Statewide Longitudinal Data Systems (Information Only)
March 7, 2019
Good morning, and thank you Chair Roth and members of the committee for the opportunity to speak today. My name is Colleen Moore and I’m the assistant director of the Education Insights Center at Sacramento State. I was asked to make comments today based on some research I did over two years examining California’s approach to tracking student progress across K-12 and higher education, and its needs going forward.
I’d like to start with the positive aspects of what I learned. While there is no statewide cross-sector data system in California, the four public education systems have longitudinal student data systems that are quite good, including the CALPADS system at CDE and data systems at each higher education system office. They contain information that is valuable for understanding student progress and outcomes within each system, with data that are updated annually in some cases, and at specified points during each term in other cases. Some information is brought into these data systems from other state agencies, including data related to financial aid from the California Student Aid Commission, and information on employment and wages from EDD. While the data in these systems are primarily used for compliance, including meeting state and federal reporting requirements, each system office has increased its efforts in recent years to make more information available through online data dashboards and other reports. They sometimes share student data among themselves to address particular needs, as you heard them describe.
So what’s wrong with having four good longitudinal data systems, but no cross-sector data system? I’ll point to three issues:
- The information we have about student progress and outcomes is disconnected. While there’s some data sharing at the system level, as I noted, and there are some voluntary data-sharing efforts across institutions in some regions, we have no way to get a comprehensive statewide picture of the education to workforce pipeline in California. From a policy and budget perspective, this reinforces California’s very siloed approach to education planning.
- The information is largely inaccessible. While each system provides some good information, it is often compliance-focused or framed in ways that suit the needs and interests of the systems. Similar information is often calculated and presented very differently across the systems, making it difficult for policymakers, educators, students and their families to interpret.
- The current approach is inefficient. As one example, we have multiple agencies and individual institutions contracting with the National Student Clearinghouse in an effort to try to find out where students go when they leave a particular institution – a boon for the Clearinghouse, but a real failure to leverage state resources.
Having good information that’s disconnected and inaccessible, with inefficient means of putting the pieces together, leaves us unable to answer critical questions about student progress and outcomes across systems. It makes it impossible to develop evidence-based policies and practices that will best serve students as they move through multiple education systems.
As to the challenges we face in developing a data system, my research suggests that they are largely political and cultural. The fact that California is coming rather late to the party in terms of developing a statewide data system is helpful in regard to the technical challenges. Most other states have or are in the process of developing such data systems, including large states with diverse populations and complex education systems, giving us examples to learn from. These states have faced issues like matching records when K-12 and postsecondary systems use different student ID numbers, and developing mechanisms to maintain data privacy and security. There are reports to read and experts to consult to understand and apply the lessons to our own context.
The bigger challenges are in deciding what basic type of data system to develop and identifying appropriate leadership for such an effort.
The bigger challenges are in deciding what basic type of data system to develop and identifying appropriate leadership for such an effort. Do we pursue a traditional Statewide Longitudinal Data System that involves connecting the records across our existing data systems? This is the approach used in other states, and has the advantage of leveraging California’s existing investments and requiring a short timeline to develop at reasonable cost. Some argue that, instead, California should try to develop a data system that involves more frequent uploads of information by individual schools and colleges, one that can be used to provide services to individual students as they transition across institutions. While there are promising local/regional examples of such systems, there are significant concerns in considering them as the model for the initial development of a statewide system, in terms of technical barriers, costs, and an extensive timeframe before comprehensive statewide data would be available.
If California pursues, at least as an initial approach, a more traditional statewide data system, the primary challenge involves deciding on a governance structure. While other states use a variety of approaches, my research concluded that California might best be served by establishing a state data agency, or a data office within an existing agency, with the specific mission of developing a centralized data warehouse, producing standard reports and data tools for various audiences, and managing access to the data for research. This structure would ensure that there’s an entity responsible for getting this done, in collaboration with the segments of course, but a neutral entity with clear accountability not tied to a specific segment. The structure would also support expanding the data system to include data from private postsecondary, health, social services or other agencies outside the public education systems.
For decades, California has been sitting on education data with no publicly available information on the outcomes of its investments across public education. The state is well-positioned to join many other states in using existing student information to create a comprehensive statewide data system that can inform student decision-making, support schools and colleges in their efforts to improve student outcomes, and meet the Governor’s vision of using data to improve and align state education and workforce policies. We should seize that opportunity. Thank you, and I’m happy to answer questions.