"If you can't even accept their basic premise, why would you think their recommendations make any sense?" she asked.

The audit was prepared by the Ohio-based audit firm SchoolMatch and was paid for by the San Bernardino County Sun.

Rialto officials said they can use some of the information about low performance to improve their curriculum. What has raised questions is the pairing of Rialto with other school districts nationwide that the audit says are demographically similar.

These districts are referred to as Rialto's "cohort group," a group of supposedly like districts that form the basis for comparison of test scores and other data. The data is used to establish the audit's "effectiveness levels" and a group average for test scores.

One of the key problems with the audit, however, is that SchoolMatch cannot define the cohort group. SchoolMatch officials could not say which school districts nationwide comprised the cohort group or even how many districts were in it.

What SchoolMatch did provide in the audit was a list Of 10 school districts nationwide that it says have demographic characteristics similar to Rialto. The characteristics include education level of school district residents, per capita income, total enrollment and the percentage of students qualifying for free or reduced-price meals.

But even here, SchoolMatch officials could not say whether those 10 districts were part of Rialto's cohort group.

Rialto school officials, however, said they question the audit's validity if those 10 districts are indicative of the school districts in the cohort group. Most of the 10 sample districts bear little resemblance to Rialto.

A review of those sample districts by the San Bernardino County Sun showed most had similar student enrollments. But key differences arise in the percentage of students who have limited English skills and the number receiving free meals, an indicator of poverty.

SchoolMatch chose not to include English proficiency as a criteria because the definitions of English-as-a-second-language students and limited-English proficiency students are not universal, said Steve Sundre, executive vice president of SchoolMatch.

"We deal with the reality of the student populations in those districts," Sundre said. "Language is not a perfect science. It's just not possible to compare them at this time." .

But Rialto officials say such information should at least be considered when comparing performance among districts.

Rialto's students speak 40 different languages, said Dorothy Spence, the district's director of special programs.

About 5,500 students, almost 22 percent, are learning English, meaning English is not their primary language. As many as 2,000 have been classified as either speaking no English or a very limited amount of English.

"It doesn't take much to know that if they don't speak the language and don't understand the words and they take the test, they're not going to do well," Spence said.

One of the districts listed as similar to Rialto, the Aiken County School District in South Carolina, has seen an increase in Latino students during the past year, spokeswoman Tammie Newman said.

But even with the increase, Aiken County's population of about 25,000 students is only 1 percent Latino, Asian and other. And only 1 percent of Aiken's students are English-learners. But the number is growing, Newman said.

"Because of migration, students have come in that spoke almost no English and now are fully entwined in a regular classroom," she said. "We have summer programs that helped just with that."

Some of the districts had less than 1 percent limited-English students.

Other differences cropped up, as well.

School districts in Kenosha, Wis., and Asheville, N.C., had roughly a quarter of their students receiving free or reduced-price meals, compared to Rialto, where more than 60 percent of students need the meals.

Such strong dissimilarities make it difficult to learn from the audit, Spence said.

"We had nothing to fear from the study," she said. "But we didn't think they picked up enough on the other factors or came up with any suggestions."

The audit's comparisons are skewed if the districts' populations aren't similar in language barriers and district spending, Newton said.

"I find tremendous fault with it." she said. "Their figures are flawed, and it damages the credibility of their report." Among the eight districts for which the data was available, Rialto's per-student spending was less than all but one - the Bay County schools in Panama City, Fla.

SchoolMatch uses a computer database of all 15,800 public school districts nationwide to generate its cohort groups for the districts undergoing the firm's audit. In some cases, only one or two schools within a district might be part of Rialto's cohort because they match demographically.

SchoolMatch began listing the sample group after requests from districts previously studied.

"Administrators would call us from across the country and say, 'I'm being compared to my neighboring school district, and I think it's not fair. It's unfair to compare me with that district because they have a different base.'

"We developed that approach and included it in the audit as a resource and only as such," Sundre said.

The audit states that officials from Rialto and the 10 sample districts can exchange information about how similar school systems "manage their resources in pursuit of effectiveness." But it does not suggest what Rialto might learn from those other districts.

Sundre responded that it is not SchoolMatch's role to outline specific steps the district should take to improve its programs.

"We are not trying to prescribe what to do to move from point A and point B. We are there to describe" he said. "For us to make a great deal of report space available to what we believe personally ought to be done may not be appropriate."

When asked what Rialto school officials could learn from the comparisons to the other districts despite the dissimilarities, Sundre said that is up to them.

"Certainly there should be something to learn from those 10 districts," he said. "What it is, we may not be sure. We're not in the business of prescribing."

Rialto Unified School District
Student enrollment	26,096
Students receiving free or reduced lunch	65.9 percent
Per-pupil spending	$4,323
Median household income	$38,608
Limited English	21.8 percent
Residents' education	43rd percentile