Written by Dr. Dan Brenner Friday, 16 October 2009 00:00
In Roslyn, we are justifiably proud of the performance of our students on a wide range of measures and assessments. As clearly demonstrated in a recent presentation to the board of education by Dr. Frank Banta, our assistant superintendent for Curriculum and Instruction, Roslyn students continue to achieve at very high levels on state English and math tests in grades 3 through 8, and on Regents and Advanced Placement exams at the secondary level.
Roslyn holds students to a higher standard than the state does, as Dr. Banta’s presentation makes clear; and we do not rely on state tests alone to tell us when students are doing well, or whether they are falling behind. Nonetheless, we remain troubled as educators by the manipulation of test data that we so quickly reference in the news media.
Consider the example cited in a recent article in The New York Times about the curving of grades on standardized tests. The Times article, which appeared on Sept. 14, focused on the math test taken by seventh-graders throughout the state. On last year’s test, students had only to answer 44 percent of the questions correctly to earn a passing grade. (Technically, students in grades 3 through 8 do not “pass” or “fail” the state tests in English and math; instead, scores are scaled from level 1 to 4, with students in levels 3 and 4 deemed to have scored at or above their grade level. Those in level 1 or 2 are given remediation to help improve their learning.)
The seventh-grade math test is just one example among many. In fact, it is not even the most extreme use of grading curves, which exist for just about every state test, including those required for students to earn a diploma. For example, on the algebra exam given in 2008 and 2009, a student had to score only 30 of 87 points (34 percent) to get a passing grade of 65.
Curving the grades on all of these exams is akin to the Long Island Rail Road changing its schedule so that more trains will appear to be arriving on time. This raises serious questions about how high the state’s academic standards really are and what level of achievement a diploma actually represents.
It has long been the contention of many administrators and teachers that the state’s shift away from a high school graduation standard that was based on the Regents Competency Test (RCT) to one now based on the more rigorous Regents examinations has been, at best, a shift in name only. In fact, the State Education Department has created Regents exams that are purposefully graded on a scale of less than 100 points (as in the algebra example above), thereby necessitating the creation of complicated conversion tables and the artificial scaling of the scores. In doing so, the State Education Department can better control the outcome and be assured that a certain percentage of students will pass the exams.
We must not lose sight of the ultimate goal: to challenge students academically and fully support those who struggle the most. In that light, we should applaud the state’s emphasis on using data to help guide instruction and support the necessary professional development for teachers who work directly with our children. We must be wary, however, of using the same data in a political context in order to support a conclusion that is secondary to instruction.
No one can reasonably argue that it isn’t a good idea to establish high educational standards, challenge students to do better in school, and hold educators accountable for their progress. However, a testing program will find little support among parents, students and the general public if the scoring method is difficult even for educators to understand. Merely declaring that high school students are earning a Regents diploma does not adequately show that students and schools are truly making progress.
As the state continues to trumpet its success in raising the number of students who are passing its exams, please examine the numbers from across the state with a healthy degree of skepticism. A “passing” grade is not what it used to be.
Ros super js