S&P Opens A Rating Service On Schools
Site Offers Public Comparative Data

By Jay Mathews
Washington Post Staff Writer
Wednesday, March 30, 2005; Page B01

With the launch of a nationwide Web site yesterday, parents in the District, Maryland and Virginia will have access to the kind of detailed in????ation about their public schools that investors have long had about Fortune 500 companies.

The free Web site, SchoolMatters.com, was developed by financial data giant Standard & Poor's and offers a searchable collection of education data, including per-pupil spending, student per????ance and classroom size.

Montgomery County parent John Hoven says the Web site needs work but is a promising resource.

Standard & Poor's got the idea for compiling the data when it noticed that some school districts were saying the company's bond ratings proved that they were doing a good job in the classroom, said Abby Potts with the Council of Chief State School Officers, one of the sponsors of the Web site. The company began to offer the education data as a better measure of teaching and learning. It also established firewalls between its school ????uators and its bond raters so that they could not influence one another, Potts said.

Early reaction from some Washington area parents suggests that the site is giving people what they want.

"I'm always in favor of giving people more in????ation," said Dick Reed, a ????er PTA president at Fairfax County's Edison High School. "The more the better, and this site does that well -- both by providing in????ation that parents, in particular, have trouble getting at all, and by providing that hard-to-get in????ation in an easily read and understood fashion."

But some school district administrators said some Web site numbers are wrong, out of date or easily misinterpreted. Sharon Ackerman, assistant superintendent for instruction in the Loudoun County schools, said staffing trends and class size numbers for 2002 to 2003 were out of touch with reality. One page of the Web site said some schools in Loudoun averaged 132 students in each classroom, she said.

"This site could be useful for parents as a starting point to find in????ation about past per????ance of students in specific schools," Ackerman said. "However, the data must be accurate."

Web site officials said the class size and other numbers came from federal education agencies.

The Web site is run by the Education Data Partnership, a collaboration that includes the Council of Chief State School Officers, Standard & Poor's, the nonprofit group Achieve Inc., which manages state education standards, and the CELT Corp., a technology company. The work is supported by grants from the Broad Education Foundation and the Bill & Melinda Gates Foundation.

The site provides test score results by race and also by economic background. It also shows the differences in how school systems spend tax dollars and allows residents concerned about how money is spent to see what portion of new revenue coming into each district is spent on instruction.

The site calls this the Instructional Spending Allocation Index, which measures the proportion of increased spending over time allocated for instruction and provides a way to track money raised with the intent of improving student per????ance. The portion of new dollars for instruction in Washington area school districts in 2002 ranged from 104.8 percent in the District, which Web site officials said spent all of its new money and then some extra from other sources on school per????ance, to 52.6 percent in Prince William County.

That index and other data developed by Standard & Poor's should be handled with care, the Web site says. It includes a warning from ????er North Carolina governor James B. Hunt, one of the leaders of the national school improvement movement and a member of a Standard & Poor's advisory board, that "these ratios should not be used alone to draw conclusions about education per????ance."

Kenneth Bernstein, a teacher at Prince George's County's Eleanor Roosevelt High School, said he thought that statement odd. "For all the warnings by Hunt and others not to use the data for comparisons, what do they expect, when the only really new thing they offer is precisely that data?" he asked.

Officials from Standard & Poor's School ????uation Services said they first tried out the data collection and presentation system in Michigan and Pennsylvania, and some school district leaders were not happy being identified as spending more per pupil with less impressive results than their neighbors. But some educators said the in????ation can help them focus their resources where they are most needed.

"By using SchoolMatters to identify schools with successful practices, principals can adjust instructional methods to further student achievement and help drive overall school improvement," said Brian Glades, principal of Fisher Elementary School in Redford, Mich.

Montgomery County parent John Hoven said the Web site was on the right track. "It doesn't deliver what it promises, but it could easily do so," he said.

He said the Web site recognizes that any fair comparison of schools must account for differences in student demographics, and it shows how easily this can be done with a simple graph called a scatterplot, which illustrates various patterns and relationships. But the Web site does not yet include a procedure for users to create their own scatterplots to compare school districts or ask such questions as whether higher spending, more rigorous standards or smaller class sizes raise student achievement.

Web site officials said that because Maryland is using new state tests, it is more difficult to get a true sense of the improvement in its schools over time than it is in Virginia and the District, which have been giving the same annual tests to all students for several years.