In a recent series of articles that involved ranking 329 elementary schools in the four largest Baltimore-area school systems as to their effectiveness in teaching reading, four Baltimore elementaries - Lakewood, Malcolm X, Montebello and Samuel F. B. Morse - should not have been ranked because of insufficient test data. In addition, the rankings of 18 other city elementary schools were based on three available test scores, not four scores as were rankings for the remaining 97 city elementary schools.
Principal Sarah Horsey spots the problem right away. One teacher says perimeter is expressed as feet. Another next door tells her class it's square feet.
"OK, we need to get this straight, so we're all teaching the children the same thing," Horsey tells a staff meeting she calls later that day. "If I'm confused, think about how the kids are trying to learn."
Just another classroom correction for Horsey, Pimlico Elementary's principal nuisance - and savior.
Let her into your class, she'll take over the lesson. Let her into your cafeteria, she'll stage a revival-style meeting to persuade children they're the smartest around. Let her into your school, she'll transform it into a place free of fights and full of learning.
"I don't stand in the background," Horsey says. "When I'm here, I'm going to be involved."
In other words, she's exactly the kind of principal needed by the Northwest Baltimore school - a school with such poor test scores in 1996 that it was on the verge of being taken over by the state.
Today, with almost half of Baltimore's schools on state probation for low test scores, Pimlico is the one closest to escaping that status. Its state test scores are far above the city averages - and not too far behind the state averages - even as its neighborhood remains mired in Baltimore's web of drugs and poverty.
"Pimlico is Pimlico again, and it's all due to Ms. Horsey," says Louise Hamilton, grandmother of two Pimlico pupils. "She came in here and did it, and for this area, this school has come a long ways."
Horsey's single-minded drive to improve instruction helped push Pimlico to a level far above its peers in reading. A Sun statistical search for exceptional schools found its students score higher on reading tests than all other city elementaries when student demographic factors are held constant.
In the past four years under Horsey, Pimlico has become the proverbial rose in the forest. Rare is the inner-city school that has risen from the depths of failure to compete on Maryland tests with the average performance of schools in much more well-off suburbs.
Like the other highly effective schools in the Baltimore area identified by The Sun's analysis, Pimlico relies on a consistent set of key practices to boost reading achievement.
These include a heavy dose of lessons in letters and sounds, or phonics; teaching in the smallest possible groups with frequent testing to catch those falling behind; and tight supervision of school time to spend as much of it as possible on reading.
Above all else, though, Pimlico has built its success on Horsey, a 55-year-old principal who sees her role as the leader of teachers and who appears to spend most of every day making sure they teach children to read.
Expectations of pupil achievement used to be high at Pimlico. Thirty years ago, it was among the city's more respected public elementaries. But as the Northwest Baltimore neighborhood succumbed to poverty, boarded-up rowhouses and corner drug salesmen, expectations at the school declined.
By the mid-1990s, Pimlico was no different than the dozens of other failing elementaries across the city. Achievement was low. Trash and graffiti were everywhere. Parents were seldom found in classrooms.
Horsey came to Pimlico in 1996 from Rognel Heights Elementary, a school in a more middle-class neighborhood. She arrived just after Pimlico was officially threatened with state takeover because barely a handful of its students were reading on grade level.
"The first day of school, I was thinking to myself, 'Oh, no. I've made a terrible mistake,'" Horsey says. "I couldn't believe what I had gotten myself into."
While Pimlico had a core of veteran teachers from the days when it was successful, much of its staff was inexperienced and transferred out as quickly as possible. Some veterans were, in Horsey's opinion, overdue for retirement, more intent on enjoying morning coffee than beginning reading lessons when the school bell rang.
So she pushed some of them out, racing up and down three flights of stairs after her morning announcements to document their lack of prompt attention to instruction.
The 'Pimlico walk'
Of Pimlico's 600 or so students, almost all now qualify for free or reduced-price lunch, and almost a quarter of them move in or out each school year.
Like many urban elementaries, Pimlico's community often spilled over into its classrooms and playground. Fights were common and classroom interruptions were the norm, gobbling up critical teaching and learning time.
So discipline became the top priority for Horsey's first few months, with reading instruction taking a back seat.
Horsey hired neighborhood parents at minimum wage for a position she called "parent technician." Their job: To sit in the classes of new teachers who were struggling with discipline, offering a firm voice from the neighborhood to keep the kids in line.
An in-school suspension classroom was set up for persistent troublemakers, taking a full-time teacher out of a regular classroom. Teachers gladly accepted slightly larger classes in exchange for losing one or two kids who routinely disrupted instruction.
"I guaranteed teachers that they would be able to spend at least 95 percent of their time on instruction, and if they can do that, achievement will improve," Horsey says. "To give them that time, I gave them permission to send out the children who were getting in the way."
In the mornings, children entering the building often couldn't be trusted to walk calmly to their classrooms without fighting, so teachers accompanied them through the halls. Horsey says that practice wasted valuable early morning preparation time, and she decided the children would be taught to walk properly.
"We were pretty skeptical when she said that children would be walking to their classes on their own, without teachers, because we had seen their behavior," says Gerald Nixon, a fifth-grade master teacher in his 29th year at Pimlico. "But she insisted that it would work, and it did. It gave us a lot more time to get ready in the mornings."
The result was the "Pimlico walk" - a proper way of marching to class, hands folded behind backs, without hitting, jostling, running or bumping into others.
"Before Ms. Horsey was here, there was a whole lot of fighting," recalls fifth-grader Keosha Herndon, 11. "You'd be afraid to walk on the playground. But when she took over, everything died down."
Focus on teaching
With order restored, Horsey turned her focus to teaching.
Research into what makes for an effective school has found again and again that success rests on the principal - or another key administrator - playing the role of instructional leader.
Successful schools - particularly schools that break the typical correlation between student achievement and family income - routinely are led by principals who recognize that they are responsible for helping teachers learn to teach better. But decades of research - and countless training seminars by school systems -have failed to establish ways of cloning such principals.
"Our job is so important that if you don't speak to somebody, their whole day is put off," Horsey says. "We have to be something to everybody at all times, yet stay focused on instruction."
The youngest of five children, Horsey grew up in the North Bentalou Street area of West Baltimore, "living in poverty but I didn't know it." Education never wavered as the family's priority, and all five graduated from college - as have Horsey's two grown children.
Like many principals, she swaps her heels for comfortable walking shoes to more easily get around to every classroom every day. But unlike most, Horsey refuses to stand quietly in the back of classes and merely watch the instruction.
She unfailingly insists on leaping in, taking over the lesson, sometimes for a minute, other times for as long as 10 minutes.
"I don't understand this square feet," Horsey says, moments after bursting into that confusing third-grade lesson. "Can someone explain the difference between perimeter and area?"
She throws question after question at the students, taking them through some quick instruction on how to measure the perimeters of zoo cages.
"I am a teacher. I can't ever forget that," Horsey says. "So when I come into a classroom, I want to teach."
Sometimes, Horsey comes in at the end of lessons, stealing a couple of minutes to ask questions - to essentially check on the quality of the teaching by seeing if the pupils can correctly recite what they should have learned.
When she hears the right answers, she often ends her visit with the request: "Students, please give your teacher a hand for doing such a good job." And they applaud.
For some teachers, it's been quite an adjustment to have a principal who comes in and takes over almost daily. But it's an adjustment most have found worth making.
"She loves to teach, and she's a hands-on person," says fifth-grade teacher Carole Rehak, a 10-year veteran who has been at Pimlico for three years. "It's her. You're not going to change it, so you accept it. The good thing is that she has some really good ideas to help you improve."
In the past few years, instruction at Pimlico has become even more closely supervised.
Horsey assigns herself one grade level to oversee. The school's assistant principal takes another grade. The school's two master teachers each take a grade, and a retired school administrator has come back to Pimlico and assumed day-to-day control over the pre-kindergarten through first-grade teaching.
Like almost all other city elementaries, Pimlico uses the Open Court program for early reading instruction - a phonics program credited with producing the recent surge in city test scores.
Open Court is building a strong track record with children from all types of neighborhoods, from the tiny rowhouses of Baltimore to the expensive McMansions of western Howard County. It helped Pimlico - a school with virtually all low-income, African-American enrollment - become one of only five city elementaries in which 50 percent or more of the children in every grade scored above the national average on the most recent round of reading tests.
But Horsey refuses to let her teachers rely solely on the city's reading programs. She insists on adding extra lessons in handwriting and spelling, fearing that those oft-neglected subjects are left wanting in texts.
And she posts detailed instructions on staff bulletin boards on how teachers should use their class time - critical directions when almost a third of the school's teachers have less than two years of experience.
Teachers are told to spend 10 or 15 minutes on children reading to themselves. Then they're to move to phonics instruction. And on and on, scripted through the day.
It's a handbook on how to keep a schoolwide focus on instruction and not waste time.
As part of that effort, class sizes for morning reading instruction have been cut far beyond levels of the citywide class-size reduction effort. Almost every teacher is a reading teacher. Master teachers, the school's physical education teacher, its art teacher and others are assigned morning classes, putting a dozen or fewer pupils in most reading groups.
One morning, Joyce Dixon smoothly moves her first-graders through an exercise - thinking up words that begin with "I" - while at the same time heating up doughnuts and distributing milk and juice for their breakfast. Like most schools in poor neighborhoods, breakfast is provided with federal funds, and it can't be ignored.
One student suggests the word nice.
"No. It must begin with the letter I," Dixon says.
"Class, say I," she says.
"I," they reply."
"Say it again."
She points to a picture of an Indian and of an ice-cream cone, and almost on cue the kids start coming up with their own "I" words.
"In," says Ashley Daney.
"It," says Bria McClain.
"Alright, time for breakfast. Let's go," Dixon says and snaps her fingers three times - pleased that she stole a few otherwise lost minutes for a bit more teaching in the basics and that her children seem to have picked up on the lesson.
A sense of ownership
When Horsey arrived at Pimlico, she saw not just a school needing uplifting, but a community needing an escape from drugs and poverty. So turning Pimlico into a highly effective school has meant going far beyond its classrooms.
Pimlico's pupils walk to school along streets lined with boarded-up houses. Many are being raised by one parent or grandparent.
Horsey has begun a grandparents group and intends to start one soon for single parents. To give pupils and their families a stronger sense that they own the school, every child's picture is posted on a hallway wall - accompanied by a certificate on which each promises to try to succeed this school year and a sampling of each one's schoolwork.
When Horsey proposed covering the walls with pupils' pictures and writings, many parents and teachers scoffed at the idea, fearing some of the children would go through the halls marking it up and tearing it down.
Instead, Pimlico's pupils have proudly watched over the displays. In the halls, kids keep their distance from their peers' work, and those few spied defacing it are typically reported by other children.
None of this has been gained easily at Pimlico. Until children begin believing they can be successful, Horsey says, they'll never break the cycle of failure - a cycle still well-established at many schools in similar settings, in Baltimore and elsewhere.
Look her in the eye
So almost every morning, she's in the cafeteria where pupils assemble before going to class - and she invariably reminds them that they can, they must, succeed.
"What's the best elementary school in Baltimore City?" she asks.
"Pimlico," the children reply.
"I can't hear you," she says. "Say it so Mayor O'Malley can hear it all the way down in City Hall.
"What's the best school in all of Baltimore?"
"Pimlico!" they shout.
As the children line up by grade to go to class - neither jostling nor fighting - she tells them they have to report to her that they have done their best in class.
And she adds that each also has to look her in the eye as they tell her that - and shake her hand, too.
"Looking an adult in the eye, having a nice, firm handshake, talking clearly and loudly, those are the kind of skills that give them confidence," Horsey says. "It's what they need to succeed in school and in life."
Tomorrow: Bushy Park Elementary: Using time efficiently.
About this series
Firmly identifying elementary schools that excel at teaching reading is a complex problem, primarily because no two student bodies are exactly alike.
But an unusually rigorous analysis of schools' test scores linked to data about particular pupils made it possible to pinpoint exceptional schools in the four largest school systems in the Baltimore area.
The goal was to identify which elementary schools within each of the districts are doing more to improve the reading skills of their pupils.
Frequently, such efforts involve nothing more than finding schools with the highest test scores. But the very top scorers are usually schools with the least poverty among their pupils.
Instead, The Sun's analysis -- performed by Geoffrey D. Borman of the Johns Hopkins University's Center for Social Organization of Schools -- used statistical techniques to essentially level the playing field and clearly highlight the varying effectiveness of schools' reading instruction.
Borman's study controlled for a long list of pupil factors that correlate with achievement, among them: family income, race, attendance, mobility, and percentages of special education pupils and non-native English speakers.
As a result, a small group of elementaries emerged in each district as much more effective at teaching reading than others. These schools -- no matter what pupils they serve -- break the typical link between achievement and such factors as family income.
At the same time, some schools with high test scores were found to be only average when it came to making a difference for their pupils. And some schools with low test scores are having more of an impact than it would seem at first glance.
In choosing one of the most effective schools to profile in each district, a mix of communities was sought -- to provide a broad basis for learning why some schools are much more successful.
One of the area's six school systems, Harford County, was unable to provide some essential test scores, and so a Harford school is not profiled. School officials there say they discovered only recently that the data had not been kept by their testing company.
And Carroll County school officials refused requests to provide test scores linked to particular students and schools, a requirement for Borman's analysis. Even though all students in his study were identified only by random numbers, not their names, Carroll school officials cited privacy concerns.
By contrast, officials of the four largest school systems in the area -- Baltimore and Baltimore, Anne Arundel and Howard counties -- were satisfied that the study protected their students' privacy.
Baltimore City schools
The impact of the reading instruction programs at each Baltimore elementary school is ranked below on a percentile scale from most effective to least effective - with the 50th percentile being average. Within five broad groupings, schools are listed alphabetically.
(100th to 81st percentile)
Dallas F. Nicholas Sr.
Sarah M. Roach
(80th to 61st percentile)
Francis Scott Key
John Eager Howard
Malcolm X Primary
Moravia Park Primary
Thomas G. Hayes
(60th to 41st percentile)
Barrister Charles Carroll
Carter G. Woodson
Dr. Bernard Harris Sr.
Dr. Martin Luther King Jr.
Dr. Rayner Browne
Elmer A. Henderson
Furman L. Templeton
Graceland Park/ O'Donnell Hgts
Robert W. Coleman
Samuel F.B. Morse
(40th to 21st percentile)
Charles Carroll of Carrollton
Commodore John Rodgers
Maree Garnett Farring
Matthew A. Henson
(20th percentile and below)
Mary E. Rodman
Walter P. Carter
NOTE: This analysis is based on test results linked to student data provided by the Baltimore school system and the Maryland State Department of Education. That included reading achievement scores from the Maryland School Performance Assessment Program tests, given annually to all third- and fifth-graders, and from a national test, the Comprehensive Tests of Basic Skills, given annually to all city elementary pupils. Particular weight was given to the test scores of pupils who remained in the same schools for several years.