D.C. cheating issue calls test-driven incentives into question
Apr 21, 2013 | By Charles Taylor Kerchner
The smoke surrounding allegations of test score cheating in the Washington, D.C., public schools burst into flame last week. In a 4,300-word blog post, titled Michelle Rhee’s Reign of Error, the veteran educational journalist John Merrow linked the former schools chancellor with documents that suggest that she knew about widespread cheating on standardized tests and looked the other way.
Merrow drew parallels between Washington and Atlanta, where former superintendent Beverly Hall and 34 others have been indicted. But the underlying question is whether school reform can successfully be driven by rewards and punishments tied to standardized tests.
First, the Watergate question: What did Rhee know and when did she know it? As U.S.A. Today reported, District of Columbia Public Schools officials have long maintained that a 2011 test-cheating scandal that generated two government probes was limited to one elementary school. Merrow, however, reported that a long-buried memo from an outside data consultant warned as far back as January 2009 that educator cheating on 2008 standardized tests could have been widespread, with 191 teachers in 70 schools “implicated in possible testing infractions.”
Consultant Fay G. Sanford noted substantial numbers of erasures on test forms where corrections were made from wrong answers to correct ones. (Erasers leave smudge marks that the machines doing the grading recognize along with the answers.) “It is common knowledge in the high-stakes testing community that one of the easiest ways for teachers to artificially inflate student test scores is to erase student wrong responses to multiple choice questions and recode them as correct,” Sanford wrote.
When asked, Rhee said that she didn’t recall getting a memo from Sanford, an assertion she repeated this week to the Los Angeles Times editorial board.
A day after the Merrow report was issued, another investigation was released citing “critical” violations of test security in 18 classrooms located in seven district schools and four charters. The test was given in more than 2,600 classrooms, and current D.C. schools chancellor Kaya Henderson said in a statement, “We are pleased that this is yet another investigation that confirms that there is no widespread cheating at DCPS.”
But Henderson is not quelling demands for a district-wide investigation. Randi Weingarten, president of the American Federation of Teachers, has called for an independent investigation, as has Washington Post writer Valerie Strauss, who said, “The memo does not offer conclusive evidence that cheating occurred, but it literally begs for a thorough probe to be conducted—this time by investigators with subpoena powers.”
Meanwhile, an in-depth evaluation of D.C. schools by a National Research Council panel continues. The study is chaired by two Californians: Carl Cohn, a professor at Claremont Graduate University and member of the State Board of Education, and Lorraine McDonnell, a professor of political science at U.C. Santa Barbara. Their work is not confined to test tampering, but considers the overall effect of governance and management changes.
Test tampering only the symptom
Education policy organizations that have generally supported Rhee were quick to minimize the test tampering as “bad judgment.” Michelle Gininger at the Thomas B. Fordham Institute cautioned that: “Critics are bound to say that all reforms that Rhee stands for — teacher evaluations, tenure reform, and school choice — should be dismissed. But let’s not abandon education reform and accountability at the expense of our students, which getting rid of testing surely would.”
But, along with her fear-based management style, reliance on test scores is the Achilles’ heel of Rhee’s reform policies. Accurate test scores form the basis for her performance-based pay system and means to determine which teachers and administrators should be kept or fired.
Unfortunately, the system Rhee advocates inherently motivates people to cheat. As Cornell Law School professor Lynn Stout wrote in the Los Angeles Times, “It’s almost impossible to design objective performance metrics that can’t be met through illegal or undesirable behavior…. And when you create a system that inadvertently incentivizes illegal or undesirable behavior, you get more of it.”
This is the lesson of organizational history, not an isolated “bad judgment” aberration. It’s about more than school test scores in the District of Columbia, Atlanta, Texas or even Rhee’s possibly outsized claims of how well her students did during the three years she taught school in Baltimore. The policies Rhee endorses create bad incentives. Bad incentives lead to disastrous results. They certainly played a part in the largest business collapses in recent history: Enron, WorldCom, Lehman Brothers and the collapse of the subprime mortgage market.
The antidote is not to abandon incentives tied to test scores and other measures of achievement, but to use them in ways that actually motivate teachers, and more importantly students, who are the real workers in the education system.
What motivates teachers most? Student success: If an organizational system of curriculum, pedagogy, professional training and school organization helps students experience success, then teachers are highly motivated. Teachers are motivated by being part of a winning team, a school that does well at its own mission, which most often is not test score maximization. Teachers are motivated by being part of an occupation that is honored and trusted. These are the lessons from a century of study.
How should we then use achievement data? First, build feedback into the lesson and project level of schooling rather than waiting for end-of-year tests. Second, use tests that actually measure student performance on substantive tasks. In other words, build data feedback into the everyday life of teaching and learning.
What policies use data effectively?
- Policies that build time into the school day for teachers to work together to interpret student results and self-correct;
- Policies that invest in smart software that adapts to student performance and build computer routines that augment teacher effort;
- Policies that build capacity into the education system.
This real road to student success is substantively at odds with Rhee’s policies. The difference is no small matter. Rhee’s StudentsFirst organization is attempting to create a political counterforce to the existing institution of public education and most current ideas about improving student achievement. In the words of Rhee’s recent book, its ideas are, indeed, Radical. Although it refers to itself as a nonprofit, giving StudentsFirst the public gloss of a thoughtful policy institute, it operates as a political organization. The Internal Revenue Service classifies it as a 501(c)4 organization, which means that contributions to it are not deductible on individual or business tax returns.
StudentsFirst is gathering supporters. Mayor Antonio Villaraigosa and former state senator Gloria Romero have signed on, as has Nevada Governor Brian Sandoval and Rhee’s mentor, former New York schools chancellor Joel Klein. Bill Cosby and Connie Chung are among its celebrity board members. Subsidiary organizations have been founded in several states.
StudentsFirst has started grading states. California received an F for its educational policies. (This same ranking grades Louisiana a top-of-the-pack A, and gives Massachusetts, which leads the world on some international measures, a D+.)
StudentsFirst also poured $250,000 into the recent Los Angeles Unified School District races, and supported candidates in Burbank and suburban Sacramento.
More than the question of test score cheating, it’s the logic and practice of Rhee’s policies and politics that deserve examination. To use the descriptors Rhee favors, they just may be “crappy” and “suck.”
• • •
Charles Taylor Kerchner is Research Professor in the School of Educational Studies at Claremont Graduate University, and a specialist in educational organizations, educational policy and teachers unions. In 2008, he and his colleagues completed a four-year study of education reform of the Los Angeles Unified School District.