come from. It’s where class rosters are maintained. It’s where attendance and grades are stored. It’s used to produce transcripts and state reports. It’s how schools get money. Nearly all K-12 software relies on data from the SIS: LMS, Assessment, Lunch Management, Transportation, SPED, etc.
Currently, in order for schools to manage accounts between systems or to move class roster or demographic data from the SIS to dependent systems, one of the following methods is typically used:
- Manual export of data to a CSV file. Manual normalization of that data to make it ready for import. Manual upload to the dependent system (e.g. LMS, assessment system, etc.)
- Manual export of data to a CSV file. Send file via email, or upload to a Dropbox account (not FERPA compliant). Vendor manually uploads data to their system.
- Scripted export process that pulls a nightly data dump from the SIS, then sends via FTP to vendor who uploads into the dependent system.
- Schools Interoperability Framework. (SIF) Automatically moves updates to data between systems. However, this has proven to be cost prohibitive for most vendors and schools.
The vast majority of schools and districts use a patchwork of the first three methods. Some of the more tech savvy and larger districts also incorporate SIF, but lack of SIF compatibility by many edtech products require that even the most technologically advanced districts maintain a kludge of methods to integrate systems together.
The result is siloed and asynchronous data. Teachers will see new students in one system but not the other, and transferred out students persist. Students can log into one system, but not the other. Time is wasted trying to get everyone logged in. IT staff spend countless hours performing manual exports and imports. Scripts are prone to malfunction and must be manually maintained. It’s a big mess, but it’s existed for so long in the same abysmal state that most have come to accept it. (like the DMV). For most dependent systems, the data that a teacher sees will be anywhere from 24 hours to weeks old.
SC: What information should a teacher have about a student on the first day of school?
FC: Before the first day of school, teachers should be able to see the most recent class roster with basic demographic information in all systems. Students should be able to log in to all systems with their existing account and see their current class schedule. Put simply, teachers and students should not have to jump through a bunch of hoops just to get to square one with new technology.
SC: Is there any data that you would NOT want stored or managed in LearnSprout?
FC: Right now we’re focused on our core strength which is pulling data from the various flavors of student information systems, but we are being pushed by our customers to consider tapping into other systems so that ALL of a students data can reside in one location.
SC: What types of data are you finding pivotal to student success?
FC: Some of the most important, and most overlooked student performance and engagement data, including: Attendance, grades, assignment scores, missing assignments, discipline, GPAs, participation in activities. Also, the number of parent and student logins to the SIS can be very useful. We’re also very interested in looking at immunizations and what the correlation is with other data like attendance, grades and discipline. (This would be abstracted, anonymous data to make sure we’re square with HIPAA) Together, this makes up most of the leading indicators for identifying at-risk students. There is such a strong focus on assessment data right now, that these fundamentals are being overlooked by many in the school reform movement.
SC: In what ways will data better equip teachers to improve student success and performance?
FC: The cruel irony here is that when LearnSprout is being utilized across a district, it will be largely invisible and taken for granted by most. Teachers will be able to use an LMS or grade book of their choice, and they’ll be able to log in and see their schedules and rosters without having to ask kids to find them or emailing keys to access a class.
So, data on it’s own is of little use. This is the current state of affairs as most schools have little or no access to the data that lives within their SIS. We plan to change that with our upcoming Dashboard tool which will provide schools with cost-free analytics that can be quickly and easily plugged into their existing SIS. More about Dashboard: http://learnsprout.com/products/dashboard and http://tmblr.co/Z0doCwYIyfSs
SC: Do you know examples of schools that are leading the way with data?
FC: We’ve been very impressed by what Rocketship Education is doing with data. I have yet to find a system as data driven as the folks at RSed.
SC: Is there a pitfall to too much data? What challenges do you see with data management, usage, etc.?
FC: Most believe that we already have too much data, and the fact that it’s so siloed and inaccessible exacerbates this notion. Bringing data together from across the various systems allows for a new generation of edtech developers to build new apps without running into a barrier to entry that keeps their new software on the sidelines. Bringing data together also makes it possible to finally perform true “Big Data” analysis where we can begin to actually see the future by analyzing the past, and make adjustments before students become at risk.