Just a few months before Artemus Werts stood before classes in Paul Laurence Dunbar High School, he had to relearn math concepts he hadn't reviewed in years.
Werts, a 2011 Rutgers graduate who joined Baltimore's teaching corps through Teach for America this year, began replacing formulas with patterns, talking himself through problems step by step. He found the skills not only came back to him, but stuck.
"I wasn't comfortable teaching math, but then it was empowering," said Werts, who studied history and eventually hopes to work in education policy. "A lot of things I was re-teaching myself are places where my students struggle. I realized that if I could do it just a few months before, my students can do it too."
A study commissioned by the U.S. Department of Education found that Teach for America teachers — recent college graduates or professionals who commit to teaching in low-performing, low-income public schools for at least two years — were as effective, if not more, at teaching math as those from similar programs or even career educators.
In Baltimore, Teach for America has drawn opposition from school officials who question whether the organization provides sufficient training — mirroring a national debate. The organization says it provides rigorous training and much-needed educators dedicated to making a difference in struggling schools.
Mathematica Policy Research, a nonpartisan research firm, conducted the study that compared a random sampling of students from across the nation, including Baltimore.
The authors found that middle and high school students taught by Teach for America teachers scored higher on assessments than those assigned to traditionally certified teachers. In some cases, student gains were equivalent to an additional 2.6 months of instruction.
The study also found that students taught by Teach for America teachers with three years or less experience outperformed students of teachers with three or more years of experience.
On average, the study suggested, secondary school principals would raise student math achievement by hiring from the Teach for America corps rather than teachers from traditional backgrounds.
Teach for America proponents said the study debunks the perception that alternatively certified teachers are inferior to their counterparts.
"Overall, the study is encouraging," said Courtney Cass, executive director for Teach for America in Baltimore. "It makes us feel like we're on the right track, and we're encouraged that we're having a positive impact on student outcomes."
Cass said she hopes the study leads to more conversations about the qualities that all educators committed to improving education share.
"It would be a great next step for this research to start looking at what are the characteristics that all teachers from all backgrounds share," she said, "and look for opportunity for us all to be facilitating the professional development, mentorship that contributed to that."
But critics said the study's conclusions are based on a small fraction of the Teach for America corps and shouldn't negate the fact that intensive preparation and experience are invaluable.
The Baltimore City school district employs 622 Teach for America teachers or alumni and 45 as principals or central office administrators.
The city school board, which approved a $1 million contract to hire and train up to 150 teachers from Teach for America this school year, recently ordered an in-house study of their effectiveness in the classroom. The study will also look at those hired through the city's teaching residency program, another alternative certification route.
Board Commissioner David Stone questioned whether the district was making a good investment in teachers who come to Baltimore with little experience. Teach for America educators have five weeks of training and then professional development throughout the year, and are required to make a two-year commitment.
He pointed to the 0.07 standard deviation difference in scores between Teach for America students and their counterparts in the study, which were considered statistically significant by Mathematica.
"You could argue that differences are not so great. ... They were actually minuscule," Stone said. "And the way they equated it to months of learning is dubious at best."
He added: "For every teacher who changes a test score, there's another that changes the trajectory of a child's life, and that comes with experience."