State investigating two Denver schools

State officials on Tuesday opened investigations into possible cheating at two Denver elementary schools, interviewing the principals and staff at Beach Court Elementary and Hallett Fundamental Academy. Principals of the two schools were placed on administrative leave.

Denver Public Schools leaders were releasing limited information about the investigation, including the names of the schools, which have been confirmed by other multiple sources.

DPS Superintendent Tom Boasberg said district staff conducted a “very thorough” analysis of 2011 assessment data for schools across the city.

“Where that analysis raised statistical concerns, we shared the information with the state Department of Education and asked the state to lead an examination,” Boasberg said. “I want to stress that the existence of this statistical analysis does not imply wrongdoing nor have we reached any conclusions.”

The Colorado Department of Education, though its legal counsel, the state Attorney General’s office, has hired a New York-based consulting firm to assist in the investigation. DPS is footing the bill. The same firm, Alvarez & Marsal, was hired in March to look into similar concerns in the Washington D.C. public schools.

“We do feel we have a duty to look further where we saw statistically unusual patterns, and that is why we asked the state to look into those cases,” Boasberg said. “Ultimately, the decisions on any potential consequences, if any wrongdoing is discovered, is for the district.”

Parents of students at the two schools were notified of the investigation and a districtwide communication to parents went out Tuesday afternoon.

Sources confirmed the analysis of DPS test results included an examination of erasure marks on student answer sheets. Results showed the two schools far exceeded district averages in the number of wrong answers erased and replaced with correct responses.

As part of their initial analysis, district officials placed testing monitors in a number of schools during the spring administration of the TCAP state exams. Last week, when third-grade reading TCAP results were released, both Beach Court and Hallett posted double-digit declines.

The principals

  • Hallett Principal Charmaine Keeton has more than 30 years experience in education, with 24 years as a classroom teacher. Hallett is her first principalship. She has led the school since 2008-09.
  • Frank Roti has been Beach Court’s principal for a decade, coming from two years as assistant principal at West High School. Before that, he was a classroom teacher and new teacher trainer in Missouri.

Beach Court dropped 40 percentage points on both the English and Spanish-language versions of the exams while Hallett, which did not administer the Spanish-language version, dropped 12 points. Remaining TCAP results will be released in late July.

Beach Court Principal Frank Roti has led the school since 2002 and Charmaine Keeton has been Hallett’s principal since 2008. Both principals were notified Tuesday of the investigation; DPS school board members were briefed Monday afternoon in closed session.

Beach Court has been the recipient of glowing media reports and district praise since 2005, when the high-poverty neighborhood school in Northwest Denver began posting strong increases in reading, writing and math.

In 2009, the district held a press conference at the school to announce DPS’ strong state test results and to applaud the work of Roti and his staff. The school also has received national praise, highlighted at NBC’s Education Nation event in 2010.

Hallett, also a high-poverty school, is a magnet program drawing students from across the district to its back-to-basics curriculum. The school was formerly known as Knight Fundamental Academy and its program was moved into the former Hallett Elementary building in Northeast Denver in 2009.

Both schools have recorded strong gains in test results, particularly Beach Court, which saw its reading proficiency rate rise from 40 percent in 2004 to 85 percent in 2011. Hallett’s reading proficiency hit 63 percent in 2004, dropped to the 50 percent range from 2005 to 2010 and then climbed from 50 percent in 2010 to 66 percent in 2011.

Beach Court is rated on the DPS performance report card as a “blue” or distinguished school, meaning it “exceeds expectations” and ranks as one of the district’s highest-performing schools. Hallett is rated as a “green” school, or one that “meets expectations” set by DPS. Both schools are rated “performance” by the state, its top rating.

It’s unclear whether additional years will be examined as part of the investigation or whether additional schools might become involved.

“This will not be a protracted investigation,” said Jo O’Brien, the state’s assistant commissioner for testing. “The due diligence on the data, initially performed by DPS, which was very thorough and very well done, has been confirmed and added to with the resources of the state’s larger metrics and methodology … We do not expect this to be long at all.”

O’Brien said it’s not unusual for a school or district to call and ask state officials to check out a statistical anomaly in the million-plus state tests administered annually. What is unusual about the data brought forth by DPS, she said, is “a level of severity that caught our eye.”

Tuesday’s action marks the first state-led cheating investigation at a Denver school, but it’s not the first time questions have been raised about gains in DPS.

Last year, USA Today conducted an analysis of reading and math scores in seven states, including Colorado, and found 69 Colorado schools where students moving from one grade to the next posted dramatic growth, or jumps greater than 99 percent of their peers in the state. Of that total, 29 percent were in DPS. Beach Court was on the list for gains made between 2006 and 2007.

But state assessment officials admitted Colorado leaders declined to pay for erasure analysis as part of their testing contract with CTB-McGraw Hill and their own statistical analysis did not flag those schools. DPS administrator Connie Casson said then that district leaders did not conduct systemic analysis of scores, such as what was done by USA Today, for potential cheating. She said they did look into incidents brought to their attention by staff in schools or by district instructional leaders poring over results.

In March, the Atlanta Journal-Constitution published a national look at cheating and cited DPS among districts with test score results warranting a second look.

“The accuracy of our student progress data is very important,” Boasberg said Tuesday. “Families … use the data to understand how kids are doing, and how much progress they’re making. Teachers use the data to inform their instruction, to know what to focus on, to know how to target their teaching, and therefore it’s very important that we have completely accurate information about how our kids are doing.”

Van Schoales, executive director of A+ Denver, a citizens advisory group to DPS, said the possibility of cheating is “incredibly disappointing and sad for the kids and families of the schools, if true.”

“It suggests that kids have a certain level of knowledge and skills that they don’t have,” Schoales said. “If you’re told in elementary school, you’re a good reader and writer and mathematician and you switch into another school and all of a sudden your scores drop, you could draw all kinds of conclusions that may not be right about why that is. The real reason why is because you don’t know those things in the first place.

“If you’re not self-aware about what you know and can do … then you’re really not in a position to get any better.”

Higher test scores can mean more money in Denver, where a performance-based system known as teacher and principal “ProComp” awards bonuses based on student growth and performance on state exams.

For example, a teacher enrolled in ProComp this year could earn bonuses topping $2,000 each if their students exceed district expectations on state exams or if their school is designated as a “high-growth” school or a “high-performing” school on the district’s annual report card.

Statewide, the full implementation of Senate Bill 10-191, the Great Teachers and Leaders Act, in 2014-15 will link student test scores with decisions about teacher and principal pay, retention and dismissal.

Because of those added consequences, as well as the state’s accountability system, which also relies heavily on the exams, O’Brien said the Department of Education this fall will debut enhanced test security policies.

Beach Court test scores

Hallett test scores