The State Office of Education knew that would be the case in Utah's smallest schools. So it devised a plan: Apply a statistical "confidence interval" to scores to ensure they are sound.

A confidence interval can be likened to an error margin, Moulding said. The smaller the school, the wider the margin.

The idea is to make it so schools aren't identified as failing to meet standards based on statistical chance alone, and so a school's performance isn't placed on the shoulders of a single student.

But this year, the rule did give some schools wiggle room.

"The confidence interval was significant in many schools having made AYP," said Laurie Lacy, No Child Left Behind specialist at the State Office of Education. "It's just right there."

For example, schools are supposed to have 65 percent of students scoring as proficient on the language arts test.

If a group has just 10 students in it, that means seven of them have to score as proficient under the state standard.

But the confidence interval makes it so as few as four could score as proficient, according to data compiled by the State Office of Education, and the state would deem that close enough.

That happened at Jordan Resource elementary, a Jordan District school for special-needs students. The school got a "pass" in language arts because just three children took the test. In math, 11 students took the test, and 55 percent scored as proficient. The standard is 57 percent.

The confidence interval narrows for larger groups.

Let's say 100 students are taking the language arts test. Though the state standard remains 65 percent proficiency, the school could have as few as 54 percent of students scoring as proficient and be OK. But a group of 400 students would have to have more than 59 percent of students scoring as proficient to meet the goal.

That confidence interval played out in the favor of some rural schools, Moulding said. But the state had few better options. It could have required a school to have, say, 80 students in each testing group in order to count for academic achievement. But that would make it so several rural schools wouldn't have to file reports at all.

Also, the confidence interval won't help schools as much once the achievement bar starts moving, Moulding said. Every two years, more students are expected to score as proficient until 2014, when every Utah schoolchild will have to be proficient.

In that sense, the confidence interval should be seen not as a free pass, but an invitation to get cracking.

"If a school falls within the confidence interval," Moulding said. "They need to pay attention."

---

Contributing: Laura Warner

E-mail: jtcook@desnews.com; terickson@desnews.com