Kim Dancy of The Georgetown Public Policy Review looks at the problems inherent in a system that uses standardized test results to make high-stakes decisions such as how schools are funded and whether teachers qualify for tenure. When the difference between good test results and bad could mean losing one’s job, a workplace transfer or even school closure, it is almost inevitable that the motivation to raise the scores via artificial means – cheating – becomes too tempting for some administrators and instructors to resist.
And so it proved. The first major standardized-test related cheating scandal concerned a round of testing in Atlanta public schools in 2009. Similar allegations about several other districts, from Los Angeles to Brooklyn rapidly surfaced. In some cases, it wasn’t even the cheating that made headlines, but the horrible pressure applied to those who refused to go along. As lurid as these stories are, they point to a pervasive problem that is growing and spreading in K-12 education.
Among numerous problems in standardized testing, oversight remains a serious issue. The decentralized nature of the education system and the lack of manpower at the federal level make it inappropriate, if not impossible, for the Department of Education to take on monitoring responsibilities. And while many State Educational Agencies do conduct on site audits of test administration, they often lack the necessary resources to do so in every individual school district.
According to Dancy, this places the oversight firmly in the hands of the local officials and district administrators — in other words, the very people who are most at risk of the test scores fall below expectation. Furthermore, not all districts place the same amount of focus on testing integrity, which means that depending on under whose auspices a particular school might fall could determine how “trustworthy” their test results are. Such variability makes statewide, much less nationwide, comparisons a farce.
Even in the cases where oversight is sufficient, determining when an incident of cheating has occurred is challenging.
Some practices are clear-cut: changing student answers is undeniably dishonest no matter what the circumstance. Still, gray areas persist. Is it cheating to provide definitions of a word like “denominator” during a math exam? To allow a student a few extra minutes at the end of the testing period? To tailor this year’s instruction to last year’s exams? These answers depend to a large degree on the policies set forth by a particular jurisdiction regarding appropriate testing behaviors, but consistency within a state or locality is crucial. Ambiguities such as these diminish the comparability of scores across students, classrooms, and districts—in this way, seemingly harmless behavior may undermine the reliability of test results.
Even if the issues of oversight are solved, says Dancy, and a consistent policy on what constitutes cheating is developed and implemented, as long as the system with the perverse incentives remains in place, cheating will continue to be a major issue. As long as high-stakes decisions will rely on test scores, instructors will either find a way to cheat or will be pressured to cheat by their superiors.