by Dr. Thomas Jefferson, Superintendent of Schools

First in a three part series

There is perhaps no greater lightning rod in the debate about public education than the role of standardized testing. In some communities, property values rise and fall based on how children perform on particular multiple-choice tests. Policy makers at the federal level are proposing that teacher pay be linked to test performance as well. Acceptance into college, and even graduation from high school, are often linked to performance on one test or another. Bromfield’s recent selection as one of the Top 100 High Schools by U.S. News and World Report, based to a large extent on test performance, has triggered much discussion about the role and value of testing.

Standardized tests fall into two primary categories. Criterion-referenced tests measure how well students perform relative to particular standards. The MCAS is an example of this type of test. Numerical standards are set to establish whether a student scores in a category of advanced, proficient, needs improvement, or warning of failure. Interestingly, despite the fact that criterion-referenced tests are not designed to rank students, schools, or districts, the Boston Globe provides an artificial ranking immediately upon public release of statewide results.

In contrast, norm-referenced tests rank students in comparison to one another. The Scholastic Aptitude Test (SAT) is the most well known of the norm-referenced tests currently used. An SAT score of 500 is by definition average, with each 100 points above or below 500 representing one standard deviation. The data we glean from analysis of both types of tests can help us identify areas where we need to provide support or re-teaching to an individual student as well as where we need to teach a particular concept or skill differently to all students.

In Harvard we have historically done fairly well on most types of standardized tests, as well we should. We have a formula that works well to support good test scores: many bright students from enriched home environments; a talented teaching faculty and support staff; a curriculum aligned with state frameworks and standards; and a school culture that values academic performance. Districts like ours that perform well on MCAS are often accused of “teaching to the test.” This phrase has developed a negative connotation, conjuring up images of drab, unimaginative classrooms that are endless drill and rote repetition.

An alternative mantra might be that a district should “teach to the standards.” In Harvard, we look to align our instruction with the state curriculum frameworks. Within each discipline, the Department of Elementary and Secondary Education has identified key concepts and skills appropriate for the grade level and subject matter. The curriculum frameworks do not limit the types of pedagogy or creativity that individual teachers can bring to their classroom. Rather, they provide an overarching structure as to what concepts and skills should be learned. Assessment tools such as the MCAS provide a mechanism to measure and track how well students are learning the concepts and skills identified in the frameworks.

People often ask me how important it is to perform well on standardized tests. The answer I offer is, “it depends.” After all, standardized test scores are only a single measure and must be considered in context with other assessment data and anecdotal evidence about a student’s performance. Many talented and capable people simply do not test well. Yet, some tests inherently do have high stakes attached to them. For example, if a student is competing for a slot at an elite college, performance on the SAT does carry significant weight. In Massachusetts, a passing performance on the 10th-grade MCAS is a requirement for graduation.

Our goal is to have all students master grade-level curriculum standards in all subjects. Much instruction is sequential and builds upon a foundation that must be mastered in previous grades. That said, the composition of tests and test-takers yields results that often fluctuate within the younger grades. The near 100 percent proficiency levels we traditionally realize in the upper grades are not as easily attainable with 8- and 9-year-olds. Yet, whenever a child falls into the “needs improvement” category in MCAS at any level, we review the child’s performance data to ensure that we’re providing the appropriate supports and instructional design. We look at student data both individually and collectively to determine where we need to work with an individual student and where we need to address the instruction within the classroom.

We seek to maintain a balanced view of the role and impact of standardized tests. We should strive to do well individually and collectively. To dismiss the role and value of standardized tests is naïve. Data should, and does, inform much of the instruction we provide students. Still, we must keep in perspective that only a portion of our success as individuals and a community can be measured through tests. As one educational theorist noted, “not everything that can be measured is important, and not everything that is important can be measured.”