In a blow to the Baltimore school system's crusade to hold principals responsible for suspected cheating in schools, independent hearing officers have recommended that two administrators who were removed from Abbottston Elementary should be reinstated amid a flawed investigation and a lack of credible evidence provided by the school system.

In documents obtained by The Sun this week, the hearing officers raised critical issues about how the system went about establishing fault on the part of the school's administrators.

Abbottston Principal Angela Faltz and Assistant Principal Marcy Isaac were both found to have committed no personal wrongdoing based on evidence presented by the system to independent hearing officers they hired.  The officers found that the system not only failed to prove they were responsible for cheating, but failed to prove that cheating took place at the school at all.

Both were recommended by the officers, hired by the school system, to be reinstated, with full back pay. Ultimately, the city school board will decide whether to accept or reject the hearing officers' opinion, and whether the administrators will be fully reinstated. The parties will go before the board in June.

The Sun's initial story on the outcome of the cases can be read here.

In hundreds of documents obtained by The Sun, jarring details about the case illustrated how key players in the city and state's initial nvestigation stopped short of asserting that the school had cheated when its scores dropped drastically in 2010. The officials also reserved judgment about who or what played a part in the drops.

These conclusions were shared with schools CEO Andres Alonso early on in the investigative process, the hearing records show. But Alonso maintained that despite interviews conducted by his investigator, and a plethora of other reasons offered for the drops, data like the school's benchmark scores (pre-tests that help predict how students will perform on the Maryland School Assessments) and further statistical analysis were just as important, he said.

The system argued strongly that the school's benchmark scores were so far off, that the principals should have known that something wasn't right, and had a duty to report their suspicions.

"Dr. Alonso agrees with the view that statistical analysis cannot show causation," the officer's report summarized.

"He also agrees that benchmark testing ... does not in itself, prove there was cheating. Dr. Alonso was aware of statements from 26 Abbottston employees, all of whom denied cheating at the school. He feels he gave the statements their proper weight, particularly given that the same kinds of statements were made about another school where it was ultimately determined that cheating had occurred."

One year later, the principals' defense team pointed out, Alonso would submit the very same factors that occurred at Abbottston -- like high student mobility, teacher turnover, small testing population -- as legitimate explanations for why schools across the system may have experienced declines in a year of heightened monitoring.

The schools chief said that he also drew conclusions based on facts that no other city school experienced drops like Abbottston, and tracking how students performed before and after attending the school raised doubt about their performance while at the school.

Parents of students who attended Abbottston at the time of the alleged cheating testified that the school's intensive test preparation methods prepared their students not only to excel in testing, but in middle school.

Tonese Smith, whose son attended the school from 2006 through 2010 and offered short testimony in the case, said she was disheartened to see the case progress. She testified in favor of Faltz at the hearings.

“As far as I was concerned, it was just a lot of shenanigans,” Smith said. “There was just not enough evidence. It was rough. I don’t think I would have been strong enough to go through that.”

Her son D’Andre described how Faltz was like a mother to him, giving him money to get a haircut when he needed it, and visited classrooms to say how well they were doing.

He said his classes at Abbottston prepared him for a rigorous middle school program that he now attends at St. Ignatius Loyola Academy.

“They gave me work that they knew we were going to do in sixth grade," D'Andre, an eighth-grader, told The Sun on the phone yesterday, "so when we got there I had all A’s, and ever since then, my grades have been good.”

The glue that held the investigation together -- and eventually pulled it apart -- was an erasure analysis, conducted by a state official, that was found to be "crude," by one officer,  and "incompetent" by another.

Alonso's witnesses, including the system's chief investigator, said that the school's scores would not have been invalidated without the erasure analysis. Others said they would not have been able to reach conclusions without it. All said that in the end, they couldn't say who did the erasing.

The most scathing part of the hearing officers' report described the methodology of the erasure analysis, pointing out that the official had no training, and conducted a manual check of the booklets that was unlike any other erasure analysis done in the country. In major cheating investigations, common practice is for systems to hire outside companies to run electronic scans.

The officers also took issue with the fact that the analysis could not be authenticated because the official destroyed all pertinent documents.

The official, however, maintained that her analysis was superior to electronic scannings, because erasure marks were more clear to the human eye. Nationally, experts debate that fact, and ultimately recommend conducting both an electronic and manual analysis.

But hearing officers also criticized the fact that the official only reviewed 167 of 485 books, which they concluded was a result of "bowing to pressure to reach conclusions quickly."

The officers also pointed out that the erasure analysis was flawed in an even more "troubling way."

An email included in the records shows that before the state official conducted her analysis, she inquired about whether the city's evidence was solid enough to support what she would ultimately find.

That led the independent officers to believe that the analysis was predetermined.

"I want to be sure that MSDE is confident in the findings and that BCPS has empirical data to support my review," the official wrote. "The BIG question is -- How much evidence is enough? A sitting principal is not likely to go quietly and she may lawyer-up at the expense of the [union.]"

Alonso acknowledged during testimony that he was aware of the flaws in the erasure analysis report, and "with this knowledge, he reviewed the report again, convened a meeting of his trusted advisors, and thereafter concluded again that something 'systemic and dramatic' had happened at Abbottston," the report summarized.

The strongest opinion handed down was in favor of Faltz, who has served in the system for more than 20 years and whose name has been attached to cheating at Abbottston, a school that had been visited by U.S. Secretary of Education Arne Duncan in 2009 for its performance.

According to email documents, the city school system reviewed the school's scores before Duncan's visit, and deemed them clean. That was later found to not be the case, the system said during the hearings.

"While there is a healthy dose of speculation, suggestion and suspicion that cheating occurred in 2009, there is a lack of credible evidence that cheating actually occurred," one hearing officer wrote. "Not one witness was able to testify with certainty that they knew what happened in Abbottston. To the contrary, witness after witness stated that they did not know what happened at Abbottston.

"Not one witness was able to testify as to a specific testing protocol that Dr. Faltz violated," the opinion continued. "In fact, every witness for the CEO testified that they could not identify any testing protocol that she violated."

erica.green@baltsun.com

twitter.com/EricaG