Report: Achievement gap closing in some Stamford, Norwalk schools
By Alexandra Fenwick, The Advocate, November 5, 2007

A statewide research group has given schools in Stamford and Norwalk high marks for closing the achievement gap between white and minority students in some schools.

Last month, the Connecticut Coalition for Achievement Now, a New Haven group, released its 2007 report, which rates the progress of elementary, middle and high schools in five categories.

The ConnCAN report measures progress, not just scores, providing an alternative to federal No Child Left Behind analyses, which label schools as passing or failing depending on whether they hit certain goals.

Using scores from the 2006 and 2007 Connecticut Mastery Tests and Connecticut Academic Performance Tests, ConnCAN ranks the Top 10 schools in performance gains, most improved, African-American scores, Hispanic scores and low-income scores.

Four Stamford elementary schools and two in Norwalk made the Top 10 list for Hispanic students scoring at or above goal on all subjects.

In Stamford, Rogers took third place on the Top 10 list, Westover took fourth, Julia A. Stark sixth and Stillmeadow eighth. In Norwalk, Kendall was seventh and Marvin was 10th.

Rogers in Stamford was among the five schools appearing most often in this year's Top 10 lists. It scored third most improved among elementary schools. Most improved is measured by the change in the average percentage of students within goal range across all subjects from 2006 to 2007.

Stamford Superintendent Joshua Starr said Rogers has been successful because expectations are high for every child.

"There is a caring environment," Starr said. "The parents and the community are very involved in the school."

With 10 mentions, Stamford was among the districts appearing most often on the lists.

Among middle schools, Scofield Magnet in Stamford took fourth on the top 10 list with 59 percent of Hispanic students scoring at or above goal. Scofield also scored eighth in percentage of black students who scored within goal range, with 59 percent.

Kendall in Norwalk increased its percentage of Hispanic students scoring within goal range to 63 percent, up from 56 percent last year. At Marvin, it was 58 percent.

In Norwalk, Nathan Hale Middle School scored 10th for high-performing Hispanic students, with 52.8 percent.

Kendall Principal Tony Ditrio said he likes accountability but it should be fair.

"I don't mind being judged by those (tests), I just want someone to take into account that traditionally some groups don't do as well as others," he said. "That's where the rubber meets the road and that's why ConnCAN is so important to us. Connecticut is mostly white and affluent and cities that are poor. Those rich communities don't have the issues we deal with. They look at NCLB and it's no big deal. They don't even get scored for half of the subgroups because don't have enough kids."

No Child Left Behind does not measure subgroups, such as black, Hispanic or low-income students, if fewer than 40 such students are in a school.

But Starr raised questions about ConnCAN's methodology, noting limitations in available data.

"Everything has to be taken with a grain of salt," he said.

The ConnCAN report measures, for example, fourth-graders' scores in 2007 against their scores as third-graders in 2006 and points out gains in narrowing the achievement gap.

"The most important indicator is growth of students' scores in schools," said Marc Porter Magee, ConnCAN director of communications and research. "If you look at a snapshot, schools like Wilton or Greenwich are at the top, but they're not necessarily adding more value. Their students come in at 90 percent and maybe go up to 91 percent. But when you look at a cohort of students, the same class from year to year, some schools are making huge gains year to year and being overlooked."

By following cohorts, progress is measured, he said.

"We think that's a really important way of looking at schools and that's something our state Department of Education doesn't do," Magee said. "It gets us a little out of the NCLB framework. Urban schools, more often or not, are going to be labeled as failing even if they're making progress."

ConnCAN's report has been criticized in part because the group supports more funding for charter schools, which it believes does the best job in closing the achievement gap. Last year, the Connecticut Education Association commissioned a study that poked holes in the way ConnCAN compiles its reports.

The study noted that, despite improvements, some schools in the most improved categories are still not very good. It also cited ConnCAN's failure to address the number of students who join or leave a class each year. The study recommended that ConnCAN use matched samples rather than compare the fifth grade of 2007 to the fifth grade of 2006, for example.

But the state Department of Education doesn't track individuals - one of only eight departments of education in the country that doesn't, Magee said.

"It's the best measure we can make based on the information our state Department of Education collects," he said.

ConnCAN got $6.5 million in the state budget for the education department to track individual students this year, but until the database is created, ConnCAN's report should not be ignored, Magee said.

"Otherwise, these urban schools are overlooked," he said. "If they generated a traditional top ten list, it would be chock full of white, suburban, middle-class schools."

The next challenge is finding out how to replicate success.

ConnCAN is visiting the top three schools in each category that have more than 75 percent low-income and minority students to determine what they are doing.

They have similar characteristics, Magee said.

"Whether charters or magnets or a traditional public schools, successful schools always have strong, effective principals with clear vision, attract great teachers who are very intense and focused on their mission, and have fostered a school culture that is about high expectations, no excuses and working together as a team to make sure every child gets a great education," he said.

Charter schools in Connecticut have an average per-pupil allocation that is about $3,000 less per child, but they outperform magnet and traditional public schools, Magee said. Part of the reason is that they have an average 18 percent more class time than elementary schools and 12 percent more than middle schools.