None of the 12 textbooks has an acceptable level of accuracy, according to N.C. State physics professor John Hubisz, the report's author.

"These are terrible books, and they're probably a strong component of why we do so poorly in science," on standardized tests, he said.

"The books have a very large number of errors, many irrelevant photographs, complicated illustrations, experiments that could not possibly work, and drawings that represented impossible situations."

The study was financed with a $64,000 grant from the Lucille and David Packard Foundation.

Liberty’s Torch in Wrong Hand

Among the books included in the study was a multi-volume Prentice Hall series called "Science," which has been used by several North Carolina school systems.

Errors in some editions of that series, according to Hubisz, include an incorrect depiction of what happens to light when it passes through a prism, a reversed photo of the Statue of Liberty showing the torch in the wrong hand, and the Ronstadt photo.

Prentice Hall acknowledges some errors, partly because states alter standards at the last minute and publishers have to rush to make changes.

"We may have to change a photograph because of a new content objection, and the caption isn't changed with the photograph," said Wendy Spiegel, a spokeswoman for Prentice Hall's parent company, Pearson Education. "But we believe we have the best practices to ensure accuracy."

Last year, the company launched a thorough audit of its textbooks for accuracy and posted a Web site with corrections, she said.

Five Hundred Pages of Errors

Hubisz enlisted a team of researchers, ranging from middle school teachers to college professors, to review the 12 books for factual errors. The researchers compiled 500 pages of errors, which were boiled down to a 100-page report.

"These are basic errors," he said. "It's stuff that anyone who had taken a science class would be able to catch."

One textbook even misstates Newton's first law of physics, which has been a staple of physical science for centuries.

Hubisz, who has received requests for copies of his study from as far away as Japan and Scandinavia, called Glencoe/McGraw-Hill books "the best of the worst."

The worst of the worst?

"Probably Prentice Hall," he said Sunday.

Teachers Under-Trained, Expert Says

Teachers, administrators, parents and curriculum specialists typically review books before they are used in a classroom. In North Carolina, a state committee approves a list of textbooks for the public schools. Each school system then picks its books from that list.

But Hubisz, president of the American Association of Physics Teachers, said many middle-school science teachers have little physical science training and may not recognize errors.

Also, many states - and local school districts within those states - tend to follow the lead of state officials in Texas, California and Florida, the three biggest textbook purchasers, Hubisz said.

"We estimated maybe 85 percent of children in the United States probably use these books," he said.

Who Writes These Books?

The study's reviewers tried to contact the authors with questions, Hubisz said, but in many cases the people listed said they didn't write the book, and some didn't even know their names had been listed. Some of the authors of a physical science book, for example, were biologists.

Hubisz said educators need to pressure publishers to get "real authors" for textbooks.

"We're really trying to get the publishers to do something," he said. "They get people to check for political correctness; … they try to get in as much cultural diversity as possible. … They just don't seem to understand what science is about."

Hubisz said the study panel contacted publishers, who for the most part either dismissed the panel's findings or promised corrections in subsequent editions.

Reviews of later editions turned up more errors than corrections, the report said.