Data, Data Everywhere, But Not a Drop to Use
Just as the weary seamen in “The Rime of the Ancient Mariner” were surrounded by a vast ocean of water yet had none to drink, educators are drowning in seas of data they cannot use.
Over the past four years, the U.S. Department of Education’s Institute of Education Sciences has awarded $264 million to 41 states and the District of Columbia for statewide longitudinal-data systems. Foundations have contributed millions of philanthropic dollars to ensure such systems’ development. And in response to the federal No Child Left Behind Act, districts have created entire departments whose sole purpose is to respond to request-for-compliance data.
Are we there yet? Do we really need more educational data? Apparently so, according to the new administration. Through the American Recovery and Reinvestment Act, another $250 million has been allocated in economic-stimulus funds for the development of state longitudinal-data systems.
The Education Department itself apparently questions the value of these state systems. A study on classroom data utilization released recently by the department found that “the vast majority of the activity around data systems and data use in districts and schools involves district, not state data systems. … [T]he hoped-for efficiencies to be gained from integrating data systems at the state, district, and school levels are not apparent from the vantage point of schools and districts.”
The American Productivity and Quality Center’s recently released findings from “Best Practices in Data-Driven Decision Making,” a benchmarking study sponsored by the Data Quality Campaign, revealed the factors that support and serve as barriers to data-driven decisionmaking. Study participants, which represented 69 districts ranging in size from 3,000 students to a million, and included five of the nation’s largest school systems, overwhelmingly agreed that it is not the lack of data that hinders data-driven decisionmaking.
To the contrary, officials in these districts described “being overwhelmed by the sheer number of data collects from the state for compliance reporting.” One participating superintendent reported that his district “has 83 state data collects a year,” and that only 2 percent of the data can be used for analyzing and improving student achievement.
It’s hard to argue with the value of longitudinal-data systems. This is especially true when one of the objectives of the federal stimulus plan is to provide much-needed data for linking K-12 and higher education information to workforce skills and training. But in an economic environment in which newspapers are filled with stories of impending teacher layoffs, proposals for four-day school weeks, and districts that are selling commercial ad space on the sides of school buses to raise cash, why would the Obama administration rank state longitudinal-data systems as one of its top priorities?
Maybe it’s because the view from 50,000 feet is always much prettier than the view at ground zero. Isolated academics, policymakers, foundations, and nonprofits may have great ideas for developing sophisticated data systems that will “help” educators. But the realities of attempting to use data for decisionmaking are much messier and more complex than they appear to be. Challenges such as slow state cycle time for turning around summative test data, lack of software interoperability, and insufficient time for teachers to discuss the data overwhelm the best of intentions.
The APQC’s benchmarking study included national focus groups of teachers and district administrators who explored barriers to data use. In one session, as administrators began to discuss a proposed bond issue to purchase newer, more sophisticated software systems, the teachers in the room began to show signs of growing frustration. After patiently listening to the conversation, one teacher blurted out, “We are just beginning to learn to use what we have now; please don’t buy a new one!”
Does the evidence indicate that state longitudinal-data systems would not have a positive impact on student success? Absolutely not. The studies cited, plus the APQC’s 30 years of experience in the field, show clearly that data systems designed with utilization in mind can create breakthrough improvement in student achievement. This means that data systems must be part of larger improvement systems to be effective.
What do I mean by an “improvement system”? It is a framework for managing an organization that ensures the organization gets the results it wants. And developing improvement systems is hard work. They require clarity of organizational purpose, definition of and focus on customers, identification and mapping of important organizational processes, development and tracking of key performance measures and metrics, benchmarking of internal and external best practices, use of tools for process improvement, and knowledge management. It is the improvement system that provides the mechanisms for utilizing the data.
A handful of successful districts nationwide have implemented data-centric improvement systems. Many were best-practice partners in our benchmarking study, including the following exemplars:
North Carolina’s Iredell-Statesville school district, the 2008 winner of the Malcolm Baldrige award in education, attributes much of its success to a deeply ingrained culture of data inquiry. All staff members know that you don’t walk into the superintendent’s office or go to any meeting unless you have the facts and figures to support whatever you came to discuss. In Georgia, the Gwinnett County public schools, another best-practice partner, has identified data owners who participate in cross-functional action teams to constantly review data validity. And at the Western Heights district in Oklahoma, the superintendent reviews multiple data “dashboards” at the start of each school day. By using these computer-generated compilations of pertinent data, the superintendent can give staff, community, and board members accurate, real-time data on just about any metric that has an impact on student achievement.
I propose that we not add a promising initiative like state longitudinal-data systems to the ever-growing pile of stagnant, random acts of educational reform. As a part of the Obama administration’s plans to help educators, we should create something that will actually affect the daily challenges of teaching students and running a school district.
As a part of, or in addition to, the money that has been allocated for state longitudinal-data systems, the administration should provide dollars for building improvement systems. Data systems designed with usage in mind can help build the internal capacity schools and districts need to survive the upheavals of economic crisis, ever-changing political platforms, and the rapidly revolving door of superintendents.
There are no silver bullets for improving the mess we have made of our nation’s education system. But as Clayton Wilcox, a former superintendent of Florida’s Pinellas school district, says, there are many promising “silver BBs” for education reform. The development and implementation of data-driven improvement systems is one shining BB that has great potential for not just improving, but transforming, the U.S. education system.
Vol. 28, Issue 33