Threatened for the last decade with bad publicity and sanctions following poor test results, school districts have been known to grab at loopholes in state testing policy and rip them wide open.

Critics say that's happening again with a new test for special education students called the California Modified Assessment.

Introduced in 2007, the CMA is a simpler version of the state's regular STAR student achievement test. It's tailored to special education students in grades three through 11 whom teachers and parents deem to have no chance at passing the regular test.

The federal government issued guidelines to the state saying the new test should be given to no more than 2 percent of students in those grades – about 100,000 children.

Four years later, almost 200,000 students are taking the test – a number that will likely grow as the CMA gains momentum.

The trend has consequences beyond special education.

CMA scores are tallied separately from scores on the regular test, the STAR California Standards Test. By removing failing students from the pool of kids taking the regular test, districts end up with a greater proportion of high-scoring students.

The CMA has inflated gains on the regular STAR test by about 25 percent statewide since 2007, according to a Bee analysis.

"It's the old business of if you want your test scores to go up, don't test the lower-scoring students," said Doug McRae, a retired testing consultant and Monterey resident who helped the state design the STAR test.

McRae has gone public with his objections to the CMA, recently telling the State Board of Education – his former employer – that districts are abusing it. He said the CMA has some value but many students should instead take the regular STAR test.

"This is an easier test," he said of the CMA. "If we don't push these kids as far as they can be pushed, they might not be functional in society."

Many special education advocates and school leaders disagree, saying the CMA gives students who fail the regular test a chance to do better.

The new test also lets schools better measure student progress, they say. Arbitrarily capping the number of students who can take the test demoralizes teachers and their students.

"At the end of the day, I think (special education students) feel a bit more successful with this," said Kristin Wright, who has a special needs child and chairs the state board's special education advisory committee. "It's like a sigh of relief for the kids that it is not overwhelming."

To qualify as a special education student in California, a student must be identified as having a physical disability such as deafness or autism, a learning disability such as dyslexia, or a serious emotional or behavioral problem. The number of special education students statewide has held steady for several years.

Generally, only special education students who performed at the "below basic" or "far below basic" proficiency levels on the previous year's regular STAR test are eligible to take the CMA. It covers the same concepts as the regular test. But it's different: Reading blurbs are shorter; fewer options are given on multiple-choice questions; pictures are used more often.

No Child Left Behind link

The argument over how to administer and count the new test is central to California's relationship with No Child Left Behind, the decade-old federal initiative that assesses school performance through testing and holds schools accountable for the results.

The primary tenet of No Child Left Behind is that all students will become proficient in math and English skills. The CMA is an acknowledgment that such a lofty goal may be unrealistic – some kids just can't pass the same test as everyone else.

The federal government allows states to give 1 percent of their students – the most disabled – a different test. In 2005, the feds also gave states permission to create additional tests like the CMA, for special education students whose disabilities are considered less severe.

But federal officials strongly recommended states offer it to no more than 2 percent of students. Otherwise, the argument went, districts might not strive to help every student reach his or her potential. To give that idea teeth, they said only scores from 2 percent of students taking an alternative test could count toward federal benchmarks.

When California approved the CMA four years ago, state education leaders endorsed that 2 percent figure, which equals about 20 percent of special education students. Then-Superintendent of Public Instruction Jack O'Connell said the new test would "provide California with important flexibility to measure the academic achievement of 20 percent of students with disabilities."

But the state didn't specifically mandate how many kids could take the CMA. And it decided not to place a cap on the number of good scores on the modified test that could be counted toward state standards, which are different from federal benchmarks.

Four years later, school districts statewide are administering the test to 40 percent of special education students – twice the federal recommended rate.

Dozens of school districts, including Sacramento City Unified, are giving students the test at almost three times the expected rate.

As a result, some districts have posted gains on the regular STAR test that really don't exist.

Sacramento City Unified, for instance, moved about 300 seventh-grade students to the new test during the past four years. With the shifts, the district's seventh-graders reported a five-percentage-point increase in English proficiency on the regular STAR test. Without the shifts, the increase would have been nil.

The new test has also inflated many schools' Academic Performance Index scores – a composite number generated from test results that lets parents compare one school against another, McRae said. The state treats the CMA the same as the regular test when computing the API.

About 40 percent of statewide API improvements since 2007 are the direct result of switching students to the new test, according to McRae's calculations.

School leaders disagree

School leaders don't argue with McRae's math, but they do question the assumptions behind it. Before the new test for special education students was introduced, they say, parents got a false picture of how their schools were performing: The old test scores were wrong, so why make comparisons to them?

"I believe in testing and using the data, but at the heart of it are children," said Mary Shelton, chief accountability officer at Sacramento City Unified. "It's not to inflate scores. It's to allow them some success if they are capable of it."

Several school officials argued that shifting special education students to the CMA has nothing to do with inflating test scores and everything to do with the impractical, one-size-fits-all approach of No Child Left Behind.

The federal program has recently come under attack from multiple quarters as more schools fail to meet its ever-rising standards. Legislators and federal officials are discussing modifications to the program.

At Center Joint Unified in north Sacramento County, half of the special education students take the CMA. At one school, Spinelli Elementary, roughly one of every six students in the entire student body takes it.

Before, these students would score "far below basic" on the regular STAR test, year after year, even if, in reality, they learned much and grew significantly, said Center Joint Superintendent Scott Loehr.

With the CMA, by contrast, "We're better able to get a true picture of how they're doing," Loehr said.

In the Sacramento City Unified district, repeated failure on the regular test was beating down many special education students, said Gabe Ross, the district's spokesperson.

"Is it more accurate to give students who have special needs a test that we know they will not be proficient in?" he asked. "How does that give you an accurate picture of student learning?"

School comparison tough

Because districts give varying proportions of students the modified test, it makes it hard to compare one school to another – a key goal of No Child Left Behind.

Schools that give most of their special education students the modified test could artificially appear to be doing a better job than schools that give the test to fewer, much like a school where only honors students take the SAT will, at a glance, look better than a school where all take the test.

In a speech earlier this year, U.S. Education Secretary Arne Duncan said tests like the CMA were regularly abused, and he vowed to "move away" from them. "We will not issue another policy that allows districts to disguise the educational performance of 2 percent of students," he said.

As for the hundreds of California school districts giving modified tests to more than 2 percent of students, federal guidelines are clear: A district "that assesses significantly more than 2 percent of its students with an alternate assessment … should prompt a review by the state."

No such reviews are forthcoming. When asked whether the state looks at the policies of districts giving the modified test to a large number of students, Rachel Perry, assessment and accountability director for the state Department of Education, said, "There is no restriction on the number of students who can take the CMA."

That number is likely to grow because the CMA is still in the final stages of its rollout. Many high schools, in particular, haven't had time to move students to the new test.

If it won't restrict the number of students who can take the modified test, the state should treat its scores much differently when calculating a school's academic performance, said McRae, the testing specialist. Otherwise, he said, districts like Sacramento City "are playing games … to float scores upward."

After McRae brought his concerns to the state Board of Education last month, the board asked staff at the California Department of Education and its special education advisory committee to look into the issue at an upcoming meeting.

Yvonne Chan, a state board member, told McRae that she is more interested in giving school districts the flexibility to assess special education students as they see fit than in imposing arbitrary limits.

"Some schools may have more; some schools may have less," Chan said at the meeting, referring to special education students who can benefit from the test. "This is not something where we can say, 'Sorry, you are over the limit.' "

© Copyright The Sacramento Bee. All rights reserved.

---

Call The Bee's Phillip Reese, (916) 321-1137.

• Read more articles by Phillip Reeseand Melody Gutierrez