Page One

Berkeley school officials believe that recently-reported “plummeting” state test scores at highly-rated John Muir Elementary School are incorrect and are seeking to have them revised by the state Department of Education. 

The accumulated summary of Muir’s fourth grade scores on the California Standards Test (CST)—taken last May but only released to the public this month—appeared to show that the students had dropped 30 percent in English Language Arts from last year to this. 

But Berkeley Unified School District (BUSD) Director of Curriculum and Instruction Neil Smith says that the score summary by New Jersey-based Educational Testing Services (ETS) is “absolutely” wrong, and does not agree with a manual adding up of the students’ actual test scores in the same report. Smith offered no theory as to how the ETS summary might have occurred, but said that he had “never seen an error like this before.” 

John Muir principal Nancy D. Waters said she was “relieved” at the discovery of the possible discrepancy between the test summary and the actual student scores, and looks forward to getting the matter cleared up. 

“My heart sank when it looked like we had gone down so much,” Waters said. “I didn’t know how it could have happened. I knew our teachers had worked 200 percent last year.” 

Waters said that a number of concerned parents contacted her office in the few days after the scores were published in local newspapers and posted on the websites of both the California Department of Education and the non-profit GreatSchools.net organization. 

CST scores are reported by the State of California under the STAR (Standardized Testing And Reporting) program, and are listed in five performance level categories: Advanced, Proficient, Basic, Below Basic, and Far Below Basic. Students in second through fifth grade are tested. 

According to the STAR summary, 23 percent (10 out of 43) of Muir’s fourth grade students scored in the lowest category, Far Below Basic, in last May’s English portion of the CST. But both Waters and Smith said that after they did a manual count of the individual student scores on which the summary is supposed to be based, only two of the students actually scored that low. 

Waters and Smith said the summary totals for Proficient, Basic, and Below Basic scores were less than the actual count of Muir students who scored in those categories. 

California public schools suffer no state rewards or penalties for the information listed on the CST summaries. A spokesperson for the California Department of Education said that the state agency bases its overall school rankings on the data taken from the reported individual student test scores themselves. And, in fact, Muir’s score on the California Academic Performance Index (API) actually went up this year—from 815 to 821 out of a possible 1,000 total score—keeping the school in the top 20 percent of all state schools with a comparable grade range. API scores are based, in part, on the individual student score results of the CST tests. 

But in an era when testing has become the popular judgment criterion for schools, the CST summaries—whether good or bad—can have a profound effect on both the staff morale and the teaching strategies of the schools themselves, as well as on the judgment of individual schools by the public. 

Lisa Rosenthal, senior editor of GreatSchools.net, said that “a large percentage” of people who access her organization’s site “are looking at the test scores as a criteria in choosing a school. We always provide the caveat: don’t judge a school by the test scores alone. There are lots of other factors. But [test scores are] definitely a first stopping point [used by parents] to decide on schools.” 

Principal Waters said that after she got the initial report of the test score summary last month, she was already “brainstorming [with colleagues] over how to proceed” to make up for what she thought were the school’s newly-discovered deficiencies. 

“I was preparing for the first staff meeting of the year, and I was trying to figure out how to motivate the staff after hitting them with such bad news right off the bat,” Waters said. 

After reviewing the John Muir test data online, Bob Bernstein, a Department of Education administrator, speculated that Educational Testing Services may have used what he called “modifications” to alter the totals in the summary, while leaving the individual test scores untouched. He said he had not previously heard about the discrepancy in the John Muir scores. 

Bernstein said that might account for why a manual adding-up of the individual scores does not bring the same figures as reflected in the summary. Bernstein said he found, for example, that nine Muir fourth graders were listed in the CST data as having needed a teacher to read to them certain passages during the test, rather than reading the passages themselves. 

Educational Testing Services could not be contacted for this story. 

GreatSchools.net’s Rosenthal called the Muir test discrepancies “puzzling,” and did some speculating herself, reasoning that “it could be a problem with the scoring. It could be a problem with the reporting. A whole host of things could be going wrong.” And she said she’d seen such problems before. 

“I was on a school board in Burlingame several years ago, there was some test that seemed way out of whack with the test results, and they went back and found that the state had made a mistake on the scores,” Rosenthal said. “When school districts get these results, it’s a good thing they go through them with a fine-tooth comb and try to figure it out, because very often the scoring is in error. So people need to question that.” 

BUSD’s Smith said a representative in the district office is expected to contact the State Department of Education today (Friday) with a formal query on the Muir test score discrepancy.