PHOENIX -- Preliminary data released by the U.S. Department of Education last month showed that only 25 percent of English Language Learners in Arizona high schools graduate in four years.
That number was well below Arizona's overall 78 percent four-year graduation rate.
It was also the lowest number reported in a spreadsheet of national data.
The stat came with a press release in which the federal agency claimed the data reflected "the first year for which all states used a common, rigorous measure" for graduation rates.
That prompted me and other reporters to begin working on stories about Arizona having the worst ELL graduation rate in the nation.
Yet, after a bit more digging, it turned out the graduation data for non-English speaking students, or Limited English Proficient (LEP) students, as they are called by the federal government, (a.k.a ELL students, here in Arizona) isn't actually comparable across states.
That's because each state got to choose how to define its LEP cohort. So while California's rate of 60 percent of LEP students graduating in four years looked much better than Arizona's rate of 25 percent, California crunched its numbers differently.
The Golden State included any student who spent a year classified as LEP at any point during high school in its tally, while Arizona only counted students who were classified as ELL in their final year of school.
This is just one example in many of why it is difficult to measure the success of teaching English as a Second Language, and compare it across states.
But what do we know about Arizona's efforts to educate non-English-speaking students, and how it is faring?
Major changes were made to the program in recent years (partly in response to a long-standing class action lawsuit), and most districts fully implemented the current ELL model in either 2007 or 2008.
A main component of the current program is the use of a four-hour mandatory block of English instruction. While defenders of the program, including state superintendent John Huppenthal, have called its quality "the best in the nation,” the program also has vocal critics.
Elements of the program -- including how students are classified as ELL, how they test out of ELL and the separation of students in the four-hour blocks -- are the subject of an ongoing Department of Justice Civil Rights Division investigation. The model has also drawn scrutiny from some education experts and at least one member of the state's English Language Learner Task Force that originally directed how the program should be implemented.
In terms of how to measure the ELL educational success, one metric is how fast ELL students are reclassified into mainstream classrooms.
In Arizona, more than 30 percent of ELL students reclassified into mainstream classes last year. That's higher than most states, and a point of pride for those who defend the state's program.
But critics say that many ELL students here are reclassified before they are English proficient, because the testing policies in place are flawed.
Federal civil rights investigators agreed with that analysis. They found that tens of thousands of ELL students may have tested out of ELL classrooms prematurely, prompting a settlement with the state of Arizona that involves both a new test and additional reading and writing services for former ELL students.
So what about ELL graduation rates since the new program came into effect?
A request for data from Arizona's Department of Education on ELL student graduation rates over the past several years has not been answered yet, but data from two of the state's biggest districts provides an interesting snapshot.
The four-year graduation rate for ELL students at Tucson Unified School District fell from 67 percent in 2008-2009, to 36 percent in 2011-2012.
At Phoenix Union High School District over the same period, the ELL student graduation rate fell from 51 percent to 22 percent.
Of course, there are a few different answers out there to explain those numbers, depending on who you ask.
For one, passing Arizona's standardized test, the AIMS, has become an even bigger part of graduation requirements over this time period. That has some ELL students needing to take extra time beyond four years to study for the test.
Another explanation, according to critics of Arizona's current ELL program, is that Arizona's current ELL model is hurting -- not helping -- graduation rates.
"The schools don't have the flexibility to deal with the students individually," ELL task force member Johanna Haver told me. Haver is a former educator and expert in English as a Second Language instruction. "They get stuck in four hours of English-language development year after year."
Sal Gabaldon, a language acquisition specialist at Tucson Unified School District, told me he believes while some students tested out prematurely from ELL instruction, others languish in the mandatory four-hour English block, leading to his district's declining ELL graduation rates.
"For English-language learners who are at the high school level, and even in middle school, it has been difficult for them to participate in four hours of English-language development and still try to find enough classes during the regular school day to keep up with the minimum requirements in content classes for graduation," Gabaldon said.
At Phoenix Union High School, spokesman Craig Pletenik said high school credits needed to graduate have increased from 20 to 22, making it that much harder for students in the four-hour English block to get the credits they need.
Johanna Haver has proposed that schools should be able to offer students who completed one year in the four-hour block but have not yet tested out of ELL, to switch to a two-hour English language block so they could access other classes.
But her colleague Alan Maguire, who chairs the state ELL task force, is a defender of the program. He told me the current state model is yielding results, and the graduation rate will reflect that over time.
"The new models are a dramatic change for teachers of English Language Learners in Arizona," Maguire said, adding that there have been successes so far, including test results that show a portion of former ELL students in younger grades are outperforming native speakers on the state AIMS test. "But it's a big change, and big changes take time."
And when it comes to those immigrant students who are new to English when they enter high school, Maguire said a diploma after four years may not be the goal.
"We have to accept the fact and be compassionate for our newly arriving ELLs that they have to learn more than a native speaker arriving at the front door of a high school as a freshman," Maguire said. "So it is going to take them longer to do that. We have to tell that is OK. We have to help them succeed."
But what measure of success will all sides agree on?