Education Sector: Some Assembly Required

7/31/2019 Education Sector: Some Assembly Required

1/12

education sector reports

www.educationsector.org

Some Assembly Required:Building a Better AccountabilitySystem for California

By Kevin Carey


2/12

ACKNOWLEDGEMENTS

This report was unded by the Stuart Foundation. Education

Sector thanks the oundation or its support. The views

expressed in the paper are those o the author alone.

ABOUT THE AUTHOR

KEVIN CAREYis director o the education policy program

at the New America Foundation and ormer policy director

o Education Sector. He can be reached at carey@

newamerica.net.

ABOUT EDUCATION SECTOR

Education Sector is an independent think tank that

challenges conventional thinking in education policy.We are a nonprot, nonpartisan organization committed

to achieving measurable impact in education, both by

improving existing reorm initiatives and by developing new,

innovative solutions to our nations most pressing education

problems.

Copyright 2012 Education Sector

Education Sector encourages the free use, reproduction, and dis-tribution of our ideas, perspectives, and analyses. Our CreativeCommons licensing allows for the noncommercial use of allEducation Sector authored or commissioned materials. We require

attribution for all use. For more information and instructions onthe commercial use of our materials, please visit our website,www.educationsector.org.

1201 Connecticut Ave., N.W., Suite 850, Washington, D.C. 20036202.552.2840 www.educationsector.org


3/12

1 Education Sector Reports: Some Assembly Required July 2012 www.educationsector.org

But when the bill reached the governors desk, he

pulled out his veto pen. When Browns rst stint as

governor o Caliornia ended in January 1983, the

states public education system was still seen as a

national model. The strict property tax limitationsand broader anti-tax movement birthed in the

Golden State in the 1970s were just beginning to

erode the nancial oundations o the states public

schools.A Nation at Risk, the landmark critique o

American public education, would be released later

that year, launching an era o bipartisan support

or standardized testing and school accountability,

culminating with the 2001 No Child Let Behind Act

(NCLB) and state-managed systems like the API.

The way people thought, acted, and spoke about

education policy had changed dramatically in theprevious three decades, and, it seemed, very little o

it was to Browns liking. During his 2010 gubernatorial

campaign, he spoke out against standardized testing.2

This led some to believe that Brown would support

the broad changes in SB 547, which reduced the

infuence o such tests on school accountability. They

didnt realize that, to Brown, those changes were not

nearly enough.

In a two-page veto message, Brown declared:

This bill is yet another siren song o school reorm...

while SB 547 attempts to improve the API, it

relies on the same quantitative and standardized

paradigm at the heart o the current system. The

criticism o the API is that it has led schools to

ocus too narrowly on tested subjects and ignore

other subjects and matters that are vital to a well-

rounded education. SB 547 certainly would add

more things to measure, but it is doubtul that it

would actually improve our schools. Adding more

speedometers to a broken car wont turn it into a

high-perorming machine.

Over the last 50 years, academic experts have

subjected Caliornia to unceasing pedagogicalchange and experimentation. The current ashion

is to collect endless quantitative data to populate

ever-changing indicators o perormance to

distinguish the educational good rom the

educational bad. Instead o recognizing that

perhaps we have reached testing nirvana,

editorialists and academics alike call or ever more

measurement visions and revisions.

On October 8, 2011, California Gov. Jerry Brown took a stand.

Throughout the 2011 session of the California General Assembly,

Senate President Pro Tem Darrell Steinberg had been pushing

legislation designed to revamp the states system for holding K-12schools accountable for student success. Californias Academic

Performance Index (API) system hadnt been updated since 1999 and

relied mostly on standardized tests of basic prociency in reading

and math. Steinbergs bill, SB 547, would have changed the system

to include graduation rates and measures of career and college

readiness.1 The bill passed both the Assembly and Senate by wide

margins and with bipartisan support, in addition to the backing of

diverse organizations including business groups, charter school

operators, and school administrators.


4/12


A sign hung in Albert Einsteins oce read Not

everything that counts can be counted, and not

everything that can be counted counts.

SB 547 nowhere mentions good character or love

o learning. It does allude to student excitement

and creativity, but does not take the qualitiesseriously because they cant be placed in a data

stream. Lost in the bills turgid mandates is any

recognition that qualityis undamentally dierent

rom quantity.

In his letter, Brown oered no real solutions to actually

improve Caliornias public schools. His aversion to

all things quantitative bordered on anti-empiricism

and was hard to square with the realities o creating

policy or a state that educates more than six million

students every day.

But the letter correctly identied the major dilemmas

acing education policymakers today, and not just in

Caliornia. Even as legislators and executives parry

in Sacramento, a similar debate has been playing out

in the White House, in Congress, and in statehouses

nationwide.

Browns lament and the broader debate come down

to three questions. First, what kind o inormation

should be used to make judgments about the success

o educators and educational institutions? Second,

what is the best way o interpreting that inormation?

Third, having interpreted the right inormation in theright way, what should be done to help more students

learn?

Yet even as pundits, politicians, and interest groups

have spent the last decade waging an oten-

ideological battle over school reorm, educators

and public ocials in states and districts across

the country and overseas have been building a

new oundation o inormation and practice that

could, i implemented correctly, satisy even the

most suspicious, Brown-like critics o education

accountability. The pieces are alling into place. All

that remains is or public leaders to put them together

in the right way.

GATHERING MOREMEANINGFUL DATA

The Caliornia API system is based primarily on

student test scores in reading/language and math.

While science and social science tests are also

included, the results receive less weight in the overall

API score than do reading and math, particularly or

elementary and middle schools. In this respect, the

system is very similar to NCLB, which requires annualreading and math tests every year in grades three

through eight and once in high school, as well as a

token science assessment in grades three through

ve, six through eight, and nine through 12.

The problems with this approach are obvious. We

expect our public schools to accomplish ar more

than to inculcate reading and math skills. In addition

to subjects including history, science, literature, arts,

oreign language, and physical education, schools

are expected to teach character, discipline, and

problem-solving, and to produce critical thinkers andcommunicators who are prepared to thrive in higher

education, the workplace, and civic lie. Accountability

systems that ignore these goals run the risk o

distorting educational decision-making by creating

incentives or educators to unduly ocus on the

subjects that are part o the system at the expense o

those that are not.

Math and language were not randomly chosen as

the ocus o school accountability, o course. They

represent the oundational skills on which most other

learning depends. Indeed, there is little evidencein national test score data to suggest that student

learning in subjects such as civics, geography,

science, and history has suered as a result o

accountability-driven preerence or reading and math

under NCLB. American ourth- and eighth-graders

scored better in history in 2010 than in 2001, or

exampleperhaps because its hard to learn history i

you dont know how to read.3

The pieces are falling into place.

All that remains is for public

leaders to put them together in

the right way.


5/12


Nonetheless, accountability systems that ignore

much o what schools expect or students suer rom

a undamental lack o legitimacy. Systems seen as

illegitimate tend to produce antagonistic relationships

with those being held accountablenot the best

environment or producing sustained excellence. The

problem, however, is that accountability systemsrequire the collection o valid, comparable inormation,

and the main tool available or gathering such

inormation is standardized testing. Tests present

problems o their own: they are expensive, time-

consuming, and risk ocusing schools on mindless

test-prep instead o authentic learning. Plus, Einstein

was right: tests are better at measuring some things

than others.

But the debate over which subjects and measures

should be included in accountability systems oten

misses a larger reality: elementary and secondary

education is a means to an end. This is not to say

that learning is not valuable or its own sake. But the

primary goal o educating children is to prepare them

to be successul learners as adultsto teach them

how to think and give them the basic knowledge

to acquire deeper expertise. Thats why the large

majority o states have moved to adopt the CommonCore State Standards, which were explicitly designed

to represent the knowledge and skills that our young

people need or success in college and careers.4

When the API and NCLB systems were being

developed in the late 1990s and early 2000s, building

accountability systems around direct measures

o student success in higher education and the

workplace was an undertaking ar beyond the logistic

capacity o most states. At that time, many states

didnt even gather basic academic inormation about

their students in a consistent manner. School, district,

and state education bureaucracies hadnt ully made

the transition rom paper to electronic records. Only a

handul o states had so-called student unit-recorddata systems that could calculate, or example, what

percentage o students graduating rom a given high

school went on to succeed in college or a promising

career.

But things have changed. Through a combination

o state and ederal investments and the relentless

march o inormation technology, most states now

have sophisticated data systems that ollow students

as they travel through the public school system into

college and the workplace.

This means that states no longer have to rely solely

on proxy measures or college and career readiness

such as standardized test scores. They can nd out

whether students actually succeeded in college by

tracking measures including college entrance rates,

persistence in higher education, and the percentage

o students who are orced to take non-credit

remedial courses in college. Similarly, states can use

Department o Labor and other data systems to nd

out what happens to students who leave the K-12

system and go into the workorce and the armed

services.5

Shiting rom proxy measures o preparation to actual

measures o success has the advantage o reducing

the amount o inormation accountability systems

need to gather. A lot goes into preparing someone

to succeed in college. Students need oundational

language and mathematics skills, the ability to

communicate orally and in writing, personal discipline,

and a acility or working with those o diverse

backgrounds. Rather than laboriously constructing

methods o assessing every one o these things

bearing the cost in time and money, and adjustingor errors o measurement reliability and validity

accountability systems can ocus on a smaller number

o bottom-line outcome measures.6

Figure 1, which appeared in a recent Education

Sector report by Anne Hyslop and Bill Tucker,

shows how inormation about college success can

signicantly alter the picture o K-12 success.7

Shifting from proxy measures of

preparation to actual measures

of success has the advantage

of reducing the amount of

information accountability

systems need to gather.


6/12


One o these demographically similar Caliornia highschools scored 778 on the API, near the 800 threshold

beyond which schools are essentially exempt rom

accountability. The other scored much lower, at 698.

But the school with the substandard API has a higher

graduation rate and ar more o those students enroll

in college. This isnt just about having more typical

college-bound studentsthe dierence in the college-

going rate among low-income students is larger still.

Brown, however, has implemented or proposed

unding cuts to several dierent Caliornia agencies

that collect education inormation.

8

The governor hasspoken o the need to create education inormation

that serves local educators making decisions on the

ground. Such inormation is denitely needed, but it

shouldnt come at the expense o centrally collected

inormation that allows or statewide analysis and

comparison. Such cuts represent a pound-oolish

approach to government. Inormation is cheap in

the grand scheme o things and vital or tracking

long-term educational results. In holding schoolsaccountable, Caliornia should use outcome measures

including rates o college remediation, persistence,

credit accumulation, and degree completion. It should

also track completion o vocational training and

apprenticeship programs, military enlistment, and

attainment o proessional licenses and certication,

to gauge success among students who go directly

into jobs.

WORKING TOWARD BETTER

DATA INTERPRETATIONGathering better inormation about student outcomes

will improve educational accountability systems.

But thats only the rst stepthe next is interpreting

the inormation. Speedometers are easy to read;

spreadsheets o student test scores less so. And one

o the crucial dimensions o educational perormance

data embedded in NCLB and the API is time.

Figure 1. Similar Schools, Different OutcomesWith more informationhigh school graduation, college enrollment, and enrollment for low-income studentsa

revised accountability system could alter current perceptions of school performance.

Source: Caliornia Department o Education.

Note: College enrollment rates include graduates that enrolled in public and private postsecondary institutions nationally within 16 months o highschool completion.

API SCORE GRADUATIONRATES

COLLEGEENROLLMENT

LOW-INCOMECOLLEGE ENROLLMENT

-80 +4.5% +19.8% +25%

77895.1%

85.7%

65.9%

81.5%

56.5%

90.6%698


7/12


American schooling is organized around units o

time, in particular the year. By convention, the

circumstances in which students learntheir peers,

their teachers, the curriculum they ollow, the building

in which they studytypically change every year.

Whether this is actually a good idea deserves more

scrutiny than it receives. But it is the way o things,and accountability systems have been designed

accordingly. NCLB and state accountability systems

all render their judgments annually.

They also ocus on absolute measures o student

learning. This is a change rom prior practice,

when students were generally evaluated on norm-

reerenced tests like the SAT-10, which yields scores

in percentiles: how students ared relative to other

students who took the test. The newer, NCLB-based

accountability systems used criterion-reerenced

tests designed to gauge student mastery o aparticular domain o learning. States dened a certain

level o mastery as procient and students met that

level or did not, irrespective o the perormance o

their peers.

The switch to criterion-reerenced exams and

prociency rates refected the conviction that there are

certain things students must know to succeed in lie,

college, and careers. It ollowed, among the designers

o NCLB, that educational accountability systems

should embody that conviction. For many, this belie

rose to the level o moral urgency: an accountabilitysystem that did not judge schools by the percentage

o students procient on a criterion-reerenced test

did not refect a legitimate commitment to the notion

that all students can learn.

The problem came when absolute measures o

student learning and annual evaluation o students,

schools, and districts intersected. Because, o course,

an absolute measure o what students know at a given

point in time refects everything that has happened

to them beore, not just what has happened to them

over the prior year in school. The obvious remedy wasto measure only how much they had learned during

the relevant time period: growth in learning over the

previous year. But as with data systems that track

student progress rom high school to college and

the workplace, the technical inrastructure needed

to calculate growth measures was lacking in many

states at the time o NCLB and API implementation.

Students werent tested in every grade, nor were there

data systems to know that, or example, the John

Smith who scored 710 on the ourth-grade math test

in one school district was the same John Smith who

moved with his parents to a new district on the other

side o the state and scored 750 on the th-grade

math test.

That, too, has changed. Most states can now track

annual progress among students on standardized

tests, which allows them to calculate growth

measures o student progress. As Richard Lee Colvin

recently wrote or Education Sector, several high-

poverty schools in Los Angeles provide excellent

examples o how growth measures open up an

important new lens on school perormance.9 Ater

a change in leadership in 2009, Audubon Middle

school in Los Angeles showed little improvement

in terms o overall prociency. Only by examining

individual student growth was the school district ableto determine that students were making substantial

progress at Audubon. They had just started so

ar below the prociency line that they hadnt yet

caught up. Instead o closing or radically overhauling

Audubon, the district worked to support the eorts

that had led to recent success.

During the 2000s, the Bush administration allowed

15 states to experiment with adding growth

measures to their NCLB accountability systems.

They required states to use a standard called

growth-to-prociency, which essentially measuredwhether growth among non-procient students was

rapid enough to bring them to prociency within a

relatively short time period, usually three or our years.

This refected a legitimate ear among advocates

or traditionally underserved students that unless

the growth measure was anchored to prociency,

schools would never be accountable or helping low-

perorming students catch up. I John Smith moves

to a new school district and his th-grade teachers

determine that, due to a substandard elementary

school education, he is reading at only the second

grade level, this represents a signicant and arguablyunair burden on the new district. But it doesnt

change the act that John Smith has no other district

to educate him, and that he needs to learn to read.

The growth model experiments had little eect

relatively ew low-perorming schools were getting

enough growth to put students on a short-run

prociency trajectory.10 Few but not all, and


8/12


create school-level growth measures grew during

the 2000s, yielding a variety o approaches to

incorporating the inormation into accountability

systems. One model subsequently adopted by

a number o states was pioneered in Colorado,

which created a user-riendly interace on the state

Department o Education website designed to show

how dierent schools compared when viewed through

lenses o prociency and growth simultaneously.

The schools in the lower right-hand quadrant o

Figure 2, which was generated rom the Colorado

Department o Education website, exhibit unusually

accountability systems that ail to recognize schools

achieving the greatest success in the most dicult

circumstances are, by denition, broken. There is an

aspirational element to any accountability system, a

conviction that schools can be better. I schools that

actually meet this goal are not recognized as such,

the accountability system suers a crisis o legitimacy,

just as it does when accountability ignores many o an

education systems most important goals or student

learning. And systems seen by the governed as

illegitimate tend not to work very well.

The number o states with the technical capacity to

Source: https://edx.cde.state.co.us/growth_model/public/index.htm#/year-2010, accessed May 3, 2011.

Figure 2. Denver School Performance - 2010


9/12


low prociency but unusually high growth. They

are Audubon-like, low-perorming or a variety o

historical, managerial, and demographic reasons,

but making swit progress relative to other, similar

schools.

USING DATA TO HELPSTUDENTS

What should happen when schools like Audubon

perorm as they do? And what about all the other

schools that lie elsewhere on the distribution

o prociency and growthnot to mention the

distribution o graduation rates, success in college

and the workorce, and other measures deemed

important enough to include in accountabilitysystems? This is a crucial questionindeed,

accountability systems exist or no other reason than

to pose it. And there are many possible answers.

Some say the inormation should simply be made

publicly available, so that parents, educators, and

policymakers can act accordingly. Others see

transparency alone as grossly insucientmany o

the schools identied as ailures by NCLB and API

were given similar labels by previous accountability

systems. Lack o student learning in many distressed

schools isnt exactly a secret that can only be

uncovered by clever accountability design.

NCLB and, to various extents, most state-designed

accountability systems, answer the what should

happen question through the application o rules

that lead to consequences or schools. Some o

those rules are interpretive, designed to distinguish,

in Browns words, the good rom the bad. NCLB

creates a standard o adequate yearly progress

and prescribes consequences or schools that ail to

meet the standard or a certain number o consecutive

years. (API assigns every school a score on a scale

rom 200 to 1,000 and mandates interventions

when scores are persistently low.) Some NCLB

consequences are automaticgiving students the

right to transer to a higher-perorming school in the

same district, or example. But or the most part, the

consequences simply trigger some kind o orced

choice among state and local policymakers, who

are required to select rom a menu o serious, semi-

serious, and not-very-serious interventions.

This model has not worked all that well.

Implementation o NCLB and state accountability

systems has revealed a strong inverse correlation

between the number o school employees likely to

lose their jobs as a result o an intervention and thelikelihood o state and local policymakers employing

it. In other words, i you give people the option o

making hard choices, they tend not to.11

Its easy to chalk this up to recalcitrance or

insuciency o resolve. But in airness to state and

local education ocials, NCLB asked them to make

decisions they were not trained or well-positioned

to make, based on an accountability system o

questionable legitimacy that they did not design. This

is not a recipe or success. There is a better way,

one that puts the hands o interpreting accountabilityinormation into the hands o people trained to use it:

school inspections.

On Her Majestys School Inspection Service

The underlying problem with rules-based

accountability systems is that the need to gather

more inormation about the complex endeavor that

is public education and the intricacies o interpreting

that inormation in a air, consistent, nuanced, and

eective manner are in constant tension with oneanother. Put another way: it is simply impossible to

design an accountability system that contains enough

inormation and interprets that inormation eectively

through the exclusive use o dened rules. Indeed,

the two most commonly voiced criticisms o the

NCLB accountability regime directly contradict one

another: it is said by critics to be both complex to the

point o inscrutability and a crude, one-size-ts-all

punishment machine.

The answer is to rely less on rules and more on highly

trained human judgment. In his veto letter, Browntouched on this idea by saying:

There are other ways to improve our schoolsto

indeed ocus on quality. What about a system

that relies on locally convened panels to visit

schools, observe teachers, interview students,

and examine student work?


10/12


Browns instinct to rely more on human judgment

or decisions about school improvement is sound. It

just needs to be trained human judgment, working

in a system with clear guidelines and a high level o

expertise.12

Such a system has been established in the UnitedKingdom or some time. As Craig Jerald wrote in

the 2012 Education Sector report On Her Majestys

School Inspection Service, the Oce or Standards

in Education, Childrens Services and Skills (Osted)

has been managing a robust school inspection

system since 1993.13 Highly trained inspectors,

many o them ormer school administrators and

teachers, conduct intensive site visits that include

structured observations o the teaching and learning

environment. Schools are provided with specic

recommendations or improvement and are judged

on their progress in ollow-up visits. Standardized testscores are an important part o the process, guiding

inspectors and helping shape the determination o

whether a school is given special measures status,

which signies the need or immediate, substantial

improvement.

But in the end, it is the judgment o the inspectors

themselves that matters most. The Osted system

recognizes the inherent complexity o educational

institutionsa complexity that cannot be captured by

a Scantron machine and a rulebook. No ormula or

set o rules can be created to automatically parse andinterpret the ull range o inormation about student

learning and long-term outcomes. Nor can such rules

perorm the kinds o holistic evaluations o school

management and culture that good inspections entail.

Only people can do thati theyre the right people.

MOVING FORWARD: BUILDINGA GREAT SYSTEM FROM

WHATS ALREADY THERESen. Steinberg took the governors challenge

seriously and introduced a new version o his

accountability reorm bill this year. SB 1458 again calls

or the reormation o the API system, reducing the

contributions o standardized tests to 40 percent and

charging state policymakers with adding measures o

college and career readiness, student advancement

through grades to graduation, as well as, possibly, a

program o school quality review that eatures locally

convened panels to visit schools, observe teachers,

interview students, and examine student work. Many

o the essential elements o a great next-generation

accountability system in Caliornia are there, along

with existing law that allows the state to move towardthe use o growth models.

However, Steinbergs new legislation still refects the

underlying faw o both API and NCLB in translating

the assembly and interpretation o data into authentic

action or reorm.

In a 2011 Sacramento Bee op-ed, Steinberg and

then-state superintendent o public instruction Tom

Torlakson wrote:

Ask a baseball an how good his teams shortstop

is, and he can point to more than two dozen

statistics, rom the number o double plays turned

to how oten the player strikes out with runners

on base. Ask about the perormance o a public

school in Caliornia, and youll get one lonely

number based solely on one set o end-o-the-

year test results.14

The baseball analogy is apt, although perhaps

not exactly in the way Steinberg intended. Ask a

baseball an what statistics he really cares about,

and he will likely respond with one lonely number:wins. From the perspective o providing broad public

inormation about school success, the API and NCLB

systems have value, as do A-F grading systems

like those used in Florida and other report-card style

measures that boil down the complex dimensions o

educational perormance to comparable, easy-to-

digest numbers and grades. Parents dont want to

spend hours poring through dozens o statistics with

little guidance, just as only obsessive sports geeks

do the same or shortstops. Summary measures

have a valuable interpretive qualityas long as

decisions about what matters most are baked into theormulas. They should be continued and publicized,

ocusing primarily on absolute measures o student

perormance and long-term outcomes in college and

the workplace.

The biggest faw o API- and NCLB-style

accountability systems is their practice o eeding

synthesized, summative measures like those


11/12


described above into mechanistic, rules-based

systems or driving school improvement policies. This

hasnt worked, and it wont work in the uture. Making

the hard, complex choices about how exactly schools

are alling short and what exactly should be done to

improve them needs to be moved rmly into the realm

o impartial, inormed human judgment or, in otherwords, to inspection systems, like the Osted system

in England.

This would require a signicant but manageable

investment in building an inrastructure o highly

trained inspectors.15 Some will balk at devoting

scarce resources to educational spending that

is not inside the classroom, particularly during

dicult budget times. But the strategy o building

accountability systems on the cheap has proved

inadequate. Inexpensive standardized tests

and laundry lists o broadly dened and largely

unenorceable interventions cannot be combined

into systems that will actually create the kind o

steady, constructive pressure to improve that

American schools need. The cost o continuing suchailure would be vast.

Caliornia has an opportunity to use methods o

gathering, interpreting, and acting on accountability

inormation that already exist and have already been

proven to work in a way that will help the state move

its battered education system back onto a path

o national and world leadership. I the system is

designed correctly, even Gov. Brown should agree.


12/12


Notes

1. Caliornia now requires graduation rates to be included in theAPI, but the update has not yet been implemented.

2. Anthony Cody, Jerry Brown to Arne Duncan: Think Again!Blog Post, Education Week, September 1, 2009, http://blogs.edweek.org/teachers/living-in-dialogue/2009/09/jerry_

brown_to_arne_duncan_thi.html

3. The Nations Report Card website, http://nationsreportcard.gov/ushistory_2010/summary.asp

4. Common Core State Standards Initiative website, http://www.corestandards.org/

5. Anne Hyslop, Data That Matters: Giving High SchoolsUseful Feedback on Grads Outcomes (Washington, DC:Education Sector, 2011.) http://www.educationsector.org/publications/data-matters-giving-high-schools-useul-eedback-grads-outcomes. See also: Education andWorkorce Data Connections: A Primer on States Status,Data Quality Campaign http://www.dataqualitycampaign.org/fles/DQC%20Workorce%20Primer_2011_Format.pd

6. College success is, o course, partially the responsibilityo colleges. Integrating college results into educationaccountability systems points toward a uture where K-12and higher education systems are held jointly responsibleor student success in the crucial transition years rom highschool to college.

7. Anne Hyslop and Bill Tucker, Ready by Design: A Collegeand Career Agenda for California (Washington, DC:Education Sector, 2012) http://www.educationsector.org/publications/ready-design-college-and-career-agenda-caliornia

8. Josh Keller, Elimination o Caliornia Agency Could LimitAccess to Student Data, Blog Post, The Chronicle of HigherEducation, August 11, 2011, http://chronicle.com/blogs/wiredcampus/elimination-o-caliornia-agency-could-limit-access-to-student-data/32826

9. Richard Lee Colvin, Measures That Matter: Why CaliforniaShould Scrap the Academic Performance Index

(Washington, DC: Education Sector, 2012) http://www.educationsector.org/publications/measures-matter-why-caliornia-should-scrap-academic-perormance-index

10. Kevin Carey and Robert Manwaring, Growth Models andAccountability: A Recipe for Remaking ESEA (Washington,DC: Education Sector, 2011) http://www.educationsector.org/publications/growth-models-and-accountability-recipe-remaking-esea

11. Rob Manwaring, Restructuring Restructuring: ImprovingInterventions for Low-Performing Schools and Districts(Washington, DC: Education Sector, 2010) http://www.educationsector.org/publications/restructuring-restructuring.See also: Sara Mead, Easy Way Out, Education Next, 7:1(Winter 2007) http://educationnext.org/easy-way-out/

12. Caliornia has used visitation panels or many years.

13. Craig D. Jerald, On Her Majestys School InspectionService (Washington, DC: Education Sector, 2012) http://www.educationsector.org/publications/her-majestys-school-inspection-service

14. Tom Torlakson and Darrell Steinberg, Bill would give schoolsa better scorecard, The Sacramento Bee, July 6, 2011,http://www.sacbee.com/2011/07/06/3749797/bill-would-give-schools-a-better.html

15. Estimates are rom $65 million to $130 million. Craig D.Jerald, On Her Majestys School Inspection Service.

Documents

Education Sector: Some Assembly Required