For parents with children in American public schools, spring is standardized testing season. Some teachers feel stressed, and communicate this to students. As my fellow educational researchers Gary Ritter and Marc Holley report, standardized testing takes just 1 percent of instructional time.1 Its effects are far greater. Advised by curriculum coaches vetted with master’s degrees from education schools, teachers devote anywhere from minutes to months to test prep. And yet, for state accountability metrics like the infamous PARCC (Partnership for Assessment of Readiness for College and Careers)2 exams imposed by the Common Core State Standards, testing companies report scores six months or more later—long after anyone feels ownership of the results and can use the data to guide improvement. As a friend who runs a successful charter school put it, kids wouldn’t care about a video game whose scores came back months later, so why would they care about standardized tests?
Given all this, why do we test?
After decades of allotting more and more money to education and getting mediocre results, the political class lost faith in the educational industrial complex. Over the past two decades Democrats and Republicans joined to impose standardized testing on largely hostile public schools. Testing strikes at all the educational industrial complex holds dear. Members of organizations like the American Educational Research Association and the National School Boards Association—I belong to both—view academic achievement as elitist, and thus unethical. They further consider it culturally insensitive to expect disadvantaged children to learn, and of course when one considers poverty, family structure, ethnicity, and special education status a great many children are disadvantaged in one way or another.3 From these core beliefs held by these organizations follows another: nontransparency is essential, since policy makers cannot comprehend the awesome work public schools do, work that has little to do with measured academic learning. Testing provides policy makers data they might use to intervene in our thing.4
Transparency from testing can embarrass ineffective educators and protect the effective from reprisals. The latter is hardly hypothetical. Central office administrators tried to close the original Knowledge Is Power Program (KIPP) campus known for large achievement gains by disadvantaged students; only the intervention of then-Houston superintendent Rod Paige saved the school.5 Paige himself has chronicled the lengths to which mediocre educators and their allies go to sabotage teachers like Jaime Escalante, of Stand and Deliver fame.6 I’ve found similar cases in fieldwork. Transparency makes such mendacity unpolitic. The transparency from standardized testing has enabled academically successful charter school networks like KIPP, YES Prep, Harmony, Basis, and Great Hearts to gradually replace less effective traditional public schools. Notably, such schools are typically founded and staffed by educators from Teach For America and other nontraditional teacher preparation organizations.
The success of these charter schools and of certain traditional public school districts like Rogers (Arkansas) reflects use of data not from state accountability tests like the molasses-slow PARCC, but from computerized assessments like the Northwest Evaluation Association’s Measures of Academic Progress (MAP). MAP tests are easy to administer, typically taking under an hour, immediately scored, cheap (about $10 per child for three tests a year), and precisely nationally normed—enabling teachers and students to set realistic goals and celebrate when they succeed. An educator related seeing two children high five after MAP showed similar achievement gains: one moved from the eighty-fifth percentile to the ninety-seventh percentile while the other went from the fifth percentile to the seventeenth. On MAP, each was a winner.
My wife and I despaired when our school district ditched MAP in a cost-cutting move. Given the wholly inadequate implementation of PARCC, that means that for well over a year we’ve had no idea if our daughter is learning. Of course we have grades, but grades range from teacher to teacher and school to school, with reasonably smart, well-behaved kids typically getting all As. In one school, a new principal nearly doubled the number of students “earning” straight As in just one year to roughly a third of the study body; such is the objectivity of grading.
Such subjective grading furnishes a final reason for standardized testing. As James Coleman pondered back in the 1990s, why do the same American students who work hard at football hardly work in class?7 In part that’s because in coursework teachers set the standards, and this motivates students, parents, and administrators to pressure teachers to lower them. Grading puts students and teachers in conflict. Grading can also lead students to resent high scorers as rate busters. In contrast, the standard for football is set by the opposing team on Friday night, and no one can lobby them to go easy. That puts players and coaches quite literally on the same team, trying to beat the external standard set by the other team; thus players cooperate when coaches make them work. Standardized testing sets up that same dynamic in foreign schools, and in certain American contexts. Academically successful schools and even individual teachers unite their students in the quest to beat the state test. Last year our son studied under a dedicated teacher who typically produces the top scorers on the state algebra tests.8 No one complained about the workload because students and teacher shared the goal of beating the other schools, whose performance set the standard.
Clearly, standardized testing does more good than harm. Yet this does not mean that all is well. Again, elites forced testing on an educational establishment that manifestly did not want it. As I have pointed out in these pages before, most public educators do not do very well on standardized tests. They do not value the content knowledge and verbal acuity that tests measure, seeing those as distracting from the real work of schooling, group projects and extracurricular activities.9 Traditional public educators are also highly risk averse, so they avoid performance data, fearing embarrassment.10 This means that too many public schools have implemented standardized testing via shortcuts, hiring pricey consultants with test prep tactics. Recently a principal I know stated his intention to increase the number of National Merit Scholars at his school through SAT preparation workshops. He did not even understand my suggestion that over time, requiring students to read books at or above grade level (rather than two or more grades below) might have far more influence. It never occurred to him that we might raise student test scores by improving student learning, as indeed successful schools have done for centuries.
In short, standardized testing merely provides data, which educators are free to use, abuse, or ignore. The problems come not with testing, but with the educators charged with implementing testing. One might as well ask pacifists to run the Pentagon as expect conventional educators to use data to improve learning.