Friday, July 8, 2011

A misplaced trust in leadership

There is a common presumption that when it comes to improving public school performance, constructive changes, whether originating at the top or the bottom, cannot take hold unless they are championed by the superintendent. The superintendent of the Atlanta Public School District is often cited as an example of the type of leadership that brings about dramatic improvement. Dr. Beverly Hall had one of the longest tenures of any large urban district superintendent, having led APS over 12 years. She won many national awards and accolades over those years and the Atlanta schools were on a steep trajectory of improvement.

Now, the Atlanta Journal-Constitution reports that trajectory was launched not by the vision of a strong leader, but was systematically manufactured by a superintendent who tolerated, and possibly encouraged, outright cheating by school staff on state standardized tests. The evidence, gathered by a governor's task force, is overwhelming, alarming, and shocking. The task force report names 178 educators, including 38 principals, as participants in cheating. More than 80 educators confessed to actions such as erasing students' wrong answers and changing them to correct answers. Cheating was confirmed in 44 of the 56 schools examined.

The consequences are yet unknown—Will Dr. Hall have to pay back performance bonuses earned by improved test scores? Will colleges stand behind acceptance decisions for which good test scores were a factor? Will the federal government hold the district accountable under the No Child Left Behind Act? Will criminal charges be brought? One clear result: Students who never learned their real scores lost opportunities to improve their weaknesses and build on their strengths.

Like the recently-exposed problems with the New York state exams, this cheating scandal raises questions about the wisdom of building education reforms around standardized tests. Some will use these developments as a reason to oppose merit pay for teachers, strengthening No Child Left Behind, or requiring private schools accepting voucher students to participate in the tests. Others will argue that a few bad actors cannot be allowed to spoil the best, albeit imperfect, method of monitoring school performance.

These opposing views each have a foothold in Wisconsin, which is currently in the process of developing a new standardized testing scheme. The public debate over the ways in which the new test is to be used for policymaking and accountability purposes is sure to be long and hard. But this debate gives rise to an opportunity to make Badgerland lemonade from other states’ lemons. Wisconsin could anticipate the potential for organized cheating on standardized tests and explicitly spell out the ways in which the Department of Public Instruction will monitor test security and integrity, and the penalties for cheating. For example, DPI might rule that any school turning in score sheets with wrong-to-right answer erasures numbering more than three standard deviations above the norm (statistically unlikely to have happened coincidentally) would trigger an automatic investigation.

Currently, according to DPI's policy and procedure manual, penalties for school staff found to be complicit in cheating are not regulated by the state; it is left to a district to administer any penalties according to local policy. The New York Times reports that in Atlanta, the school board was loath to criticize Dr. Hall, whom the directors felt had done much to improve the perception of their schools and city. Perhaps the lesson of Atlanta is that districts cannot be allowed to self-police as long as their test results are used to determine more than just student performance.

No comments: