An astonishing amount of cheating is taking place on the tests that measure progress under the federal No Child Left Behind Act. And the cheating I’m referring to isn’t coming from the kids.
It’s rare for a week to go by lately without encountering one or more news reports describing how educators have been caught cheating on their states’ federally mandated tests. Much of this reported cheating stems from the actions of school or district administrators. But a number of classroom teachers have also been identified as benders and breakers of the No Child Left Behind tests’ rules.
— Bob Dahm
The specifics of educator cheating vary from setting to setting. In some instances, we see teachers allowing students more than the allotted testing time. Even worse, some reports have teachers supplying students with heavy-handed hints about which answers are correct during a test’s administration. We also see test-preparation sessions featuring actual test items, copied surreptitiously from a supposedly secure form of the exam. The most egregious instances of cheating arise when administrators conspiratorially erase incorrect responses on students’ answer sheets, then substitute correct responses—sometimes supplying literally thousands of such “improved” responses.
The underlying reasons for this reprehensible conduct are all too obvious. If a school’s students don’t score high enough on their state’s No Child Left Behind tests, the school can (1) be placed on a sanction-laden improvement track capable of “improving” it into nonexistence, (2) be publicly embarrassed in local newspapers, or (3) be slammed simultaneously with both of these consequences. The adverse outcomes are basically the same for districts. Yet it’s possible to dodge all these calamities. Sufficiently improved No Child Left Behind test scores can save the day! Is there any mystery, then, about what has prompted the recent upswing in educators’ cheating?
But understanding why a person does something wrong doesn’t render the wrongdoing right. When educators cheat on tests (whether or not they’re caught), it sends a wretched message to students—that it’s OK for grown-ups to cheat. This kind of conduct should never be modeled in our schools.
What about the world outside school? When citizens learn that certain educators have been trying to pull a fast one with their students’ No Child Left Behind test scores, the discovery is apt to trigger an even greater erosion of confidence in public schools than currently exists. And let’s face it, our public education system is not regarded these days as one of America’s crowning accomplishments. Cheating by educators on these high-stakes tests has a host of consequences, all of them bad.
Cheaters are made, not born. No genetic marker will be discovered predestining a particular person to engage in wholesale hoodwinking. B.F. Skinner had it right on this point: People become cheaters because they’ve been reinforced for cheating. If there’s no need to cheat, that is, if there’s no payoff for people’s cheating, then the problem can be quickly extinguished. How, then, do we halt today’s expanding epidemic of educator cheating on No Child Left Behind tests?
Most states have, in recent years, installed regulations designed to minimize cheating on state-administered accountability tests. Moreover, in most instances, these prohibitions have been widely promulgated. Still, many of the cheating occurrences we read about have taken place in states where carefully crafted anti-cheating regulations exist. So another line of defense has emerged: special-focus firms whose primary mission is to assist states and school districts in minimizing cheating—whether from students or adults.
I have no problem with either of these cheating-reduction strategies. Both will surely help to some extent. But the most effective way of grappling with the problem is to tackle its real source: the tests themselves. This is where we should be putting our energy, because today’s educator cheating on federally mandated tests is, without question, being spawned by those very same tests. If the state tests being used to evaluate schooling under the No Child Left Behind law did not require cheating for educators to be successful, then why would educators need to cheat?
Almost all states’ No Child Left Behind tests are instructionally insensitive: They are incapable of detecting instructional quality, good or bad. The vast majority of teachers want to be successful. They want to be successful by teaching kids what they need to learn. When those well-taught kids then score high on the tests, there will be credible evidence attesting to the teachers’ success. But if these tests can’t distinguish between successfully and unsuccessfully taught students, and if scores on the tests are the trump card that determines a teacher’s prowess, then why should teachers strive to do a super instructional job?
Many teachers, of course, want to do a super job of teaching their students, regardless of accountability tests. But too many others, unable to improve students’ scores on tests that won’t reflect such improvements, may scurry in frustration toward fraud.
Most states’ tests are instructionally insensitive because they were created using a traditional psychometric strategy that, in the end, makes students’ socioeconomic status the most influential factor in determining which students get high or low scores. Thus, if most of a school’s students come from affluent families, we can be pretty sure the school’s students will perform well on the state’s No Child Left Behind exams. Conversely, a school serving students from poor families will rarely be able to promote sufficient improvement in its students’ test scores. This socioeconomic-status-based instructional insensitivity is particularly prevalent when the tests have been built according to a traditional psychometric approach aimed at providing comparative (norm-referenced) interpretations regarding test-takers’ achievement levels.
Even if a state has created customized, “standards based” tests in an effort to better measure students’ mastery of the state’s content standards, or curricular aims, the tests still can turn out to be instructionally insensitive. This is because a state’s curricular aspirations usually include far too many aims, and there are too few test items to accurately measure students’ mastery of those sprawling arrays of content standards. Teachers can’t really target their instruction sensibly or use students’ test results in a meaningful, diagnostic manner. In time, then, even these standards-based tests turn out to be instructionally insensitive, because they become too closely linked to students’ socioeconomic status. Instructionally insensitive No Child Left Behind tests create the circumstances in which cheating becomes attractive to certain educators.
What we must do is make sure that a state’s No Child Left Behind tests are instructionally sensitive: that they are capable of accurately ascertaining instructional quality. Instructionally sensitive tests need to possess three essential attributes:
First, to avoid overwhelming a state’s teachers with too many curricular targets, such tests need to measure students’ mastery of only a modest number of genuinely significant curricular aims—so significant, in fact, that those curricular aims subsume a number of subskills and bodies of knowledge.
Second, each of the test’s assessed curricular aims must be concisely described, with sufficient clarity that the state’s teachers can aim their instruction toward students’ mastery of these goals, and not toward students’ correctly answering a particular set of test items.
Instructionally insensitive tests create the circumstances in which test cheating becomes attractive to certain educators.
Finally, there must be enough items on the test measuring every assessed curricular aim, so that students’ mastery of each aim can be measured with reasonable accuracy, allowing teachers, students, and parents to know which curricular aims have and have not been mastered.
If a No Child Left Behind test embodied these three features, there would be no need for educators, especially competent educators, to cheat. A state’s federally mandated tests could then serve as the centerpiece of an instructionally supportive accountability system, one whose overriding function would be to help the state’s students learn better. This sort of transformation would require the vast majority of a state’s teachers to be successful—and to have the evidence to prove it: students’ improved scores on the state’s No Child Left Behind tests.
The state of Wyoming is building just such a No Child Left Behind-based accountability system, so that its instructionally sensitive tests will become catalysts to improve teaching throughout the state. As Wyoming teachers become more familiar with the new, instructionally oriented accountability tests, the need for educator cheating on the high-stakes tests will, for all practical purposes, disappear.
Would the availability of instructionally sensitive tests dissuade all educators from cheating? Realistically, no. Let’s be honest: Not all public teachers in the United States can trace their pedagogical ancestry to Socrates or Dewey. Some, even in states with wonderful No Child Left Behind tests, will be unable to get their students’ scores to soar year after year. And some may still succumb to cheating. But if appropriate tests are in place, the level of such cheating should fall dramatically.
The purpose of the No Child Left Behind law is to provide an ever-improving education for all American children. To monitor this improvement, the law requires state officials to install achievement tests of their own choosing. Sadly, many states have chosen unwisely. It’s time for state officials to replace counterproductive exams with those that can foster educators’ success, rather than their rule-breaking.