Developing database interoperability: the example of the "California Longitudinal Pupil Achievement System (CALPADS)"

As Thad and I argued recently in our study of interoperability and data exchange standards for election administration, we think this is indeed the “next big challenge” for election officials. But this morning I learned of an interesting analogy to the problem of developing data exchange standards, and working towards interoperable information management systems: the California Longitudinal Pupil Achievement Data System (CALPADS). Much to my surprise, it turns out that educators in California don’t currently have a fully-functional database of all public school students, which means that keeping track of all the public school students, not to mention their progress through the public school system, has been difficult.

On February 5th, the state superintendent of education announced that the CALPADS system had hit an important milestone, whereby they had data from 93 percent of the state’s school systems. Here is the media release:

SACRAMENTO — State Superintendent of Public Instruction Jack O’Connell announced today that 93 percent of school districts in the state have submitted important student data in preparation for the launch of a statewide system to be used for calculating enrollment and long-awaited, more accurate graduation and dropout rates. The California Department of Education is nearing completion of collecting this baseline data for the California Longitudinal Pupil Achievement Data System (CALPADS).

“CALPADS will greatly improve our state’s ability to monitor student achievement over time and ensure that all students receive the educational services they need when they need it,” O’Connell said. “We need more accurate enrollment, graduation, and dropout data in order to accurately represent the progress and achievement of student subgroups. It is absolutely critical that we receive this student data from every single school district, county office, and charter school in California. I am pleased that the vast majority of districts and charter schools have contributed to this important effort.”

This year, for the first time, detailed enrollment data has been collected from school districts, county offices, and charter schools from an annual maintenance submission on their statewide student identifiers (SSIDs). Each local educational agency is required to certify with the California School Information Services (CSIS) Program an Annual SSID Maintenance, which is a specific set of student-level data for students enrolled within its jurisdiction. SSID accuracy is critical since SSIDs are now required on all statewide assessments and will be used for enrollment, graduation rates, dropout rates, and accountability-reporting. By collecting this data annually, CSIS is also able to compare the SSID data statewide and identify and correct invalid SSIDs.

The due date for reporting on SSIDs was January 19, 2007. As of today, the state has received SSID records that represent 93 percent of California’s local education agencies and 98 percent of its student enrollment. O’Connell called on districts that have not yet submitted this crucial information to immediately focus on reporting this data.

“Submitting student enrollment records is the first step toward having better data and we must not fall short,” O’Connell said. “I do recognize the amount of work involved to submit this annual enrollment update and the ongoing work to maintain student data. I have and will continue to strongly advocate for more funding to support these important activities. Nevertheless, it is still absolutely critical that the data be submitted this year.”

O’Connell has asked Governor Schwarzenegger and the Legislature to provide $32 million – $5 per student – in the state budget to support collection of quality data at every school.

The trick, in making this happen, has been the development of unique student identifiers, as part of the “Statewide Student Identifier Project.” I’ve only today started to look into the information on how this project has been set up and how it operates; there is information on the development of the unique student identifiers at this website.

In many ways, this does look to be a good example to study closely, not just in California but in other states, to find lessons and best practices for the development of fully interoperable election administration datasystems, in particular, interoperable voter registration systems.