State to review 2012 test scores from Trace Crossings Elementary in Hoover due to concerns with irregularities

Trace Crossings Elementary School

Trace Crossings Elementary School in Hoover, Ala. (Frank Couch/fcouch@al.com)

HOOVER, Alabama – State schools Superintendent Tommy Bice said he plans to have his staff take a deeper look at a significant drop in test scores at Hoover's Trace Crossings Elementary School in 2012 due to concerns of potential irregularities.

Trace Crossings was one of three Hoover schools that failed to make "adequate yearly progress" on state accountability tests in 2012, but a former principal there said she believes some of the 2012 scores are faulty due to a test administration error.

Robin Litaker, who was principal at Trace Crossings from the summer of 2010 to November 2012, said she believes fourth-grade students in the spring of 2012 were not given enough time to take the first part of the math portion of the Alabama Reading and Mathematics Test.

The percentage of Trace Crossings fourth-graders who were deemed proficient in math fell from 89 percent in 2011 to 54 percent in 2012.

Litaker and other school officials had expected some drop in scores in 2012 but not that much. Students statewide were taking a new version of the test, and Trace Crossings in particular had gained a lot of students from lower-performing school systems, Litaker said. Some of those students were way behind grade level and struggling, she said.

But the 33-percentage point decline in proficiency was jaw-dropping. Litaker said she logged in and out of the computer program several times to make sure she was seeing the correct scores. "I realized something wasn't right," she said. "I was just baffled."

The same general group of students showed a 75 percent proficiency rate when they were in the third grade.

Litaker said she told Hoover schools Superintendent Andy Craig about the test scores within 30 minutes of seeing them on July 3, 2012. She and Deborah Camp, the district's curriculum director at the time, called in a testing expert from the University of Alabama at Birmingham to analyze the scores.

"He said there is no statistical reason for those scores to have changed this much," Litaker said. "These kids could have stayed home all year long and their test scores would not have dropped this much."

Another test the same students had taken at the end of the previous year indicated that 83 percent of them were proficient in math, and some of the math skills areas where they scored the lowest on the ARMT had been considered strengths in other assessments, Litaker said.

The week after the test scores came back, Litaker called in the teachers, even though it was summer. After working so hard to help students improve, they were distraught, she said. "They cried. They were visibly a wreck."

That's when the teachers told her that not one of the 107 children in the fourth grade had finished Part 1 of the math portion of the test, Litaker said. It's normal for a few students not to finish a standardized test, but for no students to finish it, "that is next to impossible. That does not happen," she said. "Something happened at Trace where all five teachers did not give the kids enough time to take the test."

A test administration document shows that fourth-grade students were supposed to have 20 minutes to finish Part 1 of the math test, and an additional 10 minutes if necessary, but students were not to have more than 30 minutes.

Litaker said she does not know how much time the students actually were given, but she is convinced it was not enough.

Camp on Oct. 10, 2012, wrote a draft of a letter for Craig to send to the state school superintendent and Gloria Turner, the state's director of student assessment at the time, asking that the completed tests be reviewed for inconsistences such as erasures, multiple responses or incomplete forms.

The superintendent's assistant replied via email on Oct. 12 that the letter had been scanned and posted to Bice and Turner. However, state officials say they don't recall ever receiving the letter and don't have a record of it.

Hoover's assistant superintendent for instruction, Ron Dodson, took over Hoover's investigation of the issue from Camp about the same time the letter supposedly was sent. Dodson said he's not sure what happened to the letter.

"Andy's not sure if he ever sent it. He doesn't remember," Dodson said. "It was two years ago."

Nevertheless, the state was aware of the issue. Dodson contacted Turner with the state, and Turner agreed for the state to rescore 10 students' answer sheets. The state found that the answer sheets were scored correctly and that none of those 10 students had completed the test.

For Litaker, that was confirmation that incorrect timing likely was the cause of the lower test scores. She wanted to press for all the students' answer sheets to be reviewed to verify that none of the fourth-graders had finished the test. Litaker said Dodson at first told her he would talk with Turner about the issue, but when she inquired again, Dodson told her that it was too late to file an appeal and that she should drop the matter and let it go.

Litaker, who was removed as principal at Trace Crossings in November 2012 and reassigned to other administrative duties, in April 2013 decided to retire, effective Dec. 31, 2013.

Litaker said she was sure that Trace Crossings' test scores would rebound significantly in 2013, and she was right.

The percentage of fourth-graders deemed proficient in math jumped from 54 percent in 2012 to 87 percent in 2013 – a 33 percentage-point increase. "Test scores don't fluctuate 33 points. That's statistically impossible," Litaker said.

To her, it was further proof that the 2012 fourth-grade math test scores at Trace Crossings were faulty and need to be voided.

Dodson isn't convinced.

He notes that no one at Trace Crossings reported substantive problems in the administration of the tests at the time they were given. Instead, Litaker and Camp suggested that perhaps not all of the teachers were following the state's newer math curriculum completely, he said.

New training was provided for all the teachers in the school, and a monitoring plan was put in place that included weekly walk-throughs by district officials during math instruction and teachers' planning periods.

Those walk-throughs confirmed the general impression that many of the teachers were initially unfamiliar with the newer math curriculum and instructional routines that are necessary in an inquiry-based learning program, Dodson said.

However, the teachers were eager and committed to implementing the new curriculum, and the ongoing presence and support of a new math coach that Litaker had requested began to bear fruit as new instructional routines became familiar to both students and teachers, Dodson said.

"I really did feel that there was a need to improve math instruction," he said.

The 2013 test scores were an indication that the improvement plan achieved its desired result, he said. "I felt like we were addressing the root problem," he said.

He did not – and still doesn't -- see any need to dig further into the 2012 test scores, he said. "It would not have changed the outcome at all," he said.

Incomplete answer sheets are still valid answer sheets, Dodson said. "Even if the state had determined that the answer documents were invalid, that would not have changed the accountability status because the invalid results would have still led to a failed AYP status," Dodson wrote in answers to questions about the matter. "There would not have been a retest."

Dodson said people can take the lower test scores as an indictment or take them as an opportunity to get better. He believes they did the latter, and it has paid off.

Litaker said she understands that the AYP status can't be changed, but the ARMT test is supposed to reflect what the children know and are able to do, and she believes the 2012 results were not an accurate assessment.

Significant impact

The fallout from the low test scores has been significant, she said.

Teachers were unfairly humiliated, and the school's reputation took a big hit in the community, she said. Parents expressed concern about whether their children were getting an adequate education and pressed the school board to redraw attendance zones so that Trace Crossings would not have such a high percentage of low-income students and students who had moved into Hoover apartments from lower-performing school systems.

Some Trace Crossings families moved, and others put their children in private school. Enrollment at Trace Crossings dropped from 520 in the fall of 2012 to 488 in the fall of 2013 and 439 now, despite being in a high-growth part of Hoover with new housing construction. Some residents in the community expressed concern about property values declining.

Litaker said if it's true that those students' scores dropped so much because they weren't given enough time to take the test, then they deserve to have those scores removed from their records, the teachers deserve some exoneration, and the community deserves to have the record set straight, she said.

"They deserve better than to think their school and the teachers in that school were failing them," she said. "That community deserves to believe in their school. That is a good school."

The issue could be investigated very simply by looking at all the students' answer sheets, she said. She doesn't understand why Hoover's top school administrators and the state have refused to do so, she said.

Litaker talked to the state school superintendent personally about the matter in November 2013, but months went by without anything more being done. In June, Alabama School Connection Executive Director Trisha Crain, a Hoover resident, inquired with Bice about the issue. Bice made a personal visit to Hoover schools Superintendent Andy Craig that month to look into the matter and seemed satisfied with the response he was given.

Contacted again recently, Bice acknowledged the unusual nature of a 33 or 35-percentage-point change in test score proficiency in one year. He said state officials had looked at a sampling of the test answer sheets and "we felt it not to be as big of an outlier as some people see it to be."

However, those test answer sheets are still in storage, and "we'll be glad to look deeper," Bice said. "If there are specific questions that need to be answered, then I will entertain those."

Bice said it's important that there be some benefit to students in such an investigation. However, "I'll ask my folks to look at this again."

See more news from Hoover at www.al.com/hoover