LEADING LEARNING - Education Week · district level for its second annual “Leading for...

Supplement to the September 14, 2005, issue of Education Week

LEADINGfor LEARNING

Once overlooked, central offices are now seen as playing a key role in improving student achievement.

A special report funded by The Wallace Foundation

LEADINGfor LEARNING

Not long ago, a popular theory about school

improvement went something like this: Put in

strong principals and dedicated staff

members, and then get out of their way. When

it came to improving teaching and learning,

the thinking went, the central office had little to add.

The upshot was an era of policies that limited the role

of district-level leadership in matters of instruction. Site-

based management and “whole-school reform” models

flourished in the 1990s on the premise that individual

schools alone could raise achievement.

And the idea worked. Or rather, it worked for some

schools, while others languished. As a result, a new

consensus is emerging in the field that strong district

leadership is needed to bring about large-scale

improvement—now a mandate under the federal No

Child Left Behind Act.

“Either you believe in district reform, or you’re going

to have to be extremely patient in waiting for a school-

by-school turnaround,” says Jane Hammond of the

Stupski Foundation, a Mill Valley, Calif.-based group that

helps districts with strategic planning.

Education Week is focusing on leadership at the

district level for its second annual “Leading for

Learning” special report. The key question is: What

strategies should district leaders pursue to influence the

quality of teaching and learning?

To answer that, we tell the stories of two school systems

that re-established the role of the central office in guiding

instructional improvement: the 10,000-student Gilroy

Unified schools in California and the 26,000-student

Clarksville-Montgomery County system in Tennessee.

The two districts—both of which work with the

Stupski Foundation—have sought greater consistency

across schools in content and teaching methods. They’ve

created new ways for teachers to learn together and use

student data. And, they’ve each seen more students

succeed academically.

To get a sense of how widespread such approaches are,

the Education Week Research Center also commissioned a

poll of superintendents that asked what practices they use

to improve instruction. The results show district leaders

across the country embracing many of the strategies

employed in Gilroy and Clarksville-Montgomery County.

True, districts still reflect a range of approaches. Some

are more explicit in telling schools what instruction

should look like—a method that some are now calling

“managed instruction.” Others prefer to set broader

SEPTEMBER 14, 2005 ■ EDUCATION WEEK S3

By J e f f A r c h e rI l lustrat ions by B r i a n J e n s e n

The idea that schools can improve on their own gives way to a focus on effective district leadership.

Theory of Action

Guiding HandA national survey finds districts exertingleadership over instruction. S5

Forward MotionThe Gilroy, Calif., schools have laid thegroundwork for improvement. S12

In Sharp FocusA Tennessee school district brings ananalytical eye to data. S18

boundaries and then step in where they see problems.Many experts see the growing assertiveness of district

leaders as a natural consequence of the movement forhigher academic standards that has dominated educationpolicymaking for more than a decade. It’s too much, theysay, to presume that every school has within it the capacityto bring its students to the levels of achievement nowdemanded of them.

“When you have a policy environment now that expectschange to occur at scale, that means that districts have toimprove all schools, essentially simultaneously,” saysWarren Simmons, the executive director of the AnnenbergInstitute for School Reform, located in Providence, R.I.

Tellingly, the Bill & Melinda Gates Foundation—one of thestrongest promoters of designs for small schools—has drafteda new white paper arguing that schools are most likely tosucceed if they’re part of a supportive district, or, in the case ofcharter schools, part of a larger network of schools.

“We’ve spent over a billion dollars on almost 2,000schools, and what we found is that most people don’t knowwhat to do, and how to do it,” says Tom Vander Ark, theexecutive director for education at the Seattle-basedfoundation.

Mounting evidence suggests that effective schools aremost often found in districts with strong systemwideguidance. In 2002, the Council of the Great City Schoolsidentified some parallels among improving districts in aninfluential report, “Foundations for Success.”

The council described strategies employed in Charlotte-

Mecklenburg, N.C.; Houston; Sacramento, Calif.; and thesubset of schools in New York City then known as theChancellor’s District.

Each district had a common curriculum, and had set uptraining and monitoring systems to ensure consistentapproaches toward instruction across schools. The districtsalso made frequent use of student-performance data toinform educators’ decisions.

“You have to take responsibility for the overallinstructional program,” says Michael D. Casserly, theexecutive director of the Washington-based council, “ratherthan just abandon that to the individual schools withoutproviding direction, technical assistance, or professionaldevelopment, and just hoping for the best.”

Chrys Dougherty, the director of research at the NationalCenter for Educational Accountability, says much the sameis true in most of the districts named as finalists andwinners for the annual Broad Prize for Urban Education,which recognizes improved student performance.

“When you go to effective schools that are in a districtthat has certain things in place, they will say their job wasmade infinitely easier by the fact that the district did thesethings,” says Dougherty, whose Austin, Texas-based centercollects the data used to make the Broad Prize selections.

On the surface, this larger role for the centraloffice might seem at odds with the concurrentpush to give families more options. Some of thebiggest urban districts, for example, are creatinglarge numbers of new schools with different

designs—what’s come to be called a “portfolio” model.Likewise, decentralized decisionmaking still has plenty of

proponents, as seen in the number of districts giving schoolsites more power to hire whom they want and to spend theirbudgets as they see fit.

But strong district leadership is needed for empowermentof school sites to succeed, says Joseph Olchefske, a formersuperintendent of the Seattle public schools. As a districtchief, he gave schools considerable leeway to design theirown programs, but that didn’t mean anything goes.

“You’ve got to set standards, you’ve got to create andimplement assessments that are for all kids, regardless ofthe school, and have very clear accountability, which meansconsequences,” says Olchefske, who is now the managingdirector of a new consulting group at the AmericanInstitutes for Research, located in Washington.

Michael Fullan, an expert on school systemmanagement at the University of Toronto, says one of asuperintendent’s biggest challenges is finding the rightbalance between central authority and site-basedautonomy. Ideally, he argues, schools should feelownership of a common vision of instruction.

“If you’re too loose, you don’t get the focus, but if you’retoo focused, you get prescription, and narrowness, andrebellion,” he says. “The holy grail of school reform on alarge scale is large-scale ownership.”

Whether most districts in the United States can achievethat balance remains to be seen. But as the survey resultsand the stories of the two districts in this report suggest,few district leaders are leaving things to chance. ■

S4 EDUCATION WEEK ■ SEPTEMBER 14, 2005

About This ReportThis special pullout section is the second of three

Education Week annual reports examining leadership ineducation, a topic of critical concern at a time of ever-increasing expectations for schools. Each report includes a mix of explanatoryarticles and researchfindings analyzed bythe Education WeekResearch Center.

The project isunderwritten by agrant from The WallaceFoundation, which seeks tosupport and share effective ideas and practices thatexpand learning and enrichment opportunities for allpeople. Its three current objectives are:

• Strengthen education leadership to improve student achievement;

• Enhance out-of-school learning opportunities; and• Expand participation in arts and culture.

For more information and research on these and other relatedtopics, please visit the Knowledge Center at www.wallacefoundation.org.For copies of last year’s special report, go to www.edweek.org.


WITH EXPECTATIONS FOR STUDENTachievement at an all-time high, school dis-trict leaders say they are playing a more as-sertive role in shaping instruction, accordingto a national survey of superintendents com-missioned by Education Week.

The nationally representative poll of 813top district officials shows large numbers ofthem turning in recent years to commonplanning time for teachers, new forms ofclassroom observations, and, in particular,data-driven decisionmaking as systemwidestrategies for improvement.

By showing the extent to which superin-tendents are employing specific practices,the results offer a rare snapshot of instruc-tional leadership at the district level. Amongthe obstacles faced in providing that leader-ship, superintendents most often cited lackof money and competing priorities.

At the same time, the poll reveals clear dif-ferences depending on district size. Leaders oflarger systems were more likely to favor stan-dard approaches across their schools, such as“pacing guides” that show teachers what con-tent to cover at what time throughout the year.

Other trends emerged for grade level andsubject matter. Superintendents said theymade greater use of common curricula and pe-riodic districtwide tests in the early gradesthan in high school, and in reading and mathe-matics than in other subjects.

The survey did not probe how well dis-tricts are using the strategies that theirleaders say they have put into place, and itdid not connect them to student outcomes.

But the results do suggest the kinds of man-

agement tools that leaders are reaching for asthey try to create a common understandingabout instruction in their districts, while also en-couraging their schools and teachers to becomemore targeted in how they meet students’ needs.

THE EDUCATION WEEK RESEARCH CENTERdesigned the telephone survey with BeldenRussonello & Stewart, a Washington-basedpolling organization, which conducted the re-search in June.The margin of error for the over-all results is 3.3 percentage points. Responseswere broken down by district enrollment size:

By J e f f A r c h e r

Guiding HandCurrent Instructional-Leadership PracticesThe top 10 instructional practices that a majority of superintendents useinclude a common district curriculum and induction programs for newteachers. Other methods, such as use of pacing guides or designating ateacher-leader position in each school, are less common.

Training for teachers and principals on using performance data

Common district curriculum

Instructional walkthroughs

Standard process for writing school improvement plans based on performance data

Induction programs for new teachers

Same mathematics programs across district

Same reading programs across district

Common planning time for teachers

Periodically administer own districtwide assessments

Adjust instruction based on districtwide assessments

SOURCE: Education Week Research Center

93%

92

90

81

81

80

79

71

68

60

Perc

ent

of s

uper

inte

nden

ts

In a poll, superintendentsreport more active rolesin teaching and learning.

fewer than 2,000 students, between 2,000 and10,000, and more than 10,000—with margins oferror of 5.3 percentage points for the group withthe smallest enrollments and 5.8 points for theother two.

Sixty-four percent of all of the respondentswere in their first superintendencies, but re-ported relatively lengthy tenures, despite the na-tional concern about leadership turnover. Abouthalf the superintendents in each size categoryhad been in their jobs five years or more.

Enrollment size had little bearing on some ofthe superintendents’ views of the No Child LeftBehind Act, which aims to make schools more ac-countable for raising student achievement.Roughly three-quarters of those in each groupagreed or strongly agreed that the federal mea-sure, signed into law by President Bush in Janu-ary 2002, has forced them to be more active “inguiding the kind of instruction that happens inthe classroom.”

Size mattered in whether they saw suchchange as needed, however. Seventy-eight percentof the leaders of large districts strongly agreedthat, regardless of the law, they should be moreactive in guiding instruction, compared with 68percent in medium-size districts and 56 percentin small ones.

Asked whether, in the past three to fiveyears, more instructional decisions were beingmade at the “district level, as opposed to at theschool sites,” 75 percent of those in large dis-tricts agreed or strongly agreed, compared with58 percent in medium-size ones and 52 percentin small systems.

Such differences evened out when superinten-dents were asked more generally about howmuch of a role they and their district staff mem-bers should play in “providing direction” on in-structional matters for their schools. Ninety per-cent of all respondents said “a large role,” and 10percent said “some.”

ONE OF THE MOST POPULAR WAYS OFproviding that direction is through overseeingwhat gets taught. Ninety-two percent of all re-spondents said they use districtwide curricula.The practice is much more common in the earlygrades: Ninety-six percent said it was used forelementary schools, 88 percent said for middleschools, and 77 percent said for high schools.

Leaders of large districts see the most benefitin doing so. Eighty-eight percent of superinten-dents of districts with more than 10,000 studentssaid a common curriculum could have a “greatdeal” of effect on improving student achievement,compared with 72 percent of those in systemswith fewer than 2,000 students.

About 80 percent of all superintendents sur-

veyed said they require schools across their dis-tricts to use the same reading programs; a sim-ilar proportion said the same of math. Roughlyeight in 10 also said they had all their elemen-tary schools use the same textbooks in thosekey subjects.

Although somewhat less widespread, instruc-tional-pacing guides are spreading quickly amonglarge school districts. Forty-four percent of the su-perintendents in large districts said they had hadsuch guides for at least three years, but 65 per-cent said they have them now.

Also growing rapidly is the use of “instruc-tional walkthroughs,” defined as classroom obser-vations not for teachers’ job reviews, but for im-proving their instruction. Ninety percent of allrespondents said their districts use walk-throughs, with 27 percent reporting that they hadadopted the technique in the past three years.

For the most part, however, the practice islimited to school site leaders. Among superin-tendents who said that their systems made useof such observations, 96 percent said their prin-cipals did the walkthroughs, 46 percent saidcentral-office staff members, and just 20 per-cent said teachers.

More often, district leaders use common plan-ning time to let teachers learn from one another.Seventy-one percent said their teachers’ sched-ules let them meet by each grade level or de-partment. Twenty-two percent said their dis-tricts had begun scheduling common time in thelast three years.

In a similar vein, 54 percent of the large-districtsuperintendents said each of their schools had ateacher freed from classroom duties to coach oth-ers on their instruction. That contrasted with 34percent of those in medium-size systems, and 27percent in small ones.

FEW STRATEGIES APPEAR TO BE GROWINGas fast as those related to data. The survey re-sults paint a picture of districts building uptheir technology systems and professional de-velopment so their educators can better useinformation on student performance to driveinstruction.

For example, a little more than half of all respondents said their districts had beentraining principals and teachers on how to make instructional decisions based on datafor three years or more. Another 40 percentsaid they had begun providing that trainingsince then, meaning more than 90 percent now do so.

Twenty-three percent said that three yearsago they had data-management systems thatallowed educators at school sites to accessachievement information on individual stu-


Superintendentsreport districtsare playing a larger role in shapinginstruction.


District-Level DecisionmakingA majority of superintendents report that over the past threeto five years, more instructional decisions are being made atthe district level, instead of at school sites.

23%Strongly

agree

NCLB has forced districtleaders to play a larger role

Regardless of NCLB, district leaders need to play

a more active role

4%

14%Strongly disagree

26%Somewhat disagree

33%Somewhat

agree

Don’t Know/Refuse

33%

42%

13% 12%Perc

ent

of s

uper

inte

nden

ts

The Influence of No Child Left Behind About three-quarters of superintendents say the No Child LeftBehind Act has forced district leaders to play a larger instructionalrole. But even more, 93 percent, feel that district leaders need to play a more active role than in the past in guiding classroominstruction, regardless of the federal mandates.

61%

32%

4%2%

Strongly agree

Somewhat agree

Somewhat disagree

Strongly disagree


School-Level DifferencesWhile 92 percent of superintendents say there is a commoncurriculum across their districts, and 68 percent say they administertheir own periodic districtwide assessments, such practices aremuch less prevalent in high schools or across all grade levels.

77%

88%96%

73%

Perc

ent o

f sup

erin

tend

ents

60%

78%

95%

55%

Common district curriculum Periodically administer owndistrictwide assessments

High school

Middle school

Elementary school

All grades

Lack of money

Competing priorities

Lack of district staff

Teachers’ concerns about lost creativity

Lack of proven instructional strategies

Union contracts

Principals’ concerns about lost autonomy

Obstacles to Instructional LeadershipAlmost 90 percent of superintendents report that a lack of moneyprevents them from acting as instructional leaders in their districts,and almost 70 percent cite competing priorities as a barrier.

Percent of superintendentsreporting obstacle prevents themfrom acting as instructionalleaders a “great deal”

Percent of superintendentsreporting obstacle prevents themfrom acting as instructionalleaders “somewhat”

56% 33%

24 45

22 39

8 47

14 39

16 29

5 39

23%Strongly

agree

33%Somewhat

agree26%

Somewhat disagree

14%Strongly disagree

4%

89%

69

61

55

53

45

44

dents online. But another 33 percent said they have put that kind of system in placesince then.

Responses also pointed to big changes inthe kind of performance data collected. A totalof 68 percent of the superintendents said theynow give districtwide student assessments periodically throughout the year. Twenty-seven percent of all respondents said they hadbeen doing so for less than three years.

Of those who said they do such testing—often called “benchmark” testing—95 percentsaid it was in the elementary grades, comparedwith 60 percent at the high school level. Twelvepercent said they gave the tests monthly, 47percent said every six to nine weeks, and 37percent said less often.

Data analysis has become a common part ofschools’ annual planning as well. Eighty-one per-cent of all respondents, and 98 percent of those inlarge districts, said they had a standard processfor schools to draft improvement plans that in-cludes an assessment of student performance.

Leaders of large districts expressed themost faith in data-driven decisionmaking.Ninety-one percent, for instance, said trainingdistrict staff members in using data promisedto have a “great deal” of effect—the highestpercentage among responses given to ques-tions about 15 different strategies.

BUT SMALL DISTRICTS ARE MAKING BIGchanges in their use of data. Nearly a third of




Adjust instruction based on districtwide assessment

Online data-management system to analyze student performance

An Emerging TrendRegular collection of student-performance data and use of thedata to inform instruction aregrowing trends at the district level,with significant percentages ofsuperintendents reporting that theirdistricts have engaged in suchpractices for less than three years.

Perc

ent

of s

uper

inte

nden

ts

Less than three years

Three years or more

40% 53%

27 41

30 30

33 23

Data DifferencesCompared with leaders ofmedium-size and smalldistricts, superintendents of large districts moreconsistently report that theirdistricts administer their own benchmark assessmentsand engage in other effortsrelated to analyzing student-performance data.

Large districts (enrollment above 10,000)

Medium districts (enrollment 2,000 -10,000)

Small districts (enrollment less than 2,000)




Adjust instruction based on districtwide assessments


Perc

ent

of s

uper

inte

nden

ts

99%95

91

77

9886

8472

64

7466

57

77

49


67

93%

68

60

56

the superintendents of districts with fewerthan 2,000 students said they had been usinga student-data-management system for lessthan three years. That’s in addition to 17 per-cent who said they already were using one.

ASKED ABOUT POTENTIAL OBSTACLES, 65 PER-cent of respondents in small districts said “lack ofthe kind of staff” needed at the central office pre-vented them either a great deal or somewhatfrom providing leadership on instruction. Fifty-eight percent of those in large systems said thesame thing.

Leaders of large districts were more likely tosee union contracts as a problem. Fifty-three per-cent said labor pacts adversely affect their efforts

on instruction either a great deal or somewhat,compared with 43 percent of superintendents insmall districts.

But other factors were seen as far greaterimpediments. Eighty-nine percent of all re-spondents, with few differences by districtsize, cited scarcity of money as a great deal orsomewhat of a problem. Sixty-nine percentsaid the same about distractions posed byother priorities.

In contrast, 55 percent of all the superin-tendents polled said they were prevented agreat deal or somewhat from exercising in-structional leadership by “teachers’ concernsabout lost creativity.” Fifty-three percent char-acterized “lack of research-proven strategies”as a problem. ■


Impact on Achievement A Matter of SizeAcross the board, superintendents of large districts are more likely thansuperintendents of smaller districts to believe that different instructionalleadership practices will affect student achievement a “great deal.”

Large districts (enrollment above 10,000)

Medium districts (enrollment 2,000 -10,000)

Small districts (enrollment less than 2,000)




Instructional walkthroughs

Districtwide pacing guides

Teacher-leader position in each school

Percent of superintendents indicating practice affects achievement a “great deal”

81%

57

50

67

66

7562

43

6852

49

6448

35

78

Practice vs. PerceptionSuperintendents largely believe that theinstructional leadership practices most in useare the ones that exert a “great deal” ofimpact on student achievement. However,their ratings of effectiveness diverge fromactual use for two methods—instructionalwalkthroughs and adjustments in instructionbased on benchmark assessments.


Instructional walkthroughs Adjust instruction based on districtwide assessments

Perc

ent

of s

uper

inte

nden

ts

6350

32

90%90%

51%51%60%60%

74%74%

Percent of superintendents reporting theirdistricts use the practice

Percent of superintendents reporting thepractice has a “great deal” of impact onimproving student achievement

In Gilroy, Calif.,educators have learned acommon processfor improvementplanning. The restis up to schools.

Gilroy, Calif.

When Dawn O’Connor returned to her job as a

science teacher here last fall after five years away

to raise her children, she found a very different

school district. Unlike in the past, teachers were

visiting one another’s classrooms. They were meeting regularly to

examine student performance.

“Before, when we got our scores at the beginning of the year from the

year before, we said, ‘Oh well, we didn’t get the scores that we wanted,’ ”

recalls O’Connor, who teaches at Ascencion Solorsano Middle School.

She adds, “Other than my regular [job] evaluation, I can’t remember

anyone being in my room.”

The change wasn’t by accident. Leaders of the 10,000-student Gilroy

Unified School District sought to create a collaborative environment in

which teachers make corrections throughout the year. Meeting in “data

teams,” teachers compare notes and plan adjustments in their instruction.

Nor did the shift happen overnight. At first, the district pursued a

top-down management strategy to establish a common language about

teaching and learning across its 14 schools. Teachers sometimes felt

stifled by the approach.

Now, they’ve got more breathing room. Instead of emphasizing the

use of specific teaching methods, the district is stressing a common

process of improvement planning at all schools. Teachers are held

accountable more for their results than for how they teach.

Forward Motion


By J e f f A r c h e r

Superintendent Edwin Diaz, who has led the district since 2000,

says he couldn’t have given schools the flexibility they have

now without first getting everyone on the same page. But he

also doubts that sustained improvement would result if the

central office continued to call most of the shots.

“We needed to shift from teachers just being compliant about

implementing the strategies that they were trained in, to actually having

to make decisions about which of those strategies to use and when,” he

says. “Because we think that’s where you get the next big level of growth.”


The California district is not alone inseeking the right balance between site-based management and centralizeddecisionmaking, a key issue assuperintendents concern themselvesmore with matters of instruction.Gilroy’s experience, in fact, shows how that balance can change over time.

Nestled in a fertile valleysurrounded by green hills,Gilroy calls itself the garliccapital of the world. Whileagriculture continues to be a

major employer, a growing number ofnew residents work in the high-techindustries of San Jose, about a 30-minute commute to the north.

For the school district, thatdemographic trend presents specialchallenges. Half of its students stillcome from low-income families, andabout a third are learning English. Butthe influx of high-paid professionals hasbrought with it new expectations abouteducational attainment.

Six years ago, all of Gilroy’selementary and middle schools hadmagnet programs that drew studentsfrom across the district. Each chose its own areas of academic focus,instructional philosophy, and schedule.Adding to the extent of variation amongsites was a conflict between the schoolboard and its then-superintendent that,by many accounts, left a leadershipvacuum at the central office.

Gilroy was among the lowest-performing of the 33 districts in Santa Clara County. In 2000, half of itsschools failed to meet theirimprovement targets on the AcademicPerformance Index, California’s test-based accountability system.

Diaz, a Gilroy native, saw a clear linkbetween the district’s decentralizationand its lackluster performance. In hisfirst year as superintendent, he visitedclassrooms in every school and saw greatvariation in the content and level of rigorof what was taught.

“I can remember one of ourelementary schools that had some ofthe lowest test scores,” he says. “It wasa math and science magnet. They hadlittle 2nd and 3rd graders in thesewhite lab coats running around, butnone of the kids were at grade level inreading.”

Diaz began by phasing out themagnet programs in favor ofneighborhood schools. Thedistrict rewrote its curriculumto cover what the state

expected students to learn. Newtextbooks were bought, and thesuperintendent made clear that theywere to be used.

For teachers’ professionaldevelopment, Gilroy picked a program bya Fresno, Calif.-based company calledLitConn. The training shows how todifferentiate literacy instruction basedon each student’s skill level. Althoughdesigned for teaching non-nativespeakers of English, district leadersbelieved all students would benefit.

Diaz wanted all teachers to go throughthe yearlong course, with the exceptionof high school math teachers, whoreceived other training. Accomplishingthat meant rewriting school budgets toinclude literacy facilitators—teacherswithout classroom duties who lead theLitConn seminars and provide on-sitecoaching. Every school has at least onefacilitator.

“I think we have some well-documented science now on what are themost effective practices that we can saywill result in greater student gains,” saysJacqueline M. Horejs, the assistantsuperintendent for educational services.“If that’s the case, then it’s importantthat everybody in the district knowswhat they are and implements themconsistently.”

To make sure that happened, thedistrict organized “walkthroughs,” inwhich administrators made classroomobservations. Armed with checklists,they looked for such techniques as“linkwords,” a visual aid for teachingvocabulary, and “partner talk,” in whichstudents consult one another beforegiving an answer.

In 2003, three years into the district’scampaign to bring about instructionalconsistency, not one of Gilroy’s schoolsmissed its state improvement target.Overall, the system posted greater gainsthat year on the performance index thanall but one other district in the county.

Teachers’ reactions to the district’sefforts were mixed, however. Some sawbenefits in using similar tools. Theapproach meant, for instance, that theyall knew the same way of assessing astudent’s reading abilities. Othersresented being told what to do. A few

AT A GLANCE:

Gilroy UnifiedSchool District

Superintendent: Edwin Diaz

Enrollment:10,000 students

47 percent Latino 31 percent white8.1 percent Asian

8 percent African-American30 percent

English-language learners

Poverty:52 percent come

from economically disadvantaged homes

Results: The proportion of students

proficient or above on statelanguage arts tests leveled off

in 2003-04, but scoresjumped again in 2004-05.


Socioeconomically Disadvantaged Students

Proficient and Advanced

2001-02 13.5%2002-03 18.52003-04 18.12004-05 23.8

Hispanic Students


2001-02 17.5%2002-03 22.62003-04 22.42004-05 28.6

All Students


2001-02 29.3%2002-03 33.82003-04 33.12004-05 39.0

called the walkthroughs “drive-bys.”Few within the district dispute

that the efforts prompted teachers toemploy the same practices. Not onlydid they learn the same methods intheir seminars, but their literacycoaches also arranged for them tovisit their colleagues’ rooms to seethem in action.

“We didn’t always agree with themodeling,” says Theresa Graham, a5th grade teacher at Antonio DelBuono Elementary School. “But atleast we all got together to see whatthe district wanted us to do.”

As soon became evident, thedistrict’s strategy hadlimitations. In 2004,Gilroy’s performanceleveled off. Half of its

schools again missed theirimprovement targets. Some sawtheir scores drop.

Diaz agreed that teachers neededmore discretion. At the same time,he feared going back to the dayswhen teachers acted likeindependent contractors, usingwhatever methods they felt mostcomfortable with regardless ofwhether they worked.

For help in finding a solution, thesuperintendent turned to outsideconsultants from the Center forPerformance Assessment. TheDenver-based group advisesdistricts on creating accountabilitysystems aimed at continuousimprovement.

The result was a new planningprocess for all schools that went intoeffect a year ago. Each site annuallydrafts a document that summarizesareas of greatest need and strategiesfor addressing them. The plans citetest scores and list performanceobjectives for the coming year.

At Rod Kelley Elementary School,for example, staff members promised,among other commitments, a 12percent jump in the number ofstudents who scored as proficient on adistrict writing assessment. Part oftheir plan to achieve that target was togive students clearer descriptions ofgood writing.

Within schools, similar planningtakes place in the new data teams,made up of the teachers in each

grade or department. Every fewweeks, they meet to compare how oneanother’s students are performing ona specific skill, and to brainstormways to improve.

Key to the district’s new approachwas a change in the use of studentassessments. After becomingsuperintendent, Diaz began givingdistrictwide tests three times a yearto gauge students’ progress. Lastyear, he gave schools more freedom todecide which tests to use, whilemaking it clear that they were to usetests more often.

“We had a lot of assessments thatwere in place mainly for monitoring,”

he says. “I think we had a real gap inthe assessments taking place in theclassroom that were actuallyresulting in a different lesson thefollowing day.”

At least one teacher from eachteam got trained on data-drivendecisionmaking. The StupskiFoundation, a Mill Valley, Calif.-based group that has givenplanning and financial assistance tothe Gilroy schools, paid for about 20educators to visit the Norfolk, Va.,school system, a district known forthe technique.

Teachers’ planning rooms herenow are adorned with devicesfor tracking progress. At oneschool, they’ve producedcolor-coded computer

spreadsheets of students’ scores onmultiple tests. At another, they usePost-It notes and poster boards toshow which students have masteredwhich skills.

“There’s a constant conversation

about what’s working, and what’snot,” says Graham, the teacher at DelBuono Elementary. “It’s ‘What didyou do? How did you get that tohappen?’”

Graham says her data team playeda big role in improving 5th gradereading instruction at her school lastyear. By sharing ideas on teachingpupils how to recognize an author’spurpose, they all saw jumps inreading performance over the courseof several weeks.

Teachers still have concerns.Last spring, the Gilroy TeachersAssociation filed a grievance againstthe district, arguing that data-drivendecisionmaking takes more planningtime than the teacher contractallows. But many teachers favor thecollaborative approach over thedistrict’s earlier efforts.

“It’s just now feeling like it’sstarting to smooth out, and feelnormal,” says Heidi Jacobson,who also teaches 5th grade at Del Buono.

The district's recent efforts appearto have paid off. State test resultsreleased last month showed that,after plateauing in 2004, the portionof Gilroy students achieving atproficient levels in English languagearts jumped six percentage points in2005—the highest such annual gainsince Diaz arrived.

Feeling validated, district leadersmake no apologies for their earlierdirection on instructional methods.Teachers needed to learn a new set oftools to use when students werestruggling, they argue.

“We think we’re in a muchdifferent position now,” says Horejs,the assistant superintendent foreducational services.

Even with the emphasis on school-based decisionmaking, many hereexpect that teachers will continue touse many of the same instructionalmethods. The difference, they say, isthat it will be because teacherschoose to do so.

“I don’t think we’re ever going to swing back to everyone doingtheir own thing,” says TriciaSatterwhite, a literacy facilitator atAscencion Solorsano Middle School.“I think they’ve equipped us, they’vetrained us, and they expect us towork as a team.” ■

Teams of teacherslook closely atdata on studentachievement toplan instruction.


Clarksville, Tenn.

Beth G. Unfried’s role as the principal of Kenwood Elementary

School has changed dramatically in the past four years.

The expectations in the 26,000-student Clarksville-

Montgomery County district began to shift in July 2001,

when Sandra L. Husk became the superintendent. On Husk’s first

visit to the principal’s 830-student school here, Unfried says, “She

started asking me these questions. Like, ‘What do you do when a

child doesn’t learn?’ and ‘How do you know when a child is mastering

these skills?’ ”

“When she left, I felt so lousy,” remembers Unfried, who found herself

struggling to answer. “But today, I can say with confidence that I am

the instructional leader of this building.”

Before that time, the principal says, “I was the facilities manager,

and I was busy. I felt good about what I was doing. I was keeping the

peace. Sometimes we made good test scores, sometimes we didn’t, but

everyone was happy.”

The same cultural transformation is taking shape across this

Tennessee school district, located just over an hour north of Nashville,

thanks to a concerted effort from the top that has focused on making

student achievement the centerpiece of everything the district does.

While the transition is far from finished, it illustrates the powerful

role that district leadership can play in improving teaching and

learning in classrooms.


When theClarksville, Tenn.,schools raisedexpectations forlearning, the centraloffice played aleading role instandardizingpractice andmonitoring data.

By L y n n O l s o n

In Sharp Focus

Founded in 1784, Clarksville, along the banks of the Cumberland

and Red rivers in north-central Tennessee, was the state’s first

incorporated city. Its sleepy downtown, dominated by law offices

and antiques stores, boasts restored brick sidewalks, period

lighting, and historic architecture. But the county’s small-town

atmosphere is giving way to rapid development along Interstate 24 at

the Tennessee-Kentucky border, fueled, in part, by the nearby Fort

Campbell and the relatively easy commute into Nashville.

Superintendent Husk describes her first fall in the district as chaotic.


One of her first tasks was to cut thebudget by $4 million. In September, twoplanes toppled the twin towers of theWorld Trade Center, and thousands oflocal U.S. Army families began sayinggoodbye to husbands and wivesdeployed overseas. It was the year thestate began enforcing mandatory class-size limits in grades K-12. And 2001was the same year Tennesseelawmakers required that test scores bereported separately for different racialand ethnic groups.

Suddenly, a district that had coastedjust above average on state testsdiscovered that African-American andlow-income students, in particular, werenot doing very well. “I don’t think that’sa major surprise,” says Husk, “but it wassomething that had not been seriouslytalked about.”

One of her first moves was toreorganize the central office. Well-respected principals were promoted tobecome the chief academic officer andthe directors of human resources,instruction, and assessment.

Sallie Armstrong, the director of thenewly created office of curriculum andinstruction, began working withteachers to arrange the statecurriculum into what is called a scopeand sequence, including units of studyand a pacing guide to show whatteachers should be teaching.

“We didn’t have a guaranteed andviable curriculum because no one hadled instruction that way,” saysArmstrong. “Teachers were going fromthe front of the textbook to the back ofthe textbook.”

In 2002-03, Clarksville-MontgomeryCounty embarked on a districtwideliteracy initiative, including the use ofthe same reading textbook for all itselementary schools. It also adopted a

common approach to teaching writing inkindergarten through 12th grade, the“Six Traits of Writing,” developed by theNorthwest Regional EducationalLaboratory in Portland, Ore.

To help carry out its “balancedliteracy” approach, the district hired 11full-time literacy coaches, who modelinstructional strategies in schools andwork with teachers to analyze student-performance data. The coaches visitclassrooms and consult with teachersdaily, provide professional development,and take part in school improvement

and literacy teams at schools. Threeconsultants provided initial trainingand continue to work with the districton the reading and writing programs.

Now, asked what KenwoodElementary’s motto is, a 3rdgrader pipes up: “Reading isthe cardinal rule.”

Students’ work, labeled bywhich state academic standard itaddresses, papers the hallways. Everyclassroom has a “word wall” ofvocabulary, writing folders for eachpupil, and a posted schedule of thestandards to be learned that day.

In one classroom, 3rd graders work in“literacy circles” to answer questionsabout Tales of a 4th Grade Nothing byJudy Blume. In the gymnasium, youngerchildren race across the floor to slap a“high five” on large letters laid on theground, as they spell out a word postedon the overhead projector. In the artroom, students draw illustrations basedon the book Why Is Blue Dog Blue? byGeorge Rodriguez. When they’refinished, they can go to the readingcenter, tucked away in a corner, to readindependently.

Schoolwide events celebrate children’sreading success, such as a “vote forbooks” campaign that culminated in adelegates’ assembly where studentsvoted for their favorite books.

The school has added two readingspecialists and a reading lab, staffed bythree teacher aides, to work withchildren individually and in groups.

Grade-level teams meet weekly tofocus on specific problems or issues. Theyturn in their minutes to the principal.Teachers also have volunteered to do“quick visits” to each other’sclassrooms—initially focused on whetherthe classroom environment showedevidence of the district’s reading andwriting initiatives; more recently, to lookfor evidence of good instructionalstrategies.

Unfried, the principal, has gotten lotsof support for such changes. Along withabout 80 of her colleagues—drawn fromboth the instructional andnoninstructional ranks of districtemployees—she attended a yearlongLeadership Academy to build leadershipcapacity across the school system. Lastschool year, she and other buildingprincipals—along with selected assistant

AT A GLANCE:

The Clarksville-Montgomery

CountySchool District

Superintendent: Sandra L. Husk

Enrollment: 26,000 students

65 percent white28 percent African-American

5 percent Hispanic

Poverty: Nearly 40 percent come

from economically disadvantaged homes

Results: The proportion of students

proficient or above on statelanguage arts tests increasedin the following categories:


Elementary and Middle School

Proficient or above

2003 88%2004 912005 95

2003 83%2004 872005 91

2003 82%2004 852005 92

2003 85%2004 912005 91

African-American

Proficient or above

Economically Disadvantaged

Proficient or above

Hispanic

Proficient or above

principals and teachers—alsoparticipated in “professional learningcommunities,” led by five of theircolleagues, to encourage collaborativeproblem-solving focused aroundteaching and learning. This schoolyear, all principals are fosteringprofessional learning communities intheir own schools, modeled on whatthey have learned.

“The work that the directors dowith the principals looks very similarto the work that principals are doingwith teachers and that teachers aredoing with students,” explains Husk,who argues that the system had tocreate a cohesive culture focused onhigh expectations. “So it’s pervasive,not just in the classroom andthroughout the school, but verticallyin the organization as well.”

That unifying focus, acrossboth the instructional andnoninstructional sides of thehouse, is most evident in thedistrict’s use of data to

improve the efficiency andeffectiveness of its practices.Clarksville-Montgomery County, infact, has earned certification underan internationally recognizedstandard for companies thateffectively monitor and adjust theirwork processes based on data.

Every department and everymajor initiative is driven by suchfeedback, the superintendent says.“I think we have to expand how wethink about data,” she adds. “It’s notjust annual high-stakes studenttests. It’s much broader and deeperand richer.”

Last year, for example, the districtset a goal of delivering all newlyadopted textbooks to schools by thefirst day of class. While it didn’tquite reach that target, 88 percent oftextbooks were delivered by day one,and 100 percent by the third day ofschool. The maintenance departmentset a goal of completing all workorders within 20 days on average.

By the end of last school year, itwas down to six days. Under acontract with Kelly Services, of Troy,Mich., which provides temporarystaffing, the district improved theproportion of classrooms staffed byqualified substitutes when teachers

are absent from 93 percent in 2003-04 to more than 97 percent by theend of last school year. “Thecomplaints just went away,” saysBruce Jobe, the director of humanresources. “It’s changed overnight.”

That same focus on data permitsthe district to track the fidelity withwhich schools carry out instructionalstrategies. All of its majorinitiatives—such as the writingprogram—were designed around athree-year implementation schedule,with benchmarks set for each yearabout the types of activities that

should be observed in classrooms,initial and intermediate outcomes,and the evidence schools shouldprovide if the program is on track.

The district relies upon such toolsas principals’ ratings, surveys ofclassroom teachers, daily activity logs kept by its literacy coaches,and classroom observations by itsconsultants and instructional-management team to monitor what’shappening.

Clarksville-MontgomeryCounty is less experiencedthan some other districts in its use of periodicassessments to help

inform instruction, often known as “benchmark” tests. But it is moving fast.

In the 2004-05 school year, it gaveits first benchmark assessments inlanguage arts and math in grades 3-10, aligned to state academic-contentstandards. Those tests were beingrevised over the summer. This schoolyear, the district planned to addscience tests. The tests are giventhree times each year and are builtwith the help of teachers andprincipals. Educators have access to

the results online, through a Web-based assessment platform developedby the San Francisco-based Edusoft.

Clara Patterson, the district’sdirector of educational services, saysstate test results are simply toolittle, too late, “so we started talkingabout the fact that we needed aconsistent way of monitoring, on aregular basis, where children are atthis point in time.”

Through the Web-based system,teachers also have access to statetest results and information fromDIBELS, the individualized, diagnosticreading assessment used by manydistricts across the nation. They alsocan build their own classroomassessments.

Margie Ford, the principal ofNorman Smith Elementary School,says her school’s literacy coachprovided an item analysis of the testresults to each teacher last schoolyear. “We’re going to take it a stepfurther this year,” she adds. Teamsof reading and math teachers, acrossgrades K-5, will meet after eachbenchmark test to review the scoresand come up with schoolwidesuggestions for addressing anyweaknesses. The school also plans todevelop short, monthly reading andmath tests, through Edusoft.

“We’re headed in the rightdirection because our system nowhas a vision,” says Smith, who’sheaded the school for 15 years, “andwe’re working as a whole.”

That vision is starting to produceresults. All of the district’s 30 schoolsmet their achievement targets underthe federal No Child Left Behind Actin 2004-05. Based on Tennessee’s“value added” model, which examineshow much growth individualstudents make from one year to thenext, the school system showedabove-average growth for its studentsin language arts and math, andexceptional gains in social studiesand science.

Kenwood Elementary studentsmade far more progress thanexpected in both math and reading.

“This was a totally different worlda few years ago,” says PrincipalUnfried, recalling the previous lackof focus on instruction.

Now, her tone suggests, things arelooking up. ■

Clarksville usesmany types of datato monitor andimprove its practices.


LEADING LEARNING - Education Week · district level for its second annual “Leading for...

Documents

Transcript of LEADING LEARNING - Education Week · district level for its second annual “Leading for...