Texas Tackles the Data Problem

Originally appeared in Education Next

Commentary | | November 16, 2010
Data-Driven Improvement

Terry Driscoll, executive director of information systems at Lubbock Independent School District, says he’s hardwired to resist government intrusion. And when the Texas Education Agency (TEA), along with the Michael & Susan Dell Foundation, came to town to talk about improvements to the state’s data system, he first wondered whether they even knew what they were talking about. But with the recession taking a bite out of his district’s own data initiatives, Driscoll was ready to listen. Now, almost a year later, Lubbock has become the first test site for a different type of state data system, one that aims to move districts from collecting data solely for accountability to collecting it to improve schools.

The darling of reformers, data have clear potential to help educators make better decisions. But however much they are touted, most data initiatives remain far from realizing their potential. Historically, the collection of data has been top-down, designed almost exclusively to show compliance with state and federal regulations. And while the amount of data collected continues to grow—Texas school districts respond to 104 data collections by the state each year, costing the districts in excess of $300 million—their quality and usefulness are questionable. Thus many state data systems function as de facto data morgues, used more often in autopsies of failed programs than to help educators and policymakers improve existing ones.

It is perhaps not surprising, then, that while Texas is data-rich, it is still information-poor. A 2008 TEA study found it likely that some state data are erroneous, even if the same data are accurate in the district systems. It also found that districts must constantly reformat their data to meet state requirements, adding to the cost and to the opportunity for introducing errors.

More important, once districts submit data, they receive little if anything of instructional value in return. Much of the information the state collects, such as the number of 7th graders eligible for Title I funds at a particular school, governs the flow of dollars, but it is not on its own useful for improving school operations or performance. Other data, such as Lubbock’s results on state assessments, could be useful. But that information arrives at the district office late each summer on computer disks, and it must be integrated with the district’s own system for storing student information, along with a third system that houses interim assessment results. By the time school personnel are able to compile reports for teachers, the information is “already cold,” says Kelly Trlica, Lubbock’s chief academic officer.

Because of experiences like Lubbock’s, when it came time to update the state’s 25-year-old data system, officials decided to make some big changes. Instead of gathering a group of technicians in Austin, state education officials talked to 2,200 educators and administrators across the state about the data they needed. Overwhelmingly, they said that the information had to be directly accessible to and relevant for educators. Middle-school teachers, for example, need access to special education identifications, test results, and other information to create appropriate instructional groupings and interventions. And they need that information well before school starts. Principals, for their part, want data to evaluate the many instructional software and intervention programs that are purchased each year. Moreover, frequent educator use is an important means of preventing, or catching and correcting, data errors. If those people closest to the data—teachers—are actually using the data, they will update class rosters and other student information on a regular basis.

To enable schools, districts, and state officials to more easily share and use data, the TEA is developing a more flexible information-system platform. The platform will offer smaller districts a shared, state-sponsored student-information system. It will also make it easy for districts with existing systems to connect to a new data platform that will serve as the hub for district-specific data, feeding relevant student, classroom, and campus information directly to educators and enabling seamless reporting of compliance data to the state. For example, the district might enter attendance data just once. That information would then be available to teachers and counselors, in real time and in dashboard formats, where it would flag students with potential problems. That same attendance data would be automatically reformatted for easy transmission to the state. If successful, the new system will not only reduce costs and streamline the existing accountability process, but will also equip educators with relevant information they can use to help their students.

Building a student-centric system that serves the diverse needs of the state’s 1,235 local education agencies, which range from districts with fewer than 500 students to those with more than 150,000, will not be easy. But this year, educators in Lubbock’s five high schools are getting a start. They will be the first to test-drive the early-warning dashboard, a tool that provides easy access to student attendance, assessment, and credit attainment records, as well as other data that will allow educators to quickly identify potential dropouts and get them back on track.

More than 160 educators are serving as advisors to ensure the system delivers what teachers need. And, true to the spirit of TEA’s new open approach, the public can read the specifications, make recommendations, and watch the system unfold at www.texasstudentdatasystem.org.

More in Data-Driven Improvement