Lessons From Atlanta
July 20, 2011 02:17 pm
The cheating scandal that has engulfed the Atlanta public schools (and similar scandals that are bubbling up in Baltimore, Philadelphia, Washington, DC, and elsewhere) ought to serve as a “teachable moment” to point to a path forward in how we think about holding schools and educators accountable for student performance. Unfortunately, the incident has provoked a predictable response: those who object to high-stakes testing have used the case to renew calls to ease up on accountability, while those who favor strong accountability say that all that needs to happen is to tighten up on test security.
Neither of these responses addresses the real issue. Accountability is necessary, and easing up on it will not provide children or schools the help they need. At the same time, simply improving test security is not enough.
Make no mistake: the Atlanta incident is deeply troubling. An investigation launched by the governor found cheating on state tests in forty-four of the fifty-six schools examined, and identified 178 teachers who tampered with tests; eighty-two teachers admitted culpability. According to the investigators’ report, teachers drilled students on items that were nearly identical to those on the test; changed their inflection when reading test questions to signal the correct answer; sat low-scoring students next to high scorers to encourage copying; and changed answers after students had turned in their test booklets. Moreover, the report found, administrators and district officials praised educators in schools that showed dramatic gains in test scores that should have raised eyebrows, and silenced those who raised questions or suggested impropriety.
There is no excuse for any of those actions. Yet the report also found that the district’s accountability system, which set demanding targets for improvements in test scores for each school, created enormous pressure on teachers and principals to raise scores. Of course, not all Atlanta educators responded the same way to those pressures. It’s important to note that many Atlanta educators did not cheat, and many improved teaching and learning in their schools. Nevertheless, some reached for shortcuts and acted improperly.
This could have been predicted, and in fact it was. In a 1999 paper, entitled “When Accountability Knocks, Will Anyone Answer?”, Charles Abelmann and Richard Elmore found that schools differed in their response to external accountability pressures depending on their level of “internal accountability”; that is, whether they had developed a strong sense of collective responsibility for improving student learning. In schools with strong internal accountability, teachers are capable of working together to develop and implement new practices to improve learning, while in those with weak internal accountability-those in which teachers are isolated and believe that they can do little to improve learning-are likely to respond to external accountability pressures in unproductive ways.
The lesson, then, is to put greater attention on building the capacity of all schools. As Elmore has put it, the goal should be “reciprocal accountability”: for every increment of performance I require of you, I have an equal responsibility to provide you with the capacity to produce that performance. Simply setting targets and holding schools accountable for meeting them will not achieve the results states and districts expect.
A second lesson is to get the incentives right. The goal in Atlanta and elsewhere is to hold schools accountable for student learning, and that goal is the right one. But test scores are not learning; they are an indicator of learning. If student learning improves, test scores will improve, but the converse is not necessarily true. Test scores can improve for reasons that have little to do with improved student learning. Focusing on test-taking skills, narrowing the curriculum to what is on the test only, and drilling students to write “brief constructed responses,” not teaching them to write effectively, can all raise test scores, but they do not produce improvements in learning. The same with cheating.
To reduce the incentives to cheat, then, districts and states need to create accountability systems that measure what they want schools to achieve. Test scores will be part of that equation. But other indicators will be as well. By broadening the measurement, systems can make it less likely that educators will manipulate the one indicator that’s measured. And doing so does not weaken accountability, it focuses it where it ought to be focused.
Currently Congress is considering a reauthorization of the Elementary and Secondary Education Act, which will re-examine the accountability system set up by No Child Left Behind. And states are redesigning their accountability systems based on a framework adopted recently by the Council of Chief State School Officers. As these new systems get put into place, state and district leaders ought to pay attention from the lessons of Atlanta.