"They don't believe in testing students andthey don't believe in testing teachers," she says.

But Fowler says it is the MECT specifically,and not testing as a concept, that he objects to.

"I support having a test for prospectiveteachers. It's not an unreasonable thing to askthat teachers have basic skills," he says.

Fowler, along with the report's co-authors,points to NES's lack of pre-testing or externalreview as its major problems.

Problems with the NES

While NES had begun developing the test inadvance, they still had less than five months tofinalize the exam, which was to run in a trialstate for the initial April and July sittings.

But because of Massachusetts politicking, theNES was given only five months to create the test.Further, the April and July administrations becameactual exams, not trials.

Across the nation, Massachusetts has become thecase study of what not to do in teacher testing,says Gene J. Campbell, director of licensing forthe Colorado Department of Education.

He says he believes politics unduly influencedthe process of testing and worries whether theintegrity of the test was compromised.

"The Massachusetts experience is well-knownacross the country. The unfortunate thing is thatit has put a shadow of a doubt on [teacher]testing across the nation," he says.

Campbell's state of Colorado, which has usedNES since it initiated teacher testing in 1994, isone of NES's larger clients.

While the NES-administered teacher tests havehad a successful start in Colorado, Campbell saysthe development of the Colorado test in 18 monthswas "tight."

Even considering the 18-month time frame,Campbell said there were logistical problems.

"I don't think, considering the size of ourtask, I'd feel comfortable at all [with fivemonths]," Campbell says of Massachusetts' timeframe. "I am not sure, given the time constraints,[NES] could have avoided these problems."

But in the report, Fowler and his co-authorsalso bring up the questionable work of NES inconducting a state teacher test in Alabama.