Michelle Brandon didn’t know what a “VAM score” was before it upended her classroom just two weeks into the school year and forced her to change jobs.
Despite two years of “highly effective” job ratings from her principal at Hudson Elementary in Pasco County, the score had reduced her teaching skill to a number, which tagged her as sub par. That triggered a Florida law requiring struggling schools like Hudson to shed most, if not all, of their teachers with low scores.
Three of Brandon’s fellow teachers were told they had to leave the school for the same reason. When they broke the news to students, “there were a lot of tears,” Brandon recalled. “Ours. Theirs. Boys we didn’t expect to see any emotion from. It was very difficult.”
A similar scenario played out last month at dozens of D- and F-rated schools across Florida, drawing widespread protests from educators.
While the law no longer requires the use of VAM scores for teacher job evaluations, the scores live on in another part of the state’s school accountability system. They are used to determine which teachers can and can’t work at “turnaround” schools.
Critics say the scores are creating a logistical mess in addition to being a poor measure of teaching ability. The state releases them in August, forcing many teachers to transfer just as the school year is starting.
“One of our schools was left without a regular fifth-grade teacher,” said Polk County School Board member Billy Townsend, who called the state’s approach disruptive.
Polk recently reassigned 18 teachers because of their VAM scores, just as the 2019-20 school year started. Two years ago, 48 teachers were forced into the last-minute switch.
“VAM matters in the turnaround world more than anything else, which of course you’re dealing with the most vulnerable kids, which creates the most chaos,” Townsend said.
The letters stand for value-added model, a statistical formula that attempts to quantify how much a teacher contributes to students’ academic growth in reading and math.
Previous state test scores and other factors are used to create an average “predicted score” for each teacher’s students. That score is compared to the latest test scores to see if students performed better or worse than expected. The difference is the VAM score.
One of the more controversial elements of the formula has been how it accounts for teachers who don’t teach reading or math, or whose students are too young to take state tests. In those cases, the VAM score is based on a school’s average scores. In other words, it judges teachers based on factors beyond their control.
The goal is to ensure students in those most needy schools don’t end up with teachers who aren’t helping them.
“Institutionally, I think you want your best teachers teaching at your toughest schools,” said state Sen. Manny Diaz Jr., a former district and charter school administrator who chairs the chamber’s education policy committee.
That’s not a principle the Legislature will abandon, he said. But the state could discuss how it uses VAM scores, including the timing of their release.
“We have been trying to discover the best way to evaluate teachers. It’s been something that has eluded us for a long time, because it’s not an exact science,” Diaz said. “Those issues are always issues we should be willing to take a look at.”
One possible place to start, he suggested, could be the way the teachers’ scores get distributed. Having them arrive after classes have begun leaves schools no room to make changes without disturbing learning.
“Movement at that time of year can be disruptive,” Diaz said. “Anything we can do to improve that, I would be in favor of.”
But the problems go beyond timing the score release, argued Audrey Beardsley, a professor at Arizona State University’s Mary Lou Fulton Teachers College.
Beardsley calls herself an “anti-VAMer” and provides free expert advice to teachers who have sued their school systems over the model.
She reasons that it proves nothing about a teacher’s abilities, and does not meet nationally accepted standards for research and measurement.
It is not fair, reliable, valid or transparent, she continued, so it should not be used in high-stakes decisions such as whether a teacher can keep his or her job.
She noted that anywhere from 25 percent to 60 percent of teachers will experience “extreme” fluctuations in their results from one year to the next.
For example, two Pasco County teachers forced to relocate because of VAM in past years, saw their three-year average score rise from “unsatisfactory” or “needs improvement” to “effective” in just one year.
That doesn’t make for a trustworthy valid system, Beardsley suggested.
She also pointed to disconnects between a low score and other indicators, such as positive observations by principals, parents and other educators. That type of evidence also could suggest a lack of validity in the model, she said.
Another concern she mentioned was teachers’ lack of understanding of the formula. Many don’t know what to do to improve after they get those scores, and receive no guidance.
Teachers in Houston won a lawsuit killing VAM over that latter issue, Beardsley said.
“I wish that (Florida teachers) would fight this,” she said.
Florida teachers did sue to stop VAM, but lost their case in 2015. The U.S. Court of Appeal found the Legislature had a rational basis for establishing the policy – a position that Beardsley suggested could be overcome with a more targeted complaint.
Hillsborough County teacher Erica Driver meets many of the criteria that Beardsley tagged as ripe for a challenge.
A teacher at Tampa Heights Elementary, Driver was identified for a transfer based on her low math score. But her principal didn’t want Driver, a school leader with a highly effective overall evaluation, to leave.
At the same time, Driver did not know that VAM scores could affect her job assignment, and she had no information on how she might fix her situation.
“I never got anything from the state saying I was in the ‘needs improvement’ category, and this is what I needed to do to get out of it,” she recalled. “I have no idea how it works. I have looked at the formula. Unless you are a mathematical engineering genius, I don’t see how you can figure it out.”
Hillsborough superintendent Jeff Eakins said he has negotiated with the Florida Department of Education to work around the restrictions. In some instances, the district has moved teachers to different grade levels that are not impacted by the state score.
In others, it has hired different teachers of record for students, while allowing the ones with low VAM marks to remain as classroom support while getting added training.
In still others, the district has reassigned the teachers into positions where they do not teach the subjects where they had a low score.
“The state VAM score is a single data point … that sometimes doesn’t give you a full picture of the teacher,” Eakins said.
Driver fell into the last category. She teaches all language arts courses at Tampa Heights, avoiding the math that vexed her state rating.
But she’s not happy about it.
She argued that if the state no longer requires districts to use the value-added result in performance evaluations, the state shouldn’t use it at all. She also wanted to teach one set of fifth-graders, not two.
Worst of all, Driver said, her morale is sapped. For the first time in her career, she questioned whether to remain a teacher.
“All year long we tell the kids, ‘It’s just one score. It doesn’t define you,’” she said. “Then the state uses that one score (on teachers). … It’s an emotional roller coaster.”
Pinellas Classroom Teachers Association president Mike Gandolfo acknowledged it might sound logical in theory to remove teachers with a low score from a high-needs school.
But that assumes that the value-added model has value, Gandolfo said. And he doesn’t buy that.
Rather, he suggested the model – though touted as taking into account factors outside a teacher’s control – lacks the ability to recognize how children’s past teachers did or did not prepare them.
It can’t accommodate for the youngsters who missed breakfast because of a family problem, those who didn’t sleep at night while caring for a sick relative, or other realities.
“How do you measure that? Come on!” Gandolfo said. “There’s not a mathematical formula.”
Also, children don’t all grow at the same pace, he noted. Sometimes, one lesson received from a teacher doesn’t click until years later. But that teacher made a positive difference, regardless of a test score, he argued.
The model has insulted teachers, damaged classroom relationships, and made some educators reluctant to helping the neediest children, Galdolfo said. “That’s criminal.”
Brandon, the Pasco teacher, shared those frustrations.
She agreed that it’s hard enough to get students to care about tests that don’t directly affect them and only matter to the teachers.
She pointed to one student who scored above his grade level on state exams, yet the result counted against her because it didn’t meet the VAM projection. Others made double-digit gains, but remained at the lowest level on their exams – again, a blemish against her.
“I feel like it’s unrealistic,” Brandon said.
Add into the mix the students’ daily lives, which aren’t captured in the complex formula, and the value-added model simply becomes a way to push hard-working teachers who care out of those schools, she said.
New educators arrive and have to recreate the bonds, she said. And some may hesitate to move from higher performing schools, knowing their VAM score could cause a transfer.
Fixing the timing of when the scores come out is perhaps the least the state could do, Brandon said.
Had she been warned of the possibilities, Brandon said, she would not have spent the time creating connections with students that were destined to be broken, hurting their education.
Plus, she added, she could have looked for a different school “when there were hundreds of jobs out there,” rather than the few that remained after classes had already begun.
“I don’t doubt my ability as a teacher because of those scores,” Brandon said. “Not everybody grew (in the class). But I was really proud of my scholars.”