Validity of School Tests Questioned : Education: Fewer of the new CLAS examinations were graded than required by the state’s own guidelines.

Ventura County school officials were shocked to learn that state educators scored a far smaller sample of the new student assessment tests than required by their own guidelines, calling into question the validity of the results.

And some officials said they were even more surprised to hear that the California Department of Education intentionally scored only a sample of the tests.

“I’m astonished at only 25% being scored,” county Supt. of Schools Charles Weis said.

Countywide, about 8,000 students in the fourth, eighth and 10th grades took the California Learning Assessment System tests for the first time last spring. But only a fraction of the tests they took were scored.

California education officials maintained that the samples scored would be statistically sound, measuring a school’s performance in reading, writing and math. But the state failed to meet even its own minimum standards on how many tests from each school should be scored, a Times computer analysis found.

The analysis found that results in 83% of the schools in Ventura County and half of the schools statewide were less precise than promised because too few examinations were scored. But the vast majority of the samplings fell short by only a few tests.

Some county educators said the scoring problems will surely undermine the credibility of the revolutionary new assessment test.

“If they don’t implement their guidelines, it seems to me it would ruin the results,” Weis said. “It’ll jeopardize the effect of the results without a doubt.”

Unlike other standardized exams that only ask students for the correct answer, CLAS assesses how well students write, think and solve problems. Because of the tests’ complexity, state educators decided to score only a sample of the handwritten answers to the first year’s exams.

The biggest scoring problems in Ventura County, as elsewhere in the state, were with the tests of fourth graders in elementary schools. For example, of the 114 elementary schools in the county that had students take the exams, 73 had too few reading tests scored to meet the state standards.

The problem was greater at elementary schools because they tend to be smaller than junior highs and high schools. And state guidelines called for scoring all tests at small schools, while culling only a sample from larger schools.

In most cases, the sampling of tests scored fell slightly short of state guidelines. At Anacapa Middle School in Ventura, for example, 331 students took math tests and 105 of these were scored, nine less than the state required.

But at two of the county’s schools--Sequoia Junior High in Simi Valley and Flory School in Moorpark--the shortfall was more significant.

At Sequoia, 360 eighth-grade students took the CLAS math tests last spring. Only 14--3.9%--of the tests were scored.

Despite the extremely small size of this sample, the state extrapolated from the results on those 14 tests to calculate the overall performance of Sequoia’s eighth-grade class.

And the results weren’t good.

Sequoia’s math test results were among the worst in the county; the tests indicated that 81% of Sequoia’s eighth-grade students have only a limited understanding of mathematical ideas and rarely use the correct techniques to solve math problems.

Simi Valley school officials said they suspected something was wrong with Sequoia’s test results.

“When I looked at Sequoia’s (scores), I knew automatically we had a problem because of the number of kids graded,” said Becky Wetzel, testing director for the Simi Valley Unified School District. “It stood out like a sore thumb.”

Wetzel said she called state education officials the day she received the scores, then faxed them a letter the following day asking why so few tests were scored. She has not yet received a reply.

“It’s disappointing because we don’t know what the real assessment looks like for that school,” she said.

At Flory School in Moorpark, 215 fourth-grade students took the tests. But only 52, or 24%, of their reading tests were scored. That made Flory one of only 148 schools in the state that had less than the required minimum 25% of its tests scored.

Moorpark school board member Clint Harper said he wanted to examine the Flory CLAS results more closely before saying whether the scoring problem made a significant difference.

But, Harper said, the state should have scored all of the tests at every school. “Ideally, as a board member, I’d like to see all the tests scored.”

State officials have promised on future CLAS tests to score a higher number of papers than were read in the first year of the exam. The next round of CLAS tests will be given this spring.

Harper and other school officials said they view the scoring problems on the first round of tests as a simple glitch that can be worked out.

“The test was being tested probably as much as the schools and the students were,” Harper said.

Wetzel agreed.

“I don’t think you throw the whole thing out because you question it,” she said. “I think you question it, revise it and try to make it better the next time.”

But Weis said that he believes the scoring problems prove that the state has taken on more than it can handle with the CLAS exams.

“The concept of the test is outstanding,” Weis said. But, he said, “When you get something as large as the state of California dealing with this, there can be huge logistical problems. It’s such a huge undertaking, I don’t think we have the resources to do it right.”

*

RELATED STORY: A1