Shifts in test scores puzzle schools
Ohanian Note: A question for people who say their hard work resulted in test score rises: when those scores fall, what are you going to say?
For the elementary schools that saw their standardized test scores take a huge swing up or down this year, the explanations aren't always easy to come by.
At Paul and Sheila Wellstone Elementary in St. Paul, the biggest change came in third-graders' math scores. This year, 15 percent of those students scored at or above grade level on the math portion of the Minnesota Comprehensive Assessment test. That compares with 40 percent a year ago.
"The scores we saw in the paper do not reflect the hard work we're doing right now," said Principal Raymond Yu, adding that they'll look for ways to improve. "I'm not easily discouraged."
Across town and at the other end of the performance curve, John A. Johnson Elementary's principal was talking about what they've done right. Last year, 20 percent of third-graders scored at or above grade level on the reading test. This year, 69 percent did. Other scores were up at Johnson, too.
Principal Frank Feinberg said hard work by staff and additional teacher training played a key role.
"The results leave no question that No Child Left Behind focuses us" on learning, he said. "We made a real concerted effort to know the kids and know where they started the year … and to provide the services they need."
A relatively small number of Minnesota schools experienced the year-to-year test score volatility that Wellstone and Johnson did. Principals are among the first to say they don't put too much weight on one year's worth of results. But they also know the federal No Child Left Behind law, which holds them accountable for test scores, means there's still an awful lot riding on each year's scores.
INDIVIDUAL DATA NEEDED
So what explains the type of test-score jumps, up or down, that can leave teachers wide-eyed? Principals said they won't have good answers to that question until they see the individual student data. But they know what the data suggests.
At Mississippi Magnet Elementary, third-graders' math and reading scores dropped precipitously. Though 75 percent were proficient in math last year, this year the figure was 29 percent. In reading, the figure dropped from 56 percent to 33 percent.
Isis Wambui, the school's principal, noted that 69 percent of the school's students are English-language learners, a figure that's grown from 59 percent last year. Mississippi is also one of the St. Paul district's Language Academies, a program geared toward recent immigrants who have little or no education background and don't speak English.
Still, Wambui and her staff believe there are plenty of good things going on in third-grade classrooms.
"We were very surprised when we saw the test scores," Wambui said.
This year's test scores are still higher than they were five years ago, she said, noting that the long-term trend at the school is still up. "We took a step backwards, but it just motivates us to work harder."
At Hancock/Hamline Magnet Elementary, another Language Academy site, fifth-graders' scores dropped significantly, too. Principal Marjorie Warrington said she needs to see the individual scores before trying to explain the decline. Though many of the school's fifth-graders are new to Hancock/Hamline and weren't there for the third-grade version of the test, "I don't know if those are the students who scored poorly," Warrington said.
The issue of Englishlanguage learners' performance on standardized tests looms large in St. Paul, where almost 40 percent of students qualify for the language program. In February, shortly after U.S. Secretary of Education Rod Paige attended a roundtable discussion of No Child Left Behind with Minnesota educators, the federal department altered some of its test-taking rules.
During their first year of enrollment in U.S. schools, English-language learners can take the standardized tests, but their scores don't have to be included as part of the calculation to determine if a school is making adequate progress. Also, once a student does become proficient in English, his or her test scores can be included in the subgroup of students still learning the language. That allows the schools to get credit for improving students' English skills.
At Hancock/Hamline, some of the immigrant students didn't have to take the test because they'd been in school less than a year, Warrington said. "Others took it because teachers felt they should, they felt their skills were high enough," she said. Warrington also said that her school's long-term trend line, even with this year's decline, is still up from five years ago.
WATCH SCORES OVER TIME
The other large caveat in shifting test scores — one that not all educators are comfortable talking about — is the variation in each class' abilities. The way the Minnesota Comprehensive Assessment test scores are calculated, this year's results are measuring a different group of students than did last year's results.
"You'll see large fluctuations in many schools due to changes in the incoming group of third- or fifth-graders," said Dave Heistad, the research director for the Minneapolis Public Schools, which also saw their share of elementary schools with shifts in test-taking performance in the most recent results.
The shifts have been shown to correlate with the percentage of children in poverty and the percentage of children still learning English in certain schools, he said. Of the four St. Paul public schools with the biggest test score declines, three have at least 60 percent of students in English-language learner programs.
Heistad said he tells principals of schools that have seen big shifts "not (to) get too depressed when they go down, and don't get too euphoric when they go back up." It's better, he said, to look at scores over a longer period of time.
What Minnesota educators are hoping for, he said, is the ability to include a "value-added" measurement for each group of students, which would allow a comparison of the same students year to year. That would make sure low-scoring schools get credit for gains students do make, regardless of how low they were when they started.
By 2006, when all students in grades 3 through 8 will take the Minnesota Comprehensives, that type of value-added comparison could help keep schools off the list of "failing" schools.
Until then, "we're stuck with this system for the federal purposes for adequate yearly progress," Heistad said.
Find Minnesota Comprehensive Assessment test results for schools and districts at the state Education Department Web site, education.state.mn.us.
John Welbes can be reached at email@example.com or 651-228-2175.
INDEX OF NCLB OUTRAGES