|
|
9486 in the collection
Botched Most Answers on New York State Math Test? You Still Pass
Ohanian Comment: If you needed further proof that test pass and failing scores have nothing to do with science and everything to do with politics, here is one more piece.
I wish someone who is both math savvy and kid savvy would deconstruct one of these tests--just lay it all out for us--let ordinary people decide if what's on this test is appropriate.
I say this remembering the time I took a state exit exam loony question about compound interest on a house mortgage into a number of banks. I just walked in and asked the tellers if they knew how to do this problem. . . or if they saw anything wrong with it. In every bank, the teller called the manager. So I also asked the managers if they could solve the problem.
None of them could. They all told me they used "established tables" to figure mortgage interest.
Oh well.
NOTE: Meryl Tisch, also mentioned in this article, is the daughter-in-law of Loew's CEO. Her own children went to private schools, and she concluded from their taking test prep for the SATs that test prep provides a good tool for learning. (New York Times, March 6,2002) Not surprisngly, when asked to grade the mayor for his takeover of New York City's schools, Tisch gave him a B, saying, "The whole reorganization has been a plus." (NY Post June 14, 2004)
By Javier C. Hernandez
For many students, bungling more than half the questions on a test would mean an F and all that comes with it — months of remedial work, irksome teachers and, perhaps, a skimpy allowance. But on New York State’s math exam this year, seventh graders who correctly answered just 44 percent of questions were rewarded with a passing grade.
What gives?
Three years ago, the threshold for passing was 60 percent. In fact, students in every grade this year could slide by with fewer correct answers on the math test than in 2006.
In math this year, 86 percent of students statewide passed the test, scoring at least a Level 3 (of 4), and 82 percent passed in English, the highest percentages in many years. But the lower passing scores, especially for math, have provided fodder for skeptics who believe the state has made it easier for struggling students to pass.
In New York City, which saw some of the largest gains, Mayor Michael R. Bloomberg has used the scores as evidence of his successful leadership of the schools. The jump in passing rates also helped 97 percent of the city’s schools earn A’s or B’s on their Department of Education report cards this year.
The mayor also has repeatedly pledged to hold back students who fail the test and do not meet even Level 2, a minimal standard. But the number of right answers needed to reach Level 2 has also dropped, to the point that on some tests, a student could randomly guess and still stand a good chance of moving on to the next grade.
State education officials say that they have not made it easier to pass the tests, and that the scoring thresholds have dropped for a simple reason: the test questions themselves have actually become harder. As a result, they have reduced the number of correct answers required to pass some exams to make the tests comparable over the years.
At a time when the tests are assuming an unprecedented role in classrooms across the state — used for everything from analyzing student deficiencies to determining which educators deserve cash bonuses — the debate underscores a central question: How accurate are the exams in measuring student learning and progress, and what skills should a passing grade reflect?
Already, New York is under significant pressure to overhaul its tests. The State Board of Regents has instructed CTB/McGraw-Hill, which develops the tests, to revamp the 2010 exams by testing a broader slice of the curriculum. The tests also will be given later in the year, in May, to allow classes to cover more ground. And state officials have said they plan to tinker with the scoring to make it harder for students to pass.
In addition, the state's contract with CTB/McGraw-Hill ends next year, and some testing experts are suggesting that the state entirely reconfigure the test’s design to better shield against score inflation. The company declined to comment, referring questions to the state.
Merryl H. Tisch, the chancellor of the State Board of Regents, said she supported lifting the score required to pass because, she said, the tests had become too predictable. Federal exams, she noted, do not reflect the same gains in learning over the past several years that state English and math tests do. (Some attribute this to the fact that federal tests are not modeled on state standards, so students might be tested on skills they have not yet mastered.)
“We are painfully aware of the fact that our tests do not align with the national tests,” she said. “We need to align new standards with a new set of assessments that are rigorous and dependable.”
Across grade levels, the percentage of points required to pass the math tests — or "meet learning standards" — has declined over the past three years. English tests have shown less significant decreases, and on fifth- and sixth-grade English tests, the cutoff score was in fact slightly higher this year than in 2006.
Before a test is ready for students, it goes through several layers of vetting. Teachers and other educators comment on hundreds of questions drafted by the testing company. State officials choose the best ones and give field tests to thousands of children across the state, representing a range of abilities and various ethnic backgrounds. The children take the tests after their regular state exams and are told to give them their best effort, even though they do not count.
After making sure the tests are not biased against specific ethnic groups or genders, testing experts draft an exam with a mix of easy and hard questions. If the questions, during the field testing stage, prove more difficult than the previous year’s questions, the state reduces the number of correct answers needed to pass, thus "standardizing" it with previous tests. Conversely, if the new questions are easier, the cutoff is raised.
State officials said the increasing difficulty of the tests was the result of several factors. In 2006, after the results for the state math test came in, state officials determined the questions were mostly too easy, and so they sought to toughen certain areas in 2007. More recently, the questions that educators have recommended have tended to be harder, the officials said, though there was no intentional effort to increase the rigor.
Critics say that the test-development process is prone to error, and that students are unlikely to take a field test seriously because it has no consequences.
Some have suggested adding questions to the tests that require students to show they know the material in different ways. For instance, a test might ask students to identify the theme of a fairy tale by reading several different passages and posing the question in several different ways, to make sure students have not just mastered the art of test-taking.
David M. Abrams, an assistant education commissioner, said the state was contemplating such a move. But he noted that outside testing experts had already validated the state’s testing practices.
"We have gone above and beyond the due diligence," he said.
A side effect of the adjustments in scoring is that on 5 of the 12 math and English tests this year, a student had a better-than-even chance of earning a Level 2 mark— a failing grade that reflects "partially meeting learning standards" — simply by guessing. On the sixth-grade English test, for instance, a student had an 89 percent chance of reaching Level 2 this year by randomly guessing, according to an analysis by The New York Times.
State officials disputed the importance of those findings, noting that students must earn a Level 3 to pass, and that the tests include decoy answers meant to dupe students intent on guessing.
In the city, students must earn at least a Level 2 to move on to the next grade, but Mr. Abrams said the tests were never designed to be used to make those judgments.
Jennifer Bell-Ellwanger, a senior New York City education official, defended the use of that benchmark, saying students who reached Level 2 had a better chance of academic success than those who scored worse on the test. She played down the guessing effect, saying decoy questions made it nearly impossible for a student to blindly reach Level 2.
Daniel Koretz, a professor of education at Harvard, said that the guessing phenomenon was "clearly a problem," and that it reflected a need for a test that more clearly distinguished between achievement levels.
"The public is being told that kids who hit Level 2 are doing substantively better than kids who are hitting Level 1," said Dr. Koretz, who served on a committee that advised the state on testing from 1996 to 2005. "I think that's misleading."
Still, Mr. Abrams said he was not concerned about the distinctions between levels. He said the state tried to make the tests as difficult as possible, though he declined to say whether they had become too easy.
"The tests are designed to determine if students are meeting the state standards," he said, "and I believe we have been able to track that."
Robert Gebeloff contributed reporting.
New York Times
2009-09-14
http://www.nytimes.com/2009/09/14/education/14scores.html?emc=tnt&tntemail0=y
INDEX OF OUTRAGES
Pages: 380
[1] 2 3 4 5 6 Next >> Last >>
|
