Practice PARCC test books sit on a table in the sixth grade English Language Arts and Social Studies classroom at Morgan Elementary School South in Stockport, Ohio. (Ty Wright/AP)

As the first wave of new Common Core test scores trickle out, Ohio policymakers did something that threatens to undermine the reason why Ohio and dozens of other states agreed to a set of national K-12 academic standards in the first place.

Ohio officials opted to interpret their test results differently than other states, inflating the performance of their students and making it tricky to compare Ohio with the rest of the country.

It is the latest sign that a central promise of Common Core — that states would have the same math and reading standards and use the same tests — is further unraveling.

Some states have announced they are withdrawing from the communal academic effort and will craft their own tests from now on, making future interstate comparisons nearly impossible. And in a move that could be emulated by others, Ohio has decided that students don’t have to actually meet the common college- and career-ready standards in order to be deemed proficient — they just have to “approach” those standards.

The result, Common Core opponents and even some boosters agree, is another step toward a return to the kind of gamesmanship with statistics — and exaggerated sense of student achievement — that the program was designed to eradicate.

And that’s discouraging for people who advocated for Common Core, said Frederick Hess, director of education policy research at the American Enterprise Institute. “Part of the reason they wanted the Common Core was they were worried about a race to the bottom, where you would juice up your proficiency rates because you didn’t want your neighboring state to look so much better than you,” Hess said.

Now Ohio Gov. John Kasich — a candidate for the GOP presidential nomination — can boast about Ohio’s student achievement, Hess said. “And the others are wondering ‘Why am I a sucker?’ Once the second governor follows, and the third and fifth, that’s how you get back to the race to the bottom,” he said.

Millions of students in more than half the states last spring took one of two new Common Core tests, PARCC (Partnership for Assessment of Readiness for College and Careers) and Smarter Balanced, that were developed with the help of $320 million from the Obama administration.

The new tests were widely expected to be more difficult than most states’ old tests, and proficiency rates were expected to be far lower. That has proven true, according to results released so far.

On California’s old test, for example, more than half of the state’s students scored high enough to be considered proficient in reading and math; on the new Smarter Balanced exam, just 44 percent were proficient in reading and 33 percent were proficient in math.

Common Core supporters say the new tests are giving parents and policymakers a more honest view of whether students are adequately prepared for college and careers after high school. Compared with the country’s only nationwide test, the National Assessment for Educational Progress, most states have a low bar for proficiency in math and reading. Ohio’s is among the lowest.

About 80 percent of Ohio students were proficient on previous state math and reading tests. But only about 40 percent scored high enough to meet PARCC’s national definition of proficient, according to preliminary results released by the state department of education.

According to Ohio’s interpretation of its results, however, 65 percent of its students are proficient.

“We’re sort of easing into this,” Jim Wright, of the Ohio Department of Education, told the state board of education on Sept. 15. “There has been a lot of press out there saying there’s going to be a cliff that we fall off of,” he added, referring to the expected test-score drop. By choosing to define proficiency differently, he said, “there's not that precipice.”

The decision’s main impact is on how Ohio’s student achievement is perceived publicly. Because of the shift to Common Core, test scores from 2015 and 2016 will not be tied to consequences for schools or teachers in most cases, according to state education officials.

Critics of standardized testing point to Ohio’s move as evidence that deciding how to score exams is an inherently subjective decision.

“There is no independent evidence that ‘proficient’ is the level that’s required to perform adequately in college or careers,” said Robert Schaeffer of the National Center for Fair & Open Testing. “The fact that PARCC’s new levels differ from Ohio’s doesn’t mean that either one is correct. . . . The way achievement levels have been set is a bunch of guys and gals sitting around a table. And it’s always been a profoundly political and ideological process.”

PARCC divides students into five performance levels, from 1 to 5, and to be considered proficient, students must score at least a Level 4: “Met expectations.”

But Ohio education officials decided that students who score lower, at Level 3, also will be considered proficient.

State officials said they are not trying to inflate student performance. Instead, they said they had no choice because a long-standing state law requires tests to divide students into five performance levels and spells out what they must be called. Under that state law, “proficient” is the name that must be used for students who score at Level 3.

The state plans to report how many students scored in each of the five performance levels, making comparisons possible with other PARCC states. “Ohio has not lowered the bar in any way,” said Joe Andrews, press secretary for Kasich, who has been a strong supporter of Common Core.

But the dueling interpretations are likely to be confusing for many parents and voters.

Parents will receive a report card that tells them how their children fared according to PARCC, and how they fared according to Ohio. A third-grader who answered one-third of math questions correctly? “Proficient,” according to the state, but only “approaching expectations” according to PARCC.

“If you are a parent, what are you supposed to believe?” said Karen Nussle, executive director of the pro-Core Collaborative for Student Success, who has been critical of what she calls a lack of transparency on Ohio’s part. “It’s going to be super confusing.”

Of the states that administered PARCC in the spring, Ohio and Illinois have released preliminary results. The rest plan to make them public throughout the fall.

Most PARCC states say they will use the achievement levels set by PARCC, but a couple, including New Jersey, have yet to decide.

The Common Core standards and tests have become a political lightning rod during the past two years as many conservatives have come to see them as a symbol of federal overreach. The movement to the new standards was state-led, but the Obama administration offered powerful incentives — including through the Race to the Top grant competition — for states to buy in.

Ohio already has decided to drop out of PARCC next year and craft its own exam. Three other states have done the same, leaving just seven states committed to PARCC in 2016, down from 26 in 2010. Fifteen states will administer Smarter Balanced next year, down from 18 this year.

With such low participation, meaningful comparisons across state lines will be hard to make: Just 22 states will give one of the two main Common Core tests next year, leaving 28 that will go their own way.

Common Core supporters say the country is still moving in the right direction overall: Despite the attrition, more children are taking common tests than ever before.

“We’re getting more honest,” Nussle said. “We have more transparency than we’ve ever had.”

But is there still a concern that states are finding ways to hide the truth about their students’ performance on tests?

“I think we’re always going to be battling that,” Nussle said. “It takes political courage, and there’s not a lot of that these days.”