Related articles:

42 schools fail fed assessment Career's the thing as schools evolve

Trades/Construction
Water Tec
Dispatcher Health Care
ALLIANCE IMAGING
PRN MRI TECH Health Care
RLM Services, Inc
Pharmacist Construction
ROR Construction
Residential Framing Carpenters Administrative & Professional
JEWISH FEDERATION
ADMINISTRATIVE ASSISTANT Production and Manufacturing
Industrial Tool, Die & Engineering Co.
CNC Lathe Lead Computer
Flowing Wells Schools
Computer Technician

A Harvard University study concluded that private schools perform better in 11 of 12 categories when compared with public schools, countering a U.S. Education Department report last month that suggested parity.

The study led by Paul Peterson at Harvard's Kennedy School of Government used the original data from the report issued July 14 by the Education Department's National Center for Education Statistics, and "an improved methodology" for interpreting the data, Harvard said in a statement.

Peterson and colleague Elena Llaudet "identified a consistent, statistically significant private school advantage."

The Harvard study helps counter administration critics, including teacher unions, who argued that the NCES study showed the government should give more money to public schools rather than fund private alternatives, and criticized Education Secretary Margaret Spellings for not paying more attention to it.

Spellings, who supports spending public money on private schools, "was nowhere to be seen or heard" when the NCES issued its findings on July 14, Edward McElroy, president of the 1.3 million-member American Federation of Teachers, said in a July 19 statement.

Peterson's study just two weeks later is an attempt by a "full-fledged unabashed voucher advocate" to undermine the NCES conclusions, said union spokeswoman Janet Bass.

The NCES study found fourth- and eighth-grade public school students performed comparably with private school students in reading and math when variables such as income and race were factored out. It involved comparisons of 2003 test results from more than 5,000 public schools and more than 500 private schools.

The Harvard researchers, under their interpretation of the NCES figures, found private schools performing better in 11 of 12 instances, including advantages as high as 13 "test points" in eighth-grade reading and 10 points in fourth-grade reading.

Peterson said the NCES study was flawed because of factors such as its decision to count low-income students on the basis of those who qualify for free or discounted school lunches. Many private schools don't participate in such programs, and weren't credited by the NCES with teaching their actual number of poorer students, he said.

"We don't interpret our findings as proving that private schools are better," Peterson said. "What we do show is that how your results are incredibly sensitive to the exact way you do the analysis."