[House Hearing, 109 Congress]
[From the U.S. Government Publishing Office]



 
                         NO CHILD LEFT BEHIND:
                        CAN GROWTH MODELS ENSURE
                  IMPROVED EDUCATION FOR ALL STUDENTS?

=======================================================================


                                HEARING

                               before the

                         COMMITTEE ON EDUCATION
                           AND THE WORKFORCE
                     U.S. HOUSE OF REPRESENTATIVES

                       ONE HUNDRED NINTH CONGRESS

                             SECOND SESSION

                               __________

                             July 27, 2006

                               __________

                           Serial No. 109-50

                               __________

  Printed for the use of the Committee on Education and the Workforce



 Available via the World Wide Web: http://www.access.gpo.gov/congress/
                                 house
                                   or
            Committee address: http://edworkforce.house.gov



                                 _____

                 U.S. GOVERNMENT PRINTING OFFICE

28-839 PDF              WASHINGTON : 2006
_________________________________________________________________
For sale by the Superintendent of Documents, U.S. Government 
Printing  Office Internet: bookstore.gpo.gov  Phone: toll free 
(866) 512-1800; DC area (202) 512-1800 Fax: (202) 512-2250 Mail:
Stop SSOP, Washington, DC 20402-0001




                COMMITTEE ON EDUCATION AND THE WORKFORCE

            HOWARD P. ``BUCK'' McKEON, California, Chairman

Thomas E. Petri, Wisconsin, Vice     George Miller, California,
    Chairman                           Ranking Minority Member
Michael N. Castle, Delaware          Dale E. Kildee, Michigan
Sam Johnson, Texas                   Major R. Owens, New York
Mark E. Souder, Indiana              Donald M. Payne, New Jersey
Charlie Norwood, Georgia             Robert E. Andrews, New Jersey
Vernon J. Ehlers, Michigan           Robert C. Scott, Virginia
Judy Biggert, Illinois               Lynn C. Woolsey, California
Todd Russell Platts, Pennsylvania    Ruben Hinojosa, Texas
Patrick J. Tiberi, Ohio              Carolyn McCarthy, New York
Ric Keller, Florida                  John F. Tierney, Massachusetts
Tom Osborne, Nebraska                Ron Kind, Wisconsin
Joe Wilson, South Carolina           Dennis J. Kucinich, Ohio
Jon C. Porter, Nevada                David Wu, Oregon
John Kline, Minnesota                Rush D. Holt, New Jersey
Marilyn N. Musgrave, Colorado        Susan A. Davis, California
Bob Inglis, South Carolina           Betty McCollum, Minnesota
Cathy McMorris, Washington           Danny K. Davis, Illinois
Kenny Marchant, Texas                Raul M. Grijalva, Arizona
Tom Price, Georgia                   Chris Van Hollen, Maryland
Luis G. Fortuno, Puerto Rico         Tim Ryan, Ohio
Bobby Jindal, Louisiana              Timothy H. Bishop, New York
Charles W. Boustany, Jr., Louisiana  [Vacancy]
Virginia Foxx, North Carolina
Thelma D. Drake, Virginia
John R. ``Randy'' Kuhl, Jr., New 
    York
[Vacancy]

                       Vic Klatt, Staff Director
        Mark Zuckerman, Minority Staff Director, General Counsel


                            C O N T E N T S

                              ----------                              
                                                                   Page

Hearing held on July 27, 2006....................................     1
Statement of Members:
    McKeon, Hon. Howard P. ``Buck,'' Chairman, Committee on 
      Education and the Workforce................................     1
        Prepared statement of....................................     2
    Miller, Hon. George, Ranking Minority Member, Committee on 
      Education and the Workforce................................     3
        Prepared statement of....................................     5

Statement of Witnesses:
    Haycock, Kati, director, the Education Trust.................    32
        Prepared statement of....................................    34
    Klein, Chancellor Joel, New York City Department of Education    20
        Prepared statement of....................................    22
    Sanders, William L., Ph.D., senior manager, value-added 
      assessment and research, SAS Institute, Inc................    37
        Prepared statement of....................................    39
    Shaul, Marnie S., Director, Education, Workforce, and Income 
      Security Issues, U.S. Government Accountability Office.....     7
        Prepared statement of....................................     9
    Weaver, Reg, president, National Education Association.......    24
        Prepared statement of....................................    27


                         NO CHILD LEFT BEHIND:


                        CAN GROWTH MODELS ENSURE
                  IMPROVED EDUCATION FOR ALL STUDENTS?

                              ----------                              


                        Thursday, July 27, 2006

                     U.S. House of Representatives

                Committee on Education and the Workforce

                             Washington, DC

                              ----------                              

    The committee met, pursuant to call, at 10:03 a.m., in room 
2175, Rayburn House Office Building, Hon. Howard McKeon 
[chairman of the committee] presiding.
    Present: Representatives McKeon, Castle, Biggert, Platts, 
Keller, Wilson, Kline, Inglis, Miller, Kildee, Owens, Scott, 
Woolsey, Hinojosa, Tierney, Kind, Kucinich, Wu, Holt, Davis of 
California, McCollum, Van Hollen, and Bishop.
    Staff present: Amanda Farris, Professional Staff Member; 
Steve Forde, Communications Director; Jessica Gross, Press 
Assistant; Richard Hoar, Professional Staff Member; Lindsey 
Mask, Press Secretary; Chad Miller, Coalitions Director for 
Education Policy; Deborah L. Emerson Samantar, Committee Clerk/
Intern Coordinator; Alice Cain, Legislative Associate/
Education; Denise Forte, Legislative Associate/Education; 
Lauren Gibbs, Legislative Associate/Education; David Hartzler, 
Junior Technology Assistant; Lloyd Horwich, Legislative 
Associate/Education; Thomas Kiley, Communications Director; 
Ricardo Martinez, Legislative Associate/Education; Joe Novotny, 
Legislative Assistant/Education; and Mark Zuckerman, Staff 
Director/General Counsel.
    Chairman McKeon [presiding]. A quorum being present, the 
Committee on Education and the Workforce will come to order.
    We are holding this hearing today to hear testimony on ``No 
Child Left Behind: Can Growth Models Ensure Improved Education 
for All Students?''
    With that, I ask unanimous consent for the hearing record 
to remain open 14 days to allow members' settlements and other 
extraneous material referenced during the hearing to be 
submitted in the official hearing record. Without objection, so 
ordered.
    Good morning. Today marks the latest in our series of 
hearings on the No Child Left Behind Act, as we work to lay the 
foundation for next year's reauthorization of this landmark 
law.
    As always, I would like to take a moment to thank my 
colleagues for taking part in this important hearing. I would 
like to extend a special note of gratitude to our committee's 
senior Democrat, Mr. Miller, and the Education Reform 
Subcommittee's chairman, Mr. Castle, and ranking member, Ms. 
Woolsey, for providing leadership throughout the worthwhile 
series of hearings.
    We have only just begun exploring the various aspects of No 
Child Left Behind, yet I think we can all say that we have 
learned a great deal already. Each hearing has been 
constructive and informative, and I believe they will prove to 
be invaluable as we forge ahead into next year's 
reauthorization.
    Today's hearing will evaluate the implications of using 
growth models to determine if schools are making adequate 
yearly progress under No Child Left Behind.
    Additionally, we will be discussing a new Government 
Accountability Office report on the benefits and challenges of 
using growth models for accountability purposes under NCLB.
    The reliability and utility of growth models is the focus 
of an ongoing debate, and for this committee to gather input on 
the subject is both reasonable and responsible as part of our 
ongoing series. We are not necessarily here to embrace the 
concept nor to refute it. Instead, we are simply here to listen 
and to learn.
    Under current No Child Left Behind guidelines, school 
districts use a status model to compare the performance of 
students in a specific grade against the performance of the 
students of that same grade in the previous year. This is done 
to determine if schools and districts are meeting adequate 
yearly progress.
    Some have raised concerns about the reliability of the 
status model and have suggested that a growth model would be 
more useful. To be clear, growth models differ from status 
models by comparing the achievement of the same students over 
time.
    Today, we will be hearing from witnesses on their views of 
growth models, whether they are effective monitors of school 
performance and progress, and whether or not growth models meet 
or can be tailored to meet the objectives of No Child Left 
Behind.
    We will have an opportunity to hear from expert witnesses 
on whether growth models determine more than just improvement, 
and if they can be tailored to determine if all students are, 
in fact, reaching proficiency.
    Additionally, we will be considering whether or not growth 
models can ensure that achievement gaps between groups of 
students are closing. After all, that is the fundamental 
principle of the No Child Left Behind Act.
    I believe today's hearing will be very insightful and will 
help us better understand the benefits and challenges of using 
growth models for NCLB accountability purposes. I am looking 
forward to this hearing and the additional hearings we will be 
having in this series.
    And I now yield to my friend, Mr. Miller, for his opening 
statement.
    [The prepared statement of Mr. McKeon follows:]

    Prepared Statement of Hon. Howard P. ``Buck'' McKeon, Chairman, 
                Committee on Education and the Workforce

    Good morning. Today marks the latest in our series of hearings on 
the No Child Left Behind Act, as we work to lay the foundation for next 
year's reauthorization of this landmark law. As always, I'd like to 
take a moment to thank my colleagues for taking part in this important 
hearing.
    I'd also like to extend a special note of gratitude to our 
Committee's senior Democrat, Mr. Miller, and the Education Reform 
Subcommittee's Chairman, Mr. Castle, and ranking Member, Ms. Woolsey, 
for providing leadership throughout this worthwhile series of hearings.
    We have only just begun exploring the various aspects of No Child 
Left Behind, yet I think we can all say that we have learned a great 
deal already. Each hearing has been constructive and informative, and I 
believe they will all prove to be invaluable as we forge ahead into 
next year's reauthorization.
    Today's hearing will evaluate the implications of using growth 
models to determine if schools are making adequate yearly progress 
under No Child Left Behind. Additionally, we will be discussing a new 
Government Accountability Office report on the benefits and challenges 
of using growth models for accountability purposes under NCLB.
    The reliability and utility of growth models is the focus of an 
ongoing debate, and for this Committee to gather input on the subject 
is both reasonable and responsible as part of our ongoing series. We're 
not necessarily here to embrace the concept, nor to refute it. Instead, 
we're simply here to listen--and to learn.
    Under current No Child Left Behind guidelines, school districts use 
a ``status model'' to compare the performance of students in a specific 
grade against the performance of the students of that same grade in the 
previous year. This is done to determine if schools and districts are 
meeting adequate yearly progress.
    Some have raised concerns about the reliability of the status model 
and have suggested that a growth model would be more useful. To be 
clear, growth models differ from status models by comparing the 
achievement of the same students over time,
    Today, we'll be hearing from witnesses on their views of growth 
models, whether they are effective monitors of school performance and 
progress, and whether or not growth models meet or can be tailored to 
meet the objectives of No Child Left Behind.
    We'll have an opportunity to hear from expert witnesses on whether 
growth models determine more than just improvement--and if they can be 
tailored to determine if all students are, in fact, reaching 
proficiency. Additionally, we'll be considering whether or not growth 
models can ensure that achievement gaps between groups of students are 
closing. After all, that's the fundamental principle of No Child Left 
Behind.
    I believe today's hearing will be very insightful and will help us 
better understand the benefits and challenges of using growth models 
for NCLB accountability purposes. I am looking forward to this hearing 
and the additional hearings we'll be having in this series. And I would 
now like to yield to my friend, Mr. Miller, for his opening statement.
                                 ______
                                 
    Mr. Miller. Thank you very much, Mr. Chairman, and thank 
you so much for not only scheduling this hearing on, I think, 
one of the most important subjects we will be considering 
during reauthorization, but also for the series of hearings 
that you have scheduled to help the committee better understand 
the challenge we confront with the reauthorization and 
important consideration of various subject matters that we will 
have to take into account.
    I believe that today's hearing focuses on one of the most 
important decisions we will face in reviewing the No Child Left 
Behind law, whether or not to reform the law's current 
accountability system.
    I can think of no question more central to the 
reauthorization goals of the law. As one of the original 
authors of No Child Left Behind, I am often asked how would I 
like to see the law changed.
    The short answer is that I would like to see us be 
responsive to the legitimate concerns while maintaining the 
core values of the law--that is, providing an equal educational 
opportunity for all children and an excellent education to 
every single child.
    There are some well-founded concerns with the current 
accountability system. One widespread concern relates to 
schools that are not making adequate yearly progress under the 
law, even though their students are making impressive academic 
progress.
    For example, take a 5th-grader who reads at the 1st-grade 
level. Their school could make great strides in helping the 
student read and, over the course of a year, improve enough to 
read at 3rd-grade level, but the school would miss making 
adequate yearly progress under the law if the student and 
others were not reading at grade level.
    The second major concern is that the different students are 
measured each year, so the achievement of this year's 5th-
graders is measured against the achievement of last year's 5th-
graders. As a result, a gain or loss of a percentage of 
students who are proficient could be the result of factors 
beside the school.
    We need to carefully weigh and address these and other 
concerns. It is important that any accountability system 
identify schools that need extra help in the most fair and 
accurate way possible, so that they can qualify for the 
additional resources, so that children can qualify for the 
additional academic opportunities such as tutoring or 
transferring to another public school.
    It is absolutely necessary that Congress appropriate the 
funds promised to make these services available to the children 
who need them.
    Here is the question we must start answering today: Are 
growth models a feasible alternative to the current 
accountability system? I have an open mind on growth models. I 
came to this idea reluctantly, but I think we have three basic 
questions that we have to deal with.
    First, do states have the data capacity and the expertise 
they need to ensure the information gathered to determine 
whether a school has made adequate progress is both valid and 
reliable?
    Second, do growth models appropriately credit improving 
schools, or do they overstate academic progress? In other 
words, are they a step forward in offering a fairer, more 
reliable means of accountability, or are they a step backwards, 
simply another loophole that hinders accountability?
    Third, and most importantly from my point of view, is are 
growth models consistent with No Child's Left Behind ultimate 
goal of assuring that all children can read and do math and 
science at grade level by 2014? It is imperative that the 
growth be pegged to proficiency.
    No Child Left Behind's goal of an excellent education for 
all children and equality of educational opportunity for all 
children are goals that our nation has been pursuing for at 
least 40 years, ever since the signing of the Elementary and 
Secondary Education Act of 1966.
    We have yet to achieve them. Poor and minority children are 
still often assigned to less challenging classes with less 
qualified teachers. As a result, fewer than half of the 
minority children can read at grade level according to the 
National Assessment of Educational Progress.
    And black and Latino 17-year-old students, on the average, 
taught math at the same level as white 13-year-old students. 
This is unacceptable.
    For these children, a good education is often their best 
and only hope for a prosperous future. That is why we must stay 
true to No Child Left Behind's promise to provide opportunity 
and an excellent education to every child, even as we make the 
necessary adjustments to the law.
    I want to thank our panelists for being here today.
    I think you have assembled a wonderful panel, Mr. Chairman. 
And I would like to extend a special thanks to Marlene Shaul 
from the GAO, who I understand is now retiring.
    One lousy report on NCLB and you are retiring, right? My 
gosh, it is that difficult? Maybe we will all think about 
joining you. But thank you so much for all your work in the GAO 
office.
    And with that, I look forward to the hearing.
    And again, thank you, Mr. Chairman, for starting this 
process this year before we get into reauthorization next year.

  Prepared Statement of Hon. George Miller, Ranking Minority Member, 
                Committee on Education and the Workforce

    Good morning. I want to thank Chairman McKeon for scheduling 
today's hearing on one of the most important decisions we face in 
reviewing the No Child Left Behind law: whether or not to reform the 
law's current accountability system. I can think of no question more 
central to the reauthorization and goals of the law.
    As one of the original authors of No Child Left Behind, I am often 
asked how I would like to see the law changed. The short answer is that 
I would like to see us be responsive to legitimate concerns while 
maintaining the core values of the law--providing equal educational 
opportunities for all children and an excellent education to every 
single child.
    There are some well-founded concerns with the current 
accountability system.
    One widespread concern relates to schools that are not making 
adequate yearly progress under the law even though their students are 
making impressive academic progress.
    For example, take a fifth grader who reads at the first grade 
level. Their school could make great strides in helping the student 
read and, over the course of a year, improve enough to read at the 
third grade level. But the school could miss making adequate yearly 
progress under the law if that student, and others, are still not 
reading at grade level.
    A second major concern is that different students are measured each 
year, so the achievement of this year's fifth graders is measured 
against the achievement of last year's fifth graders. As a result, a 
gain or loss in the percentage of students who are proficient could be 
a result of factors besides the school.
    We need to carefully weigh and address these and other concerns.
    It is important that any accountability system identify schools 
that need extra help in the most fair and accurate way possible, so 
they can qualify for additional resources, and so their children can 
qualify for extra academic opportunities, such as tutoring or the 
transferring to another public school.
    It is absolutely necessary that Congress appropriate the funds 
promised to make these services available to the children who need 
them.
    Here's the question we need to start answering today: are growth 
models a feasible alternative to the current accountability system?
    I have an open mind about growth models, and have three basic 
questions.
     First, do states have the data capacity and expertise they 
need to ensure that information gathered to determine whether a school 
has made adequate progress is both valid and reliable?
     Second, do growth models appropriately credit improving 
schools, or do they overstate academic progress? In other words, are 
they a step forward in offering a fairer, more reliable means of 
accountability? Or are they a step backward--simply another loophole 
that hinders accountability?
     Third, and most importantly, are growth models consistent 
with No Child Left Behind's ultimate goal of ensuring that all children 
can read and do math and science at grade level by 2014? It is 
imperative that growth be pegged to proficiency.
    No Child Left Behind's goals--of an excellent education for all 
children and equality of educational opportunity for all children--are 
goals that our nation has been pursuing for at least 40 years, ever 
since the signing of the Elementary and Secondary Education Act in 
1966. But we have yet to achieve them.
    Poor and minority children are still often assigned to less-
challenging classes and less qualified teachers. As a result, fewer 
than half of minority children can read at grade level, according to 
the National Assessment of Educational Progress.
    And black and Latino 17-year-old students are, on average, taught 
math at the same level as white 13-year-old students.
    This is unacceptable.
    For these children, a good education is often their best and only 
hope for a prosperous future.
    That's why we must stay true to No Child Left Behind's promise to 
provide opportunity and an excellent education to every child, even as 
we make the necessary adjustments to the law.
    I thank our panelists for being with us today and for the light you 
will shed on these important questions. I would like to extend a 
special thank you to Marnie Shaul--who I understand is retiring--for 
her outstanding work at the Government Accountability Office, including 
all of her work on the very useful report that we will hear about 
today.
    I also thank Chairman McKeon and his staff for the bipartisan 
process that led to this hearing. I hope that today's hearing will help 
us make great progress towards our reauthorization of No Child Left 
Behind.
                                 ______
                                 
    Chairman McKeon. Thank you, Mr. Miller. I hope that wasn't 
your retirement you were announcing.
    Mr. Miller. Not me, Mr. Chairman. Not as long as you are 
here.
    [Laughter.]
    Chairman McKeon. We work well together. I would hate to see 
him leave, most of the time.
    We have a distinguished panel here today, and I would like 
to introduce them now at this time.
    First we will hear from Dr. Marlene Shaul, the director of 
education, workforce and income security for the Government 
Accountability Office. Dr. Shaul is responsible for the studies 
GAO undertakes on early childhood programs and elementary and 
secondary education programs. Before her work in the Federal 
Government, she worked for the state of Ohio on community and 
business development issues and at the Kettering Foundation. 
Dr. Shaul holds a Ph.D. in economics from the Ohio State 
University.
    I would also like to take a moment to recognize Dr. Shaul 
for all of her work on behalf of GAO. And as Mr. Miller has 
pointed out, she will be retiring next month, and I want to 
thank her for all her efforts on what you have done.
    Then we will hear from Mr. Joel Klein, the chancellor of 
the New York City Department of Education. Chancellor Klein 
oversees more than 1,450 schools with over 1.1 million 
students. When I served on the school board in California, our 
district had 10,000 students, and the way they divided this up 
at the state level was small districts, large districts, and 
Los Angeles. And Los Angeles is much smaller than New York when 
it comes to the number of children that you have.
    Since becoming chancellor, Mr. Klein has enacted his reform 
program, Children First, which provides academic support for 
students who are struggling as well as new supports for 
parents, in addition to improving school safety. A native of 
New York, he is active in community service work and has a 
strong interest in educational issues. He studied at New York 
University School of Education and later taught math to 6th-
graders at a public school in Queens.
    Then we will hear from Dr. Reg Weaver, the president of the 
National Education Association. Dr. Weaver is currently serving 
his second term as president of the NEA. Prior to his position 
at the NEA here in Washington, Dr. Weaver served as president 
of his local NEA in Harvey, Illinois and as president of the 
NEA affiliate in Illinois.
    A 30-year classroom veteran, Dr. Weaver also serves on a 
number of boards, including the executive board of the National 
Council for the Accreditation of Teacher Education and the 
board of directors of the National Board of Professional 
Teaching Standards.
    Then we will hear from Ms. Kati Haycock, who is the 
director of The Education Trust. Established in 1992, Ed Trust 
provides a voice for young people, especially those who are 
poor or members of minority groups. Ed Trust also provides 
hands-on assistance to educators who want to work together to 
improve student achievement in pre-Kindergarten through 
college. Prior to coming to The Education Trust, Ms. Haycock 
served as executive vice president of the Children's Defense 
Fund, the nation's largest child advocacy organization.
    And finally, we will hear from Dr. William Sanders, 
research fellow with the University of North Carolina system 
and a manager of value-added assessment and research for the 
SAS Institute Inc in Cary, North Carolina. Over the last 20 
years, Dr. Sanders and his colleagues have developed and 
refined a methodology to measure the influence that school 
systems, schools and teachers have on the academic progress of 
students.
    He has served as an adviser to policymakers at the Federal 
level and has worked with many states and school districts 
interested in developing a value-added component to leverage 
their testing data into more precise and reliable information. 
Many of his suggestions concerning measurement of student 
outcomes were incorporated into Tennessee's Educational 
Improvement Act of 1992.
    Thank you all for being here.
    I would like to remind you, as you begin your testimony, 
there is a little light that will come on in front of you. 
Green means go. Yellow means you have a minute left. And red 
means wrap it up if you haven't already done so.
    And your full written testimony will be included in the 
record.
    Dr. Shaul?

 STATEMENT OF MARLENE S. SHAUL, DIRECTOR, EDUCATION, WORKFORCE 
  AND INCOME SECURITY ISSUES, U.S. GOVERNMENT ACCOUNTABILITY 
                             OFFICE

    Ms. Shaul. Mr. Chairman and members of the committee, thank 
you for the opportunity to present the findings of the report 
we did for this committee on growth models and how they might 
be used to meet the goals of the No Child Left Behind Act, such 
as having all students proficient by 2014.
    Our findings are based on GAO's March 2006 survey of the 50 
states and the District of Columbia about their use of growth 
models. We also analyzed academic performance data in several 
states.
    As the committee requested, we have used a broad definition 
of growth models that includes year-to-year comparisons of 
schools, cohorts of students and individual students.
    My remarks today focus on three topics: How many states are 
currently using some form of growth models and why; how growth 
models can measure progress toward achieving key goals of the 
law; and the challenges states face in using growth models to 
measure schools' progress, especially to meet the law's 
requirements.
    Before I go over our findings on each of these, let me give 
you our bottom line. It is possible for states to use growth 
models to meet the goals of the law, but they face considerable 
challenges in doing so.
    Now I would like to turn to each of the topics. As you 
know, states use growth status models to measure AYP and many 
of them use growth models for their own purposes such as 
grading schools and rewarding teachers.
    As the map shows on the screen, in March of 2006 26 states 
were using growth models. Those are shown in the dark green. 
Another 22 states were either considering using growth models 
or in the process of implementing them. And those are shown in 
the medium green.
    Most states are currently using school level growth models. 
They compare changes in proficiency levels over time at a 
particular school. Using a hypothetical example, let's see how 
this works. It is a three-step process.
    In year one, 50 percent of the students assessed were 
proficient or above. In year two, 60 percent of the students 
assessed were proficient or above. Growth is measured as that 
10 percent difference, and states could use this measure of 
growth, for example, to compare schools that are similarly 
situated.
    Seven states currently track individual students and can 
compare growth for a cohort of students--that is the same group 
of students over time--or for individual students. These models 
require substantially more data and are more involved than 
models that measure results at the school level.
    In these models, the state may compare actual student 
achievement to prior student achievement to determine if 
progress is sufficient.
    We examined two state models that measure growth over time 
and measure progress toward the law's goals. Massachusetts' 
model sets school targets that are achieved when the number of 
students at various levels of achievement increases.
    The model sets different targets for the school and for 
each subgroup in the school, and this model has been approved 
by the Department of Education for use in meeting AYP.
    Tennessee's model uses individual student data and projects 
future academic achievement using information from past 
performance. It was recently approved by the department for its 
new growth model pilot program.
    Let's take a look at a very simplified example. My 
apologies to Dr. Sanders here for a very simplified example. In 
4th grade, we just have three students on the screen, Students 
A, B and C, and none of them are shown as meeting the 4th-grade 
proficiency target.
    Next, a projection is made based on past performance, and 
you can see the projection to 7th grade shows that two of these 
three students are projected to make the state's 7th-grade AYP 
target.
    So two of these three students can be counted as 
contributing to meeting the 4th-grade AYP target. Now, you may 
wonder why two students with similar scores would have 
different projections. That is because the projection is based 
on prior assessments, and those could be quite different.
    As you might gather, using growth models can be 
complicated. To undertake growth models, states must have 
assessments that are comparable across grades, and at least 2 
years of assessment data and the capacity to analyze this 
greater amount of data.
    States need personnel who have technical skills and can 
explain the results of growth models to teachers and to 
parents.
    Using growth models for AYP purposes may pose a risk for 
some lower performing schools. Lower performing schools that 
initially make good growth will need to sustain high levels of 
growth rates to meet future targets.
    If Title I schools do not sustain high growth, they may be 
disadvantaged by not receiving federally required assistance 
during the years that the growth target was met.
    On the other hand, if fewer schools need federally required 
assistance, resources could be concentrated in the lowest 
performing schools.
    Mr. Chairman, this concludes my remarks. Thank you.
    [The prepared statement of Ms. Shaul follows:]

Prepared Statement of Marlene S. Shaul, Director, Education, Workforce, 
   and Income Security Issues, U.S. Government Accountability Office

    Mr. Chairman and Members of the Committee: I am pleased to be here 
today to discuss our report, which describes how states use growth 
models to measure academic performance and how these models can measure 
progress toward achieving key goals of the
    No Child Left Behind Act of 2001 (NCLBA). With annual expenditures 
approaching $13 billion dollars for Title I alone, NCLBA represents the 
federal government's single largest investment in the education of the 
48 million students who attend public schools. The NCLBA-the most 
recent reauthorization of the Elementary and Secondary Education Act of 
1965-requires states to improve academic performance so that all 
students are proficient by 2014 and achievement gaps among groups such 
as economically disadvantaged students close. The upcoming 
reauthorization of the law presents an opportunity to discuss some key 
issues associated with the act.
    To measure whether schools are making adequate yearly progress 
(AYP) toward having all students proficient by 2014, states set annual 
proficiency targets using an approach known as a status model, which 
calculates test scores 1 year at a time. With status models, states or 
districts determine whether schools make AYP based on performance for 
the year while generally not taking into account how much better or 
worse the school did than during the previous year. Thus, a school that 
is showing significant improvement in student achievement but has too 
few students at the proficient level would not likely make AYP.
    In addition to determining whether schools meet proficiency 
targets, some states have interest in also recognizing schools that 
make progress toward NCLBA goals. Growth models can measure progress in 
achievement or proficiency over time and vary in complexity, such as 
calculating annual progress in a school's average test scores from year 
to year; estimating test score progress while taking into account how 
factors such as student background may affect such progress; or 
projecting future scores based on current and prior years' results. 
While growth models are sometimes defined as tracking the same students 
over time, because of the committee's interest in the range of models 
states are using to measure academic improvement, we define a growth 
model as a model that measures changes in proficiency levels or test 
scores of a student, group, grade, school, or district for 2 or more 
years. We included models that track schools and student groups in 
order to provide a broad assessment of options that may be available to 
states.
    My testimony today will focus on how growth models may provide 
useful information on academic performance. Specifically, I will 
discuss (1) how many states are using growth models and for what 
purposes, (2) how growth models can measure progress toward achieving 
key NCLBA goals, and (3) what challenges states face in using growth 
models especially to meet the law's key goals.
    My written statement is drawn from our recent report on growth 
models, which we completed for the committee.\1\ For this report, we 
conducted a survey of all states to determine whether they were using 
growth models. We conducted telephone interviews with state and local 
education agency officials in eight states that collectively use a 
variety of growth models, and conducted site visits to California, 
Massachusetts, North Carolina and Tennessee. For Massachusetts and 
Tennessee we analyzed student-level data from selected schools to 
illustrate how their models measure progress toward key NCLBA goals. We 
conducted this work in accordance with generally accepted government 
auditing standards.
Summary
    In summary, nearly all states were using or considering growth 
models for a variety of purposes in addition to their status models as 
of March 2006. Twenty-six states were using growth models, and another 
22 were either considering or in the process of implementing them. Most 
states that used growth models did so for schools as a whole and for 
particular groups of students, and 7 also measured growth for 
individual students. For example, Massachusetts measured growth for 
schools and groups of students but does not track individual students' 
scores, while Tennessee set different expectations for growth for each 
student based on the student's previous test scores. Seventeen of the 
states that used growth models had been doing so prior to passage of 
the NCLBA, while 9 began after the law's passage. States used their 
growth models for a variety of purposes, such as targeting resources 
for students that need extra help or awarding teachers bonus money 
based on their school's relative performance.
    Certain growth models are capable of tracking progress toward the 
goals of universal proficiency by 2014 and closing achievement gaps. 
For example, Massachusetts uses its model to set targets based on the 
growth that it expects from schools and their student groups. Schools 
can make AYP if they reach these targets, even if they fall short of 
reaching the statewide proficiency targets set with the state's status 
model. Tennessee designed a model, different from the one used for 
state purposes described above, that projects students' test scores and 
whether they will be proficient in the future. In this model, if 79 
percent of a school's students are predicted to be proficient in 3 
years, the school would reach the state's 79 percent proficiency target 
for the current school year (2005-2006).
    States face challenges in developing and implementing growth models 
that would allow them to meet NCLBA goals. Technical challenges include 
creating data and assessment systems to meet the substantial data 
requirements of growth models and having personnel that can analyze and 
communicate growth model results. For example, states need to have 
tests that are comparable from one year to the next to accurately 
measure progress. Further, some models require sophisticated data 
systems that have the capacity to track individual student performance 
across grades and schools. Using growth models can present risks for 
states if schools are designated as making AYP while still needing 
assistance to progress. For example, one school in Tennessee that did 
not make AYP under the status model would make AYP under the state's 
proposed growth model. This school is located in a high-poverty, inner-
city neighborhood and has been receiving federal assistance targeted to 
improving student performance. If the school continues to make AYP 
under the growth model, its students would no longer receive federally 
required services, such as tutoring or the option of transferring to a 
higher performing school. On the other hand, the school's progress may 
result in its making AYP in the future under the state's status model. 
U.S. Department of Education (Education) initiatives may help states 
address these challenges. For example, Education started a pilot 
project for states to use growth models that meet the department's 
specific criteria to determine AYP. Education also provided grants to 
states to support their efforts to track individual test scores over 
time.
    By proceeding with a pilot project with clear goals and criteria 
and by requiring states to compare results from their growth model with 
status model results, Education is poised to gain valuable information 
on whether or not growth models are overstating progress or whether 
they appropriately give credit to fast-improving schools. In comments 
on a draft of our recent report, Education expressed concern that the 
use of a broader definition of growth models would be confusing. GAO 
used this definition in order to reflect the variety of approaches 
states have been taking to measure growth in academic performance.
Background
    The NCLBA \2\ requires states to set challenging academic content 
and achievement standards in reading or language arts and 
mathematics\3\ and determine whether school districts and schools make 
AYP toward meeting these standards.\4\ To make AYP, schools generally 
must:
     show that the percentage of students scoring at the 
proficient level or higher meets the state proficiency target for the 
school as a whole and for designated student groups,
     test 95 percent of all students and those in designated 
groups, and
     meet goals for an additional academic indicator, such as 
the state's graduation rate.
    The purpose of Title I Part A is to improve academic achievement 
for disadvantaged students. Schools receiving Title I federal funds 
that do not make AYP for 2 or more years in a row must take action to 
assist students, such as offering students the opportunity to transfer 
to other schools or providing additional educational services like 
tutoring.
    States measure AYP using a status model that determines whether or 
not schools and students in designated groups meet proficiency targets 
on state tests 1 year at a time. States generally used data from the 
2001-2002 school year to set the initial percentage of students that 
needed to be proficient for a school to make AYP, known as a starting 
point. From this point, they set annual proficiency targets that 
increase up to 100 percent by 2014. For example, for schools in a state 
with a starting point of 28 percent to achieve 100 percent by 2014, the 
percentage of students who scored at or above proficient on the state 
test would have to increase by 6 percentage points each year, as shown 
in figure 1.\5\ Schools that do not reach the state target will 
generally not make AYP.
figure 1: hypothetical example of annual proficiency targets set under 
                             a status model


    The law indicates that states are expected to close achievement 
gaps, but does not specify annual targets to measure progress toward 
doing so. States, thus, have flexibility in the rate at which they 
close these gaps. To determine the extent that achievement gaps are 
closing, states measure the difference in the percentage of students in 
designated student groups and their peers that reach proficiency. For 
example, an achievement gap exists if 40 percent of a school's non-
economically disadvantaged students were proficient compared with only 
16 percent of economically disadvantaged students, a gap of 24 
percentage points. To close the gap, the percentage of students in the 
economically disadvantaged group that reaches proficiency would have to 
increase at a faster rate than that of their peers.
    If a school misses its status model target in a single year, the 
law includes a ``safe harbor'' provision that provides a way for 
schools that are showing significant increases in proficiency rates of 
student groups to make AYP. Safe harbor measures academic performance 
in a way that is similar to certain growth models do and allows a 
school to make AYP by reducing the percentage of students in designated 
student groups that were not proficient by 10 percent, so long as the 
groups also show progress on another academic indicator. For example, 
in a state with a status model target of 40 percent proficient, a 
school could make AYP under safe harbor if 63 percent of a student 
group was not proficient compared to 70 percent in the previous year.
Nearly All States Reported Using or Considering Growth Models to 
        Measure Academic Performance
    Twenty-six states reported using growth models in addition to using 
their status models to track the performance of schools, designated 
student groups, or individual students, as reported in our March 2006 
survey. Additionally, nearly all states are considering the use of 
growth models (see fig. 2).
 figure 2: states that reported using or considering growth models, as 
                             of march 2006


    Of the 26 states using growth models, 19 states reported measuring 
changes for schools and student groups, while 7 states reported 
measuring changes for schools, student groups, and individuals, as 
shown in table 1.
table 1: types of growth models and states using them, as of march 2006


    For example, Massachusetts used a model that measures growth for 
the school as a whole and for designated student groups. The state 
awards points to schools in 25-point increments for each student,\6\ 
depending on how students scored on the state test. Schools earn 100 
points for each student who reaches proficiency, but fewer points for 
students below proficiency. The state averages the points to award a 
final score to schools. Growth in Massachusetts is calculated by taking 
the difference in the annual scores that a school earns between 2 
years. Figure 3 illustrates the growth a school can make from one year 
to the next as measured by Massachusetts model.
             figure 3: illustration of school-level growth


    Tennessee reported using a growth model that sets different goals 
for each individual student based on the students' previous test 
scores. The goal is the score that a student would be expected to 
receive, and any difference between a student's expected and actual 
score is considered that student's amount of yearly growth,\7\ as shown 
in figure 4.
  figure 4: example of higher-than-expected growth for a fourth-grade 
                    student under tennessee's model


    In addition, Tennessee's model, known as the Tennessee Value-Added 
Assessment System, estimates the unique contribution-the value added-
that the teacher and school make to each individual student's growth in 
test scores over time.\8\ The state then uses that amount of growth, 
the unique contribution of the school, and other information to 
determine whether schools are below, at, or above their level of 
expected performance. The model also grades schools with an A, B, C, D, 
or F, which is considered a reflection of the extent to which the 
school is meeting its requirements for student learning.
    Seventeen of the 26 states using growth models reported that their 
models were in place before the passage of the NCLBA during the 2001-
2002 school year, and the remaining 9 states implemented them after the 
law was passed. States used them for purposes such as rewarding 
effective teachers and designing intervention plans for struggling 
schools. For example, North Carolina used its model as a basis to 
decide whether teachers receive bonus money. Tennessee used its value-
added model to provide information about which teachers are most 
effective with which student groups. In addition to predicting 
students' expected scores on state tests, Tennessee's model was used to 
predict scores on college admissions tests, which is helpful for 
students who want to pursue higher education. In addition, California 
used its model to identify schools eligible for a voluntary improvement 
program.
Certain Growth Models Can Measure Progress toward Key NCLBA Goals
    Certain growth models can measure progress in achieving key NCLBA 
goals of reaching universal proficiency by 2014 and closing achievement 
gaps. While states developed growth models for purposes other than 
NCLBA, states such as Massachusetts and Tennessee have adjusted their 
state models to use them to meet NCLBA goals. The Massachusetts model 
has been used to make AYP determinations as part of the state's 
accountability plan in place since 2003. Tennessee submitted a new 
model to Education for the growth models pilot that differs from the 
value-added model described earlier. This new model gives schools 
credit for students projected to reach proficiency within 3 years in 
order to meet key NCLBA goals. Our analysis of how models in 
Massachusetts and Tennessee can measure progress toward the law's two 
key goals is shown in table 2.
table 2: how a status model and certain growth models measure progress 
                     in achieving key nclba goals*


    *Note: Additional requirements for schools to make AYP are 
described in the background section of our report. Massachusetts refers 
to proficiency targets as performance targets and refers to growth 
targets as improvement targets.
    \a\ The information presented in this table reflects the model 
Tennessee proposed to use as part of Education's growth model pilot 
project, as opposed to the value-added model it uses for state 
purposes. The information is based on the March 2006 revision of the 
proposal the state initially made in February 2006.
    Massachusetts designed a model that can measure progress toward the 
key goals of NCLBA by setting targets for the improvement of schools 
and their student groups that increase over time until all students are 
proficient in 2014. Schools can get credit for improving student 
proficiency even if, in the short term, the requisite number of 
students has yet to reach the state's status model proficiency targets. 
For example, figure 5 illustrates a school that is on track to make AYP 
annually through 2014 by reaching its growth targets. While these 
growth targets increase at a faster pace than the state's annual 
proficiency target until 2014, they do provide the school with an 
additional measure by which it can make AYP.
 figure 5: targets for a selected school in massachusetts compared to 
                       state status model targets


    The model also measures whether achievement gaps are closing by 
setting targets for designated student groups, similar to how it sets 
targets for schools as a whole. Schools that increase proficiency too 
slowly-that is, do not meet status or growth targets-will not make AYP. 
For example, one selected school in Massachusetts showed significant 
gains for several designated student groups that were measured against 
their own targets. However, the school did not make AYP because gains 
for one student group were not sufficient. This group-students with 
disabilities-fell short of its growth target, as shown in figure 6.
figure 6: results for a selected school in massachusetts in mathematics


    Tennessee developed a different model that can also measure 
progress toward the NCLBA goals of universal proficiency and closing 
achievement gaps. Tennessee created a new version of the model it had 
been using for state purposes to better align with NCLBA.\9\ Referred 
to as a projection model, this approach projects individual student's 
test scores into the future to determine when they may reach the 
state's status model proficiency targets.
    In order to make AYP under this proposal, a school could reach the 
state's status model targets by counting as proficient in the current 
year those students who are predicted to be proficient in the future. 
The state projects scores for elementary and middle school students 3 
years into the future to determine if they are on track to reach 
proficiency, as follows:
     fourth-grade students projected to reach proficiency by 
seventh grade,
     fifth-grade students projected to reach proficiency by 
eighth grade, and
     sixth-, seventh-, and eighth-grade students projected to 
reach proficiency on the state's high school proficiency test.
    These projections are based on prior test data and assume that the 
student will attend middle or high schools with average performance (an 
assumption known as average schooling experience).\10\ At our request, 
Tennessee provided analyses for students in several schools that would 
make AYP under the proposed model. To demonstrate how the model works, 
we selected students from a school and compared their actual results in 
fourth grade (panel A) with their projected results for seventh grade 
(panel B) (see fig. 7).
figure 7: results for selected students in mathematics from a school in 
                               tennessee*


    *Note: The same students are presented in both panels (for example, 
student A in panel A is the same student as student A in panel B). 
While these data reflect the scores of individual students, Tennessee 
provided data to GAO in such a way that student privacy and 
confidentiality were ensured. Data are illustrative and are not meant 
to be a statistical representation of the distribution of students in 
this school.
    Tennessee's proposed model can also measure achievement gaps. Under 
NCLBA, a school makes AYP if all student groups meet the state 
proficiency target. In Tennessee's model, whether the achievement gap 
is potentially closed would be determined through projections of 
students' performance in meeting the state proficiency target.
States Face Challenges in Implementing Growth Models
    States generally face challenges in collecting and analyzing the 
data required to implement growth models including models that would 
meet the law's goals. In addition, using growth models can present 
risks for states if schools are designated as making AYP while still 
needing assistance to progress. Education has initiatives that may help 
states address these challenges.
    States must have certain additional data system requirements to 
implement growth models, including models that would meet NCLBA 
requirements.
    First, a state's ability to collect comparable data over at least 2 
years is a minimum requirement for any growth model. States must ensure 
that test results are comparable from one year to the next and possibly 
from one grade to the next, both of which are especially challenging 
when test questions and formats change. Second, the capacity to collect 
data across time and schools is also required to implement growth 
models that use student-level data. This capacity often requires a 
statewide system to assign unique numbers to identify individual 
students. Developing and implementing these systems is a complicated 
process that includes assigning numbers, setting up the system in all 
schools and districts, and correctly matching individual student data 
over time, among other steps. Third, states need to ensure that data 
are free from errors in their calculations of performance. While 
ensuring data accuracy is important for status models, doing so is 
particularly important for growth models, because errors in multiple 
years can accumulate, leading to unreliable results.
    States also need greater research and analysis expertise to use 
growth models as well as support for people who need to manage and 
communicate the model's results. For example, Tennessee officials told 
us that they have contracted with a software company for several years 
because of the complexity of the model and its underlying data system. 
Florida has a contract with a local university to assist it with 
assessing data accuracy, including unique student identifiers required 
for its model. In addition, states will incur training costs as they 
inform teachers, administrators, media, legislators, and the general 
public about the additional complexities that occur when using growth 
models. For example, administrators in one district in North Carolina 
told us that their district lacks enough specialists who can explain 
the state's growth model to all principals and teachers in need of 
guidance and additional training.
    Using growth models can present risks for states if schools are 
designated as making AYP while still needing assistance to progress. On 
the basis of growth model results, some schools would make AYP even 
though these schools may have relatively low-achieving students. As a 
result, some students in Title I schools may be disadvantaged by not 
receiving federally-required services.
    In two Massachusetts districts that we analyzed, 23 of the 59 
schools that made AYP did so based on the state's growth model, even 
though they did not reach the state's status model proficiency rate 
targets in 2003-2004.\11\ Consequently, these schools may not be 
eligible to receive services required under NCLBA for schools in need 
of improvement, such as tutoring and school choice. Because these 
schools would need to sustain high growth rates in order to achieve 
universal proficiency by 2014, it is likely that their students would 
benefit from additional support.
    In Tennessee, 47 of the 353 schools that had not made AYP in the 
2004-2005 school year would do so under the state's proposed projection 
model. One school that would be allowed to make AYP under the proposed 
model was located in a high-poverty, inner-city neighborhood. That 
school receives Title I funding, as two-thirds of its students are 
classified as economically disadvantaged. The school was already 
receiving services required under NCLBA to help its students. If the 
school continues to make AYP under the growth model, these services may 
no longer be provided.
    Education's initiatives, such as the growth model pilot project, 
may facilitate growth model implementation. In November 2005, Education 
announced a pilot project for states to submit proposals for using a 
growth model-one that meets criteria established by the department-
along with their status model, to determine AYP. While NCLBA does not 
specify the use of growth models for making AYP determinations, the 
department started the pilot to evaluate how growth models might help 
schools meet NCLBA proficiency goals and close achievement gaps.
    For the growth model pilot project, each state had to demonstrate 
how its growth model proposal met Education's criteria, many of which 
are consistent with the legal requirements of status models. In 
addition to those requirements, Education included criteria that the 
proposed models track student progress over time and have an assessment 
system with tests that are comparable over time. Of the 20 proposals, 
Education approved 2 states-North Carolina and Tennessee-to use growth 
models to make AYP determinations in the 2005-2006 school year. States 
may submit proposals for the pilot again this fall.
    In addition to meeting all of the criteria, Education and peer 
reviewers noted that Tennessee and North Carolina had many years of 
experience with data systems that support growth models. These states 
must report to Education the number of schools that made AYP on the 
basis of their status and growth models. Education expects to share the 
results with other states, Congress, and the public after it assesses 
the effects of the pilot.
    In addition to the growth model pilot project, Education awarded 
nearly $53 million in grants to 14 states for the design and 
implementation of statewide longitudinal data systems-systems that are 
essential for the development of student-level growth models. While 
independent of the pilot project, states with a longitudinal data 
system-one that gathers data such as test scores on the same student 
from year to year-will be better positioned to implement a growth model 
than they would have been without it. Education intended the grants to 
help states generate and use accurate and timely data to meet reporting 
requirements, support decision making, and aid education research, 
among other purposes. Education plans to disseminate lessons learned 
and solutions developed by states that received grants.
Conclusion
    While status models provide a snapshot of academic performance, 
growth models can provide states with more detailed information on how 
schools' and students' performance has changed from year to year. 
Growth models can recognize schools whose students are making 
significant gains on state tests but are still not proficient. 
Educators can use information about the academic growth of individual 
students to tailor interventions to the needs of particular students or 
groups. In this respect, models that measure individual students' 
growth provide the most in-depth and useful information, yet the 
majority of the models currently in use are not designed to do this.
    Through its approval of Massachusetts' model and the growth model 
pilot program, Education is proceeding prudently in its effort to allow 
states to use growth models to meet NCLBA requirements. Education is 
allowing only states with the most advanced models that can measure 
progress toward NCLBA goals to use the models to determine AYP. Under 
the pilot project, which has clear goals and criteria that requires 
states to compare results from their growth model with status model 
results, Education is poised to gain valuable information on whether or 
not growth models are overstating progress or whether they 
appropriately give credit to fast-improving schools.
    While growth models may be defined as tracking the same students 
over time, GAO used a definition that also includes tracking the 
performance of schools and groups of students. In comments on our 
report, Education expressed concern that this definition may confuse 
readers because it is very broad and includes models that compare 
changes in scores or proficiency levels of schools or groups of 
students. GAO used this definition of growth to reflect the variety of 
approaches states are taking to measure academic progress.
    Mr. Chairman, this concludes my prepared statement. I would be 
happy to respond to any questions that you or members of the committee 
may have.
                                endnotes
---------------------------------------------------------------------------
    \1\ GAO, No Child Left Behind Act: States Face Challenges Measuring 
Academic Growth That Education's Initiatives May Help Address, GAO-06-
661 (Washington, D.C.: July 17, 2006).
    \2\ Pub. L. No. 107-110 (Jan. 8, 2002).
    \3\ The law also requires content standards to be developed for 
science beginning in the 2005-2006 school year and science tests to be 
implemented in the 2007-2008 school year.
    \4\ States determine whether schools and school districts make AYP 
or not. For this report, we will discuss AYP determinations in the 
context of schools.
    \5\ States were able to map out different paths to universal 
proficiency subject to certain limitations. For example, states must 
increase the targets at least once every 3 years and those increases 
must lead to 100 percent proficiency by 2014. See GAO, No Child Left 
Behind Act: Improvements Needed in Education's Process for Tracking 
States' Implementation of Key Provisions, GAO-04-734, (Washington, 
D.C.: Sept. 30, 2004).
    \6\ Students with disabilities are generally included in these 
calculations. The state is allowed to give different tests to students 
with significant cognitive impairments and to count them differently 
for calculating points awarded to schools.
    \7\ Tennessee's growth model mentioned here is not used to make AYP 
determinations under NCLBA. However, Tennessee developed a different 
growth model to determine AYP for Education's growth model pilot 
project. That model is discussed later in this testimony.
    \8\ The state calculates the unique contribution of schools and 
teachers by using a multivariate, longitudinal statistical method where 
results are estimated using data specific for students within each 
classroom or school.
    \9\ Tennessee continues to use its original model to rate schools 
based in part on the unique contributions-or the value added-of school 
to student achievement.
    \10\ While Tennessee's model estimates future performance, other 
models are able to measure growth without these projections. For 
example, Florida uses a model that calculates results for individual 
students by comparing performance in the current year with performance 
in prior years.
    \11\ Another 11 schools also met the growth target, but these 11 
schools made AYP under NCLBA's safe harbor provision.
---------------------------------------------------------------------------
                                 ______
                                 
    Chairman McKeon. Thank you very much.
    Chancellor Klein?

     STATEMENT OF JOEL I. KLEIN, CHANCELLOR, NEW YORK CITY 
                    DEPARTMENT OF EDUCATION

    Mr. Klein. Thank you, Mr. Chairman, Mr. Miller, members of 
the committee. I appreciate the opportunity to be before you 
today.
    I want to acknowledge my friend, Major Owens, who has 
served this country with distinction for the past 24 years, who 
will be leaving this Congress at the end of the term. And it is 
a loss to our city, and you will be missed, sir. I just want 
you to know that.
    You have my testimony, so let me highlight a few of the 
points that I think are critical, because I agree with the 
committee. I think this is a key issue on No Child Left Behind.
    Mr. Miller, I think you are right. I think we have many 
problems in K-12 education, none greater than the achievement 
gap between black and Latino kids on the one hand and white and 
Asian kids on the other. It is the shame of this nation. I used 
to think it was merely a moral crisis. Now it is an economic 
crisis as well.
    And I think NCLB, which is an act that needs to be 
improved, nevertheless put in place an accountability system 
that forced our nation to come to grips, after decades of 
reform in public education that had not moved the needle--to 
come to grips with this fundamental issue of a massive 
achievement gap.
    And to me, whatever else happens, this Congress and this 
nation should not throw out the baby with the bath water when 
it comes to the fundamental framework and the accountability 
that is built in NCLB.
    People like you should be very tough on people like me. 
Excuses for why our kids are not succeeding in public education 
are not going to help us solve the problem.
    That said, my view is an accountability system need not 
only be tough, it needs to be fundamentally credible. And the 
current NCLB system I don't think is ultimately long-term 
credible and sustainable.
    We are almost a decade out from 2014 so its impact is not 
fully clear yet, but let me tell you what I think the behaviors 
it is leading to in the trenches where people like I am working 
with those 1.1 million kids.
    By focusing on proficiency, what we do is we create 
perverse incentives in terms of those kids that the school 
system focuses on. The kids who are in my system--level three 
is proficient. Kids who are close to the line get the most 
attention, because you are going to get the credit for bumping 
them over.
    You take a kid from level 2.1 to 2.9--that is a huge gain. 
You don't get any credit for it. You take a kid from 2.9 to 
three--that is a big credit event.
    People out in the field get that, because AYP and city 
status matters to them. If you take a kid from level three to 
level four, which is where we want our kids, future leaders of 
this country--we want our highest performing kids to be moving 
forward, too.
    Level three to level four--you get no credit for it in the 
system. We group them together. It is a mistake. It is a major 
mistake. Every kid has got to be moving forward.
    The derivative effect of that is we often engage in what I 
call zip code educational policy, because in my school system 
we have some communities where kids are coming to school ahead 
of the game and are performing at level three and four before 
they got to school. And yet that school is perceived to be a 
good school whether it is moving those kids forward or not.
    We have got other kids--schools where there are high 
concentrations of English language learners, special ed kids 
and so forth. Those kids may have a longer road to travel, but 
they may be moving along that road and still not getting the 
kind of recognition they deserve.
    It is incredibly important to close this gap, but you can't 
leap across this gap in one fell swoop. And schools that are 
bringing over those years a sustained growth deserve credit at 
every end of the spectrum.
    I don't want us to leave those level threes at level three 
in New York, and I want those schools to get credit. That is 
precisely what we are doing in our city. We are putting in 
place--building on NCLB an accountability system that will give 
real credit for schools that are moving forward on growth.
    We are going to do this in a way that looks at exactly the 
kind of thing that Dr. Shaul said. Year to year, how did this 
year's 3rd-graders do in the 4th grade, do in the 5th grade, do 
in the 6th grade?
    Now, we are going to take into account where schools start 
and how they move forward. And we are not going to allow the 
kind of what I think is perverse responsive behavior of using 
gimmicks to try to move the number of kids to level three. We 
are going to look at every single kid in the system.
    We are going to give extra credit to those people who are 
moving the most challenging, the hardest-to-serve kids in the 
system. But we are going to look at every single kid.
    And we are going to provide the data and the information to 
the schools. An accountability system that is a gotcha game is 
not going to be sufficient for what we need. The information 
has got to be used by the schools.
    Schools that use assessment take the data and use that to 
improve their teaching practices and to make sure that we build 
the solid foundation for kids before they move forward.
    So we are creating a massive data bank and investing 
enormous amounts of resources in professional development, so 
our principals, assistant principals and teachers can use the 
data to drive accountability.
    At some point, an accountability system has to be a 
spotlight, not a hammer. We can do that.
    [The prepared statement of Mr. Klein follows:]

 Prepared Statement of Chancellor Joel Klein, New York City Department 
                              of Education

    Good morning. Thank you Chairman McKeon and Congressman Miller for 
inviting me to testify today. I would also like to thank Congressman 
Owens for his many years of service to New York City. You will be 
missed.
    This morning, I will discuss the promise of growth models from the 
vantage point of New York City, where we are implementing a hybrid 
system that tracks both growth and proficiency levels. But first, I'd 
like to step back and put my comments into context.
    For decades, school reformers have tried and failed to fix 
education in America, to ensure that American students do not fall 
behind their international peers and to ensure that all students in 
this country, no matter what their race or socio-economic status, are 
receiving the high-quality educations they deserve. The law that we're 
discussing today, No Child Left Behind, might not be perfect, but it is 
incredibly valuable because it recognizes that the achievement gap-the 
gap that separates our African-American and Latino students from their 
white and Asian peers-is the chief, though certainly not the only, 
problem in American schooling. When Congress passed NCLB, it helped 
America finally take responsibility for the fact that white and Asian 
students are performing four years ahead of African-American and Latino 
students in high school. Four years. And this law finally puts muscle 
behind the attempt to close that gap. It forces us to report student 
performance in grades three through eight by race. We can no longer 
mask the deficiencies of some students with outsized gains by others.
    Now, NCLB has problems, and I will talk about those today. But at 
the same time, it is critical that we all remember that NCLB is not 
just important. It's fundamental. I've never met a law that couldn't be 
improved. But to criticize the heart of No Child Left Behind is to 
refuse to take responsibility for the achievement gap-the most serious 
civil rights, social, and economic crisis facing America today. We 
should, of course, learn from our experiences and make a good thing 
better, but we should not consider diluting or destroying a law that 
forces us to confront our problems head on. We must not yield to the 
critics of NCLB because, I believe, their complaints are missing the 
law's broader significance.
    Now, to the topic of the day. For a long time, we have heard that 
we face a choice between absolute standards, like those in NCLB, and 
value-added approaches, like the ones we are discussing today. I don't 
think we have to choose. We need to keep our eye on the ultimate goal, 
which is ensuring that every child is at least proficient in reading, 
math, and the other core academic areas; if a school is not helping all 
children achieve high standards, the school still has work to do. But 
we also need to recognize that schools sometimes succeed in helping 
kids make real, substantial progress without boosting them all the way 
to proficiency. We all know it is wrong to call a school an unqualified 
success when children are not meeting standards in reading and math, 
but it is also wrong to call a school an unqualified failure when it is 
helping its students grow academically. Both absolute achievement and 
growth should count when we are judging schools.
    While it's important to strive for 100% proficiency, focusing on 
proficiency alone and ignoring gains can lead to serious negative 
consequences.
    First, the current the law, which shortchanges growth, motivates 
educators to help certain types of students and not others. The way the 
law is designed, we get credit if we move a child across the threshold 
of ``proficiency.'' So, if a school wants to succeed under No Child 
Left Behind, it makes sense to dedicate more time and energy helping 
students who are just on the cusp of meeting standards than to students 
who are far away from proficiency or to students who are already safely 
above the threshold. Gains achieved by students who make progress but 
fail to reach proficiency are not rewarded. If children are so far 
behind that they have little prospect of achieving proficiency, it may 
make more sense under NCLB to encourage them to drop out than to help 
them achieve at the highest levels possible. Similarly, it's not worth 
giving the brightest kids-the students who could become the future 
leaders of this nation-the extra support and attention they need to 
achieve their potential because boosting children further above 
proficiency brings no added benefit under the law.
    Second, the law can give a misleading picture of how well a school 
is doing. We have schools in New York filled with kids who score above 
the proficiency threshold even though the schools are doing very little 
to help students progress beyond where they started on the first day of 
school. These schools pass under NCLB. At the other extreme, we have 
other schools that make real contributions to student learning and 
consistently help students move in the right direction, but don't get 
them all the way to proficiency. These schools do not get credit under 
NCLB.
    I know some people worry that growth models will give weak 
educators an excuse to give weak schools passing grades just because 
they are showing marginal gains. I say it depends on the growth model. 
It's true that a growth model could be designed as a smokescreen, but 
it could also be crafted to accurately gauge student learning. The 
latter is what we are creating in New York.
    Mayor Bloomberg and I have been very tough on accountability. We 
eliminated social promotion in our elementary and middle schools. We 
shut down failing high schools. We created ``Empowerment Schools,'' 
schools whose principals receive greater autonomy in exchange for 
entering specific performance contracts agreeing to be held accountable 
for results. More than 300 principals volunteered, knowing that they 
could lose their jobs if they were unable to raise student achievement.
    The sophisticated growth model we are crafting is another way we 
hold our schools accountable for providing New York City schoolchildren 
with the educations they need and deserve. Our new measure of progress 
will be more precise than what is required under No Child Left Behind. 
We will measure all year-to-year gains, even those that don't boost 
students to proficiency, and all losses. I think this will be a vast 
improvement that will start placing incentives where they belong: on 
educating all students. Here's how it will work: According to NCLB 
Student A's progress from a 2.0 to a 2.9 on New York's four-point scale 
counts for nothing and Student B's drop from a 2.8 to a 2.2 does not 
count against his school. Student C's progress by a tenth of a point, 
however, from a 2.9 to the point of proficiency, a 3.0, matters. If 
progress is our goal, NCLB's incentives system is out of whack. New 
York City's new accountability system will provide more appropriate 
incentives. We want teachers to focus on all students, not just those 
who are most likely to jump across an arbitrary threshold.
    Our new value-added information will feed into annual school 
progress reports, which will be completely transparent and publicly 
available for all of our 1,400 schools. These reports will measure 
schools on three quantitative factors: School Environment, Performance, 
and Progress. School Environment is a combination of attendance and 
safety data, as well as student, teacher, and parent survey results. 
Performance is what NCLB now relies on-snapshots of student performance 
in a given year. Progress is our ``value added'' measurement. It 
measures aggregated individual student performance over time. Based on 
these three factors, with Progress counting most, our schools will 
receive grades of A, B, C, D or F.
    These grades will tell parents how well their students' schools are 
performing-and whether they are making progress. Grades will also help 
administrators take responsibility for the schools, ensuring that 
children don't remain in failing schools. We will intervene to make 
improvements, and if we need to, we will shut down failing schools or 
change school leadership. Under our growth model, we expect that about 
15% of schools will receive Ds or Fs.
    We need to use data not only to evaluate how schools are 
performing, but also to enable principals and teachers to identify 
their strengths and weaknesses, to share the practices that work and 
improve the ones that don't. So, starting next year, we will also 
implement a comprehensive achievement reporting system, which will 
include all the information from the state standardized exams as well 
as information from periodic in-class assessments, which are no-stakes 
diagnostics that teachers will use to check in on their students. Using 
these systems, teachers will be able to give mini-assessments to learn 
whether students have grasped what they've been taught. If they have, 
teachers will move to the next subject. If they haven't-or if certain 
students are still struggling-teachers can intervene. These assessments 
will help educators measure progress over the course of the year and 
make mid-course corrections for students or classes with difficulties. 
They will no longer have to wait for end-of-year standardized exams to 
learn if students needed extra help.
    Some of our schools are already using data to drive performance, 
but too many of our educators are estimating and guessing, even though 
we're working in an age of technology, an age when educators don't have 
to guess what the problem is and experiment until they find the right 
solution. In its current form, NCLB does not motivate educators to help 
all children achieve at the highest levels possible. I believe New York 
City's new system's will help to make this possible.
    Thank you. I welcome your questions.
                                 ______
                                 
    Chairman McKeon. Thank you very much.
    Dr. Weaver?

    STATEMENT OF REG WEAVER, PRESIDENT, NATIONAL EDUCATION 
                          ASSOCIATION

    Mr. Weaver. Thank you very much. Thank you for the 
opportunity to join you this morning to share the views of 2.8 
million members of the National Education Association.
    The NEA is the largest professional association in the 
country, and our membership is diverse, but we have a common 
mission and values based on our belief that a great public 
school is a basic right for every child.
    NEA and our members have long supported the Elementary and 
Secondary Education Act. And despite what some say, we have 
always supported the laudable goals of No Child Left Behind, 
closing the achievement gaps, raising overall student 
achievement and ensuring all students have a qualified teacher.
    NEA did not oppose annual testing, nor did we oppose the 
passage of No Child Left Behind. We did, however, make a clear 
statement as to our objection to accountability systems based 
solely upon test scores.
    And I hope that this committee understands that expressions 
of concern about the impact of this law are not a rejection of 
the goals of No Child Left Behind, nor do we want to do away 
with the law.
    In fact, I just returned from our representative assembly 
where almost 9,000 delegates voted on NEA's priorities for 
elementary and secondary act reauthorization. And they didn't 
vote to repeal nor do away with No Child Left Behind.
    Instead, they voted on a comprehensive set of proposals 
designed to fix what is not working with the law and add to it 
the kind of initiatives that will make our goals a common 
reality.
    And our report spells out seven key components of a great 
public school: Quality programs and services; high expectations 
and standards; quality conditions for teaching and lifelong 
learning; qualified, caring, diverse and stable staff; shared 
responsibility for appropriate school accountability by 
stakeholders at all levels; parental, family and community 
involvement and engagement; and adequate, equitable and 
sustainable funding.
    Growth models--well, we applaud Secretary Spellings' 
decision to pilot a growth model project, and we thank her for 
listening to educators and being willing to explore new 
options.
    We believe that accountability systems should reward 
success and support educators to help students learn. And 
measuring student growth over time will be more helpful than 
the current snapshot approach which measures student 
achievement on 1 day at a time throughout the year.
    I taught middle school science. And as a veteran classroom 
teacher, I would welcome the opportunity to use my students' 
test results to guide my instructional practice.
    But I seriously question the logic of any system that 
mandates tests but does not also mandate that the results of 
those tests be given in time to make any adjustments in 
instruction.
    And if someone had told me that my class the next year 
would be tested in the spring, and their scores would be 
compared to my students' scores from last year, I would have 
said that there is something inherently wrong with that system.
    An accountability system designed to measure performance 
cannot compare apples and oranges. And as a science teacher, I 
know that such a system will not yield any meaningful data.
    The implementation phase of No Child Left Behind has 
highlighted a critical void in assessing student progress, 
measuring student progress over time, and providing the 
resources and tools that educators need to get the job done.
    In our opinion, it does not recognize that children are 
human beings and not cardboard cutouts and that teaching them 
is both an art and a science.
    Including growth models in No Child Left Behind's 
accountability system would not mean abandoning requirements 
that all students read and complete math problems on grade 
level. Quite the opposite is true. Growth models hold greater 
promise to demonstrate whether a student is learning and 
provides data to educators in order to inform their 
instruction.
    Better data will show when instructional techniques are not 
working and will allow teachers to make adjustments to meet 
student needs.
    Growth models also reward success in teaching and learning 
by giving schools credit both for moving a child from below 
basic to basic, as well as moving a child from proficient to 
advanced.
    I would like to offer one caution, however. Including a 
growth model as part of adequate yearly progress--it is not a 
panacea. Complexities will continue to arise for some English 
language learners and certain students with disabilities who 
take alternate assessments.
    And we will also continue to need much more research about 
growth models as well as technical assistance to states and 
local districts and educators, but the ultimate goal should be 
to help classroom educators use the data to inform instruction.
    And I have made closing the achievement gap one of our 
organization's highest priorities. And it is not only something 
that I care deeply about, but I believe it is the right thing 
to do.
    My written testimony highlights the many ways in which the 
NEA has marshaled our resources to assist our state and local 
affiliates to help close the achievement gaps. In short, we 
have devoted millions of dollars and thousands of hours of 
staff time to that effort.
    And I would like to conclude by sharing a story from one of 
our members. It is a story about a boy named Cesar and his ESL 
teacher, Mary Beth Solano in Fort Collins, Colorado.
    And she writes: ``Cesar, a 3rd-grade student, came to me in 
August with not a word of English. And together with his 
classroom teacher, significant work from me and a fantastic 
group of peers, he learned English amazingly well.
    ``And I was almost going to exempt him from the test, but 
two things stopped me: One, part of No Child Left Behind that 
says that any child who doesn't take the test counts as a zero 
on our school's report card, and Cesar himself, who set as his 
personal goal knowing enough English to pass Colorado State's 
mandated No Child Left Behind test.
    ``And reluctantly but with Cesar's terrific desire to 
exceed coaxing us on, we had him take the test. And he 
struggled and he struggled, reading every word, and read it 
over and over again until he understood and thought he had 
understood well enough to answer the question.
    ``And he worked so hard and so long on the test, it brought 
tears to my eyes. And he was so proud of himself after he 
completed it. You should have seen his face.
    ``Well, the scores came out. While we celebrated, his 
family was devastated. Cesar earned a score that was only two 
tiny points below the cutoff for partially proficient, but he 
was labeled an unsatisfactory learner. Nothing could be farther 
from the truth, for he had gone from basically zero to almost 
proficient in just 6 months.
    ``And I tried to explain that to his parents and tried even 
harder to lift Cesar's spirits. But I am not sure how much 
success that he had.
    ``So without changes in the structure and the process of 
reporting scores, stories like Cesar's will continue to deflate 
and demoralize the best and the brightest.''
    So I encourage every one of you, every one of the members 
of the committee, to talk to your local educators about their 
experiences and ask them about their frustrations. But more 
importantly, ask them about their successes.
    And they all have a Cesar story, and they all want every 
one of their students to succeed. And you know what? They go 
every day above and beyond the call of duty, time and time 
again, to make that goal a realty for America's children.
    Thank you.
    [The prepared statement of Mr. Weaver follows:]

    Prepared Statement of Reg Weaver, President, National Education 
                              Association

    Good morning Chairman McKeon, Congressman Miller, and Members of 
the Committee.
    Thank you for the opportunity to join you this morning and share 
the views of the 2.8 million members of the National Education 
Association (NEA).
    NEA is the largest professional association in the country, 
representing public school educators-teachers and education support 
professionals, higher education faculty, educators teaching in 
Department of Defense schools, students in colleges of teacher 
education, and retired educators across the country. While our 
membership is diverse, we have a common mission and values based on our 
belief that a great public school is a basic right for every child.
    Our members go into education for two reasons-because they love 
children and they appreciate the importance of education in our 
society. We want all students to succeed. Our members show up at school 
every day to nurture children, to bring out their full potential, to be 
anchors in children's lives, and to help prepare them for the 21st 
century world that awaits them. It is their passion and dedication that 
informs and guides NEA's work as we advocate for sound public policy 
that will help our members achieve their goals.
I. NEA Principles for Great Public Schools
    Today's hearing focuses primarily on the use of growth models in 
the No Child Left Behind (NCLB) accountability framework. Before I 
address that specific point, however, I would like to take a moment to 
make a few broader points about NEA's principles and goals for ensuring 
great public schools.
    NEA and its members have long supported the Elementary and 
Secondary Education Act (ESEA). During the last reauthorization of 
ESEA, we supported the laudable goals of No Child Left Behind-closing 
achievement gaps, raising overall student achievement, and ensuring all 
students have a qualified teacher. We also supported a number of 
specific elements in the new law, including the targeting of Title I 
funds to the neediest schools and students; disaggregation of test data 
by subgroup; and programs for dropout prevention, after-school learning 
opportunities, and math and science education. We continue to support 
all of these elements.
    NEA did not at any time oppose annual testing nor did we oppose 
passage of the No Child Left Behind Act. We did, however, make clear to 
Congress our objection to accountability systems based solely upon test 
scores. We also made clear that any tests used in an accountability 
system had to be valid and reliable, aligned with the curriculum, and 
designed to inform instruction, and that the system had to allow for 
accurate and fair measurement of test results.
    During the debate on No Child Left Behind, we suggested two ways to 
craft a more workable, accurate, and fair accountability system. First, 
we suggested the use of multiple measures of student achievement and 
school quality to determine school effectiveness. Second, we suggested 
creating accountability systems that not only required certain 
proficiency levels, but that also measured growth in student 
achievement over time. We supported, and continue to support, these 
polices because, while we deeply believe that all children can learn, 
we know that not all children learn at the same rate or in the same 
way.
    Let me be perfectly clear that our criticisms during initial debate 
and our continuing expressions of concern over implementation of the 
law are not rejections of the goals of No Child Left Behind. Nor do 
they reflect a desire to do away with the law.
    In fact, I have made closing the achievement gaps one of NEA's 
highest goals. It is not only something about which I care personally; 
it is the right thing to do. As someone who taught for 30 years, I know 
that change doesn't happen overnight. But, I also know that if we are 
to achieve the change we seek, we cannot ignore the experiences of 
those working in our classrooms every day. Rather, we must translate 
the lessons we learn from our nation's educators into sound, workable 
policies that will help us meet our goals.
    I just returned from NEA's annual meeting where almost 9,000 
delegates voted on NEA's priorities for ESEA reauthorization. They 
didn't vote to repeal or do away with NCLB. Instead, they voted on a 
comprehensive set of proposals designed to fix what's wrong with the 
law and add to it the kinds of initiatives that will make our common 
goals a reality. A copy of that report is attached as Appendix I. I 
hope it will help guide the committee as you approach reauthorization.
    Our report spells out what we believe to be the seven key 
components of a great public school:
    Quality programs and services that meet the full range of all 
children's needs so that they come to school every day ready and able 
to learn;
    High expectations and standards with a rigorous and comprehensive 
curriculum for all students;
    Quality conditions for teaching and lifelong learning;
    A qualified, caring, diverse, and stable workforce;
    Shared responsibility for appropriate school accountability by 
stakeholders at all levels;
    Parental, family, and community involvement and engagement; and
    Adequate, equitable, and sustainable funding.
    The priorities detailed in the report are very consistent with the 
views of our general membership. In fact, our recent member poll found:
    57 percent of our members want major changes to No Child Left 
Behind, 21 percent want minor changes, 17 percent favor repeal, and 
just 4 percent want to keep the law as it is;
    95 percent of our members want the Association to be active in 
working to change the law, while only 4 percent disagree;
    85 percent of our members believe there is too much reliance on 
standardized testing.
    Our members rated highest the following proposals to improve NCLB:
    Use multiple measures instead of just tests (71% strongly favor);
    Measure student achievement over time instead of just the day of 
the test (70% strongly favor);
    Ensure that employee rights under contracts and laws are respected 
(65% strongly favor);
    Fully fund mandates (64% strongly favor);
    Restore the class size reduction program (63% strongly favor).
II. Growth Models and Effective Accountability Systems
    I would now like to turn to the focus of today's hearing, the use 
of growth models in measuring student progress and school 
effectiveness. As I have stated, we believe that accountability systems 
should be based upon multiple measures, including local assessments, 
teacher-designed classroom assessments collected over time, portfolios 
and other measures of student learning, graduation/dropout rates, in-
grade retention, percentage of students taking honors/advanced classes 
and Advanced Placement exams, and college enrollment rates. We strongly 
believe that the current one-size-fits-all system is unacceptable and 
that states need the flexibility to design systems that produce 
results, including deciding in which grades to administer annual 
statewide tests.
    Accountability systems should reward success and support educators 
to help students learn. To this end, any improved accountability system 
should allow for use of growth models and other measures that assess 
student learning over time and recognize improvement on all points of 
the achievement scale. These measures should then be used as a guide to 
revise instructional practices and curricula, provide individual 
assistance to students, and tailor appropriate professional development 
for teachers and other educators. They should not be used to penalize 
teachers or schools.
    We applaud Secretary Spellings' decision to pilot a growth model 
project. Her decision signaled that she has heard what our members have 
had to say, and we thank her for that. We also applaud her decision to 
allow states to propose their own growth models for peer review, rather 
than prescribing a certain type of model. This flexibility was 
particularly welcome given that all states were testing prior to 
enactment of NCLB and 15 states were already testing annually in grades 
three though eight. We have recently completed a policy brief on the 
growth model pilot program and the process used by the Department of 
Education to approve proposals by two states (NC and TN). This policy 
brief is attached as Appendix II to my statement.
    Our members believe that measuring student growth over time will be 
more helpful than the current snapshot approach, which measures student 
achievement on one day out of the year. A growth model approach will 
allow for a more accurate reflection of student learning and will help 
inform instruction.
    I taught middle school science for 30 years. If someone had told me 
that my students would be given a state standardized test in the spring 
and that I would not receive the results of those assessments in time 
to make any instructional adjustments, I would have seriously 
questioned the logic of the central testing office. If someone had then 
told me that my class the next year would be tested in the spring and 
that their scores would be compared to my students from last year, I 
would have said there was something inherently wrong with the system.
    An accountability system designed to measure performance cannot 
compare apples and oranges. As a science teacher, I know that such a 
system simply will not yield any meaningful data. The children I teach 
in any given year will have completely different educational needs than 
the children I teach in the following year. NCLB fails to recognize 
that children learn in different ways and at different rates. It fails 
to recognize that children are human beings, not widgets in a factory, 
and that teaching them is both an art AND a science.
    One of our members from Rockford, IL has noted the illogical 
consequences of the current system:
    ``Jackson Elementary School teachers worked tirelessly in the first 
year of corrective action to bring up scores to the level set by NCLB. 
The students made incredible gains, unfortunately they missed AYP by 
less than one percent. This translates to one or two students that made 
gains, but not enough to bring them to the prescribed level. Therefore, 
they are in their second year of corrective action and labeled as a 
failing school.''
    The current system simply fails to provide useful, timely data for 
diagnosing learning problems and facilitating instructional changes. 
Rather, students who are tested in one grade move on to the next grade, 
and their new teacher receives their test results-results that have 
virtually no relevance to the choices that new teacher will make in 
instructional strategies.
    Not only is the current underlying system flawed, but 
implementation is also troubling. NCLB requires assessments to be built 
upon states' content standards, which in turn are to be aligned with 
statewide assessments. Yet, four and a half years into the law, only 
ten states have received full approval from the Department of Education 
for their content standards and assessment systems. To educators, this 
translates as a lack of interest in what is tested and whether the test 
content has actually been taught in the classroom. It appears that the 
goal is simply to administer tests and assign accountability labels. 
This is demoralizing to educators and contradictory to sound 
educational practice.
    NEA is not alone in supporting an improved accountability system 
that allows for use of more accurate measures. We have led an effort to 
develop consensus on a broad set of principles for ESEA 
reauthorization. To date, 87 organizations have endorsed these 
principles, one of which calls for use of growth models as part of an 
accountability framework (See Appendix III for the complete Joint 
Organizational Statement on NCLB with the list of signatories). In 
addition, nine bills that would allow use of growth models are 
currently pending in Congress. Some of those bills were introduced by 
members of this committee, including Representatives McCollum, Wu, and 
Andrews. Several other committee members, including Representatives 
Grijalva, Ryan, and Woolsey, have cosponsored bills that would allow 
states to utilize growth models.
    Governors and state legislators have also called on a bipartisan 
basis for more flexibility to use growth models. The National Governors 
Association's (NGA) proposals for the ESEA reauthorization, issued in 
March 2006, state that, ``Maximum flexibility in designing state 
accountability systems, including testing, is critical to preserve the 
amalgamation of federal funding, local control of education, and state 
responsibility for system-wide reform.''
    Similarly, the National Conference of State Legislatures issued a 
report in February 2005 calling on Congress to make substantial changes 
to the law. The report states:
    ``Administrators at the state, local and school levels are 
overwhelmed by AYP because it holds schools to overly prescriptive 
expectations, does not acknowledge differences in individual 
performance, does not recognize significant academic progress because 
it relies on absolute achievement targets, and inappropriately 
increases the likelihood of failure for diverse schools.''
    By allowing inclusion of growth models in NCLB's accountability 
system, Congress would not have to abandon the requirement that all 
students read and complete math problems on grade level. Quite the 
opposite is true. We believe that growth models hold greater promise to 
demonstrate whether a student is learning. They would provide a more 
accurate measure by giving schools credit both for moving a child from 
below basic to basic as well as moving a child from proficient to 
advanced. They would also offer a way to recognize highly effective 
schools that have an influx of students who are not performing at grade 
level.
    Growth models will also help overcome the all-or-nothing approach 
of measuring Adequate Yearly Progress (AYP). Currently, a typical 
school has to meet 37 criteria to make AYP. A school that falls short 
on just one of the 37 is treated in the same manner as a school that 
fails all 37 criteria. (See tables below) Growth models that offer more 
common sense ways to measure student achievement, in particular for 
students with disabilities and English Language Learners (ELL), will 
ameliorate this problem.

                                    CURRENT AYP FAILURE SCHOOL: MISS 1 OF 37
----------------------------------------------------------------------------------------------------------------
                                                                                                       Other
                                        Reading           95%            Math           95%          indicator
----------------------------------------------------------------------------------------------------------------
All Students......................  Yes             Yes             Yes            Yes            Yes
Black/African American............  Yes             Yes             Yes            Yes
Hispanic..........................  Yes             Yes             Yes            Yes
Native American...................  Yes             Yes             Yes            Yes
Asian.............................  Yes             Yes             Yes            Yes
White.............................  Yes             Yes             Yes            Yes
ELL...............................  Yes             Yes             Yes            Yes
Poverty...........................  Yes             Yes             Yes            Yes
Disability........................  Yes             Yes             Yes            No
----------------------------------------------------------------------------------------------------------------


                                      CURRENT AYP FAILURE SCHOOL: MISS ALL
----------------------------------------------------------------------------------------------------------------
                                                                                                       Other
                                        Reading           95%            Math           95%          indicator
----------------------------------------------------------------------------------------------------------------
All Students......................  No              No              No             No             No
Black/African American............  No              No              No             No
Hispanic..........................  No              No              No             No
Native American...................  No              No              No             No
Asian.............................  No              No              No             No
White.............................  No              No              No             No
LEP...............................  No              No              No             No
Poverty...........................  No              No              No             No
Disability........................  No              No              No             No
----------------------------------------------------------------------------------------------------------------

    However, the use of growth models will not completely eliminate 
this problem. If one subgroup makes AYP using a growth model, while 
another subgroup in the school does not make AYP (even using the growth 
model), the school will still be designated as failing AYP. In this 
instance, the use of the growth model doesn't eliminate the ``all or 
nothing'' approach.
    I would also note that there are differences among growth models, 
with varying levels of complexity. Some states, like North Carolina, 
have had to implement a different growth formula for students not 
already proficient than for students who are proficient. This was 
necessary because of the federal requirement that proficiency be the 
end result of any growth trajectory. Obviously, such a model does not 
work for students who already are proficient.
    I would like to close my comments on growth models by reminding the 
committee that having a growth model as part of the AYP process is an 
improvement but it will not be a panacea. Getting certain students on 
track to proficiency within a four-year timeline, as is required under 
North Carolina's approved model, will still be a challenge for many 
schools. In addition, complexities will continue to arise for some ELL 
students or certain students with disabilities who take alternate 
assessments. We will also continue to need much more research about 
growth models as well as technical assistance to states, local 
districts, and educators to evaluate and use data, evaluate the models 
themselves, and replicate successful efforts. The ultimate goal should 
be to help classroom educators use data to inform instruction.
III. NEA's Work to Close Achievement Gaps
    It has been a majority priority of mine to marshal NEA resources to 
assist our state and local affiliates in seeking policy changes at the 
state and local level to help close achievement gaps. Our work has 
included:
    Committing more than $6 million through NEA Foundation grants to 
close achievement gaps in urban school districts. Those grants fund 
programs with clear goals of improving literacy and math and science 
achievement; helping stabilize quality staff; and involving families 
and communities involved in the learning process. In two of the grant 
sites, Hamilton County, Tennessee and Milwaukee, Wisconsin, school 
faculties use growth data to assess progress in reading and mathematics 
and determine appropriate interventions for instructional improvement. 
Reaching ambitious growth targets provides confidence and positive 
reinforcement to teachers and students who have large gaps to overcome 
and helps teachers and administrators set continuous benchmarks for 
progress and observe what works in changing instructional practice. For 
example, last year in Hamilton County, Tennessee, the five schools 
targeted under the NEA grant set and achieved a goal of 115 percent of 
the expected growth according to state standards per annum in reading 
and mathematics achievement. While these schools have not yet all 
reached high levels of achievement compared to the state's affluent 
schools, they have made greater gains than many of the top-ranked 
schools. By significantly accelerating the rate of achievement, low-
performing schools can close achievement gaps, while all schools 
continue to make progress.
    Delivering trainings and products on a variety of instructional 
issues, including closing the achievement gaps, to our members and 
leaders across the country.
    Sponsoring statewide National Board Certified Teacher summits 
focused on recruiting and retaining accomplished teachers in high-need, 
high-poverty schools with low student achievement.
    Developing and sharing with all NEA affiliates our Closing 
Achievement Gaps: An Association Guide-a blueprint for closing the 
gaps.
    Awarding grants to ten states (Colorado, Minnesota, Illinois, 
Indiana, Michigan, Missouri, New Mexico, Nebraska, Oklahoma, and 
Pennsylvania) focused on closing gaps through changes in state policies 
such as early childhood education, coaching for new principals, 
statewide teacher induction programs, and revised state professional 
development guidelines. We have also provided grants to eight states 
for their work to encourage highly skilled teachers to move to and stay 
in high-need schools.
    Continuing the partnership established in 2005 with the Tom Joyner 
Foundation to increase the percentage of highly qualified Black 
teachers in targeted high-needs communities. The Teacher Licensure 
Scholarship Program, funded by the Tom Joyner Foundation, provides 
financial assistance to historically Black colleges and universities to 
assist minority teachers in preparation for licensure exams. More than 
250 scholarships have already been awarded nationally through this 
program.
    Developing training modules on closing the achievement gaps for use 
in community conversations. This work has been piloted in Mississippi 
and Florida and will be used by the end of this program year in three 
of the seven states that have been awarded grants to convene community 
conversations.
    Surveying, through the Center for Teacher Quality, teachers in 
three states (Kansas, Arizona, Ohio,) to identify the necessary working 
conditions to achieve optimum teaching and learning environments.
    Developing online professional development focused on helping 
teachers become more effective with a diverse student body.
    Awarding grants to recognize model teacher retention program 
through a Saturn-UAW-NEA partnership.
    Launching and maintaining an easily accessible, interactive Website 
to help our affiliates and the general public research and locate 
resources about the achievement gaps (www.achievementgaps.org).
    In the near future, NEA will develop a program guide outlining 
effective support strategies for minority candidates pursuing National 
Board Certification, with the goal of increasing the percentage of 
minority National Board Certified teachers in high-need schools. In the 
next school year, we will convene community conversations in seven 
states as part of our Public Engagement Project (PEP) initiative. We 
will also develop additional educational materials for state affiliates 
on teaching and working conditions, and we will be announcing 
additional grants for 2006-07. Finally, NEA is building a state-by-
state database to identify policies, practices and programs that help 
close the achievement gaps. I would be happy to share additional 
information on any of our projects with the committee.
    Before I conclude, I would like to share just one of the many 
stories from countless NEA members about the impact of ESEA on them and 
their students. It is a story about a boy named Cesar and his ESL 
teacher, Mary Beth Solano in Fort Collins, CO. She writes:
    ``One of my recent student's stories is a prime example of how NCLB 
legislation, by labeling students unfairly, is demoralizing and needs 
to be changed. Cesar, a third grade student, came to me in August with 
not word one of English. Together with his classroom teacher, 
significant work with me, and a fantastic group of peers, he learned 
English amazingly quickly.
    I was almost going to exempt him from the test, but two things 
stopped me. One, the part of NCLB that says that any child who doesn't 
take the test counts as a zero for our school's report card, and Cesar 
himself, who set as his personal goal knowing enough English to pass 
CSAP (Colorado's state mandated NCLB test). Reluctantly, but with 
Cesar's terrific desire to succeed coaxing us on, we had him take the 
test. He struggled and struggled, reading every word * * * over and 
over again until he thought he understood well enough to answer each 
question.
    He worked so hard and so long on that test, it brought tears to my 
eyes. He was so proud of himself after he completed it, you should have 
seen his face. Remember, he had only been working in English since 
August and the reading test was given in February * * * six short 
months with a new language, and he took the same test as native 
speakers did. He kept asking what his score was, and actually looked 
forward to the day his parents would get `the letter'. Well, the scores 
came out, and while we celebrated (understanding statistics and 
scoring), he and his family were devastated. Cesar earned a score that 
was only two tiny points below the cut off for partially proficient on 
CSAP, but below the cut off it was, so he was labeled an 
`Unsatisfactory' learner.
    Nothing could be farther from the truth for he had gone from 
basically zero to almost proficient in just six months (something no 
politician has ever done), but the federal government didn't care about 
that effort or progress. To the feds, the state and the public he was 
unsatisfactory. I tried explaining it all to his parents, and tried 
even harder to lift Cesar's spirits, but I'm not sure how much success 
I had. Without changes in the structure and process of reporting 
scores, stories like Cesar's will continue to deflate and demoralize 
the best and brightest students.''
    On behalf of all 2.8 million members of the National Education 
Association, I want to thank you for this opportunity. We look forward 
to working with you throughout the reauthorization process. I encourage 
every member of this committee to talk to your local educators about 
their experiences. Ask them about their frustrations. But more 
importantly, ask them about their successes. When you do, you'll get as 
clear a sense as I have. They all have a ``Cesar'' story. They all want 
every one of their students to succeed. And they go above and beyond 
the call of duty time and time again to make that goal a reality for 
America's public school students.
    Thank you. I will be happy to answer any questions.
                                 ______
                                 
    Chairman McKeon. Thank you.
    Ms. Haycock?

    STATEMENT OF KATI HAYCOCK, DIRECTOR, THE EDUCATION TRUST

    Ms. Haycock. Mr. Chairman, Mr. Miller, and members of the 
committee, thank you for inviting me to come speak with you 
this morning about whether the addition of a growth component 
can improve the NCLB accountability system.
    As you know, I head the Education Trust, and in that role I 
spend an awful lot of time with educators around the country, 
including on accountability matters. I also serve as one of 10 
members of the peer review committee assembled by the secretary 
of education to review the proposals from states.
    My simple answer to your overarching question this morning 
is yes, incorporating growth models can improve accountability 
systems. Such improved systems won't necessarily guarantee, in 
and of themselves, more quality education, but they can help.
    The reason for that is because accountability systems that 
grant appropriate credit for growth can help in four ways.
    First, they can help local and state education leaders to 
prioritize the schools that need the most help by 
distinguishing between schools where the kids may not yet be at 
proficient, but they are on their way to being proficient in a 
few years, and those schools where kids are on a trajectory to 
get nowhere at all.
    Second, they can help ensure that the problem that 
Chancellor Klein talked about, schools focusing just on kids 
right below the proficient bar, doesn't happen as often, that 
they are working on growth for all kids.
    Third, they can help education leaders set stretch goals 
for all schools, including the higher achieving schools that my 
friend Bill Sanders here talks about as the slide-and-glide 
schools.
    And finally, they can help to get rid of the perception by 
some people at least that the NCLB accountability system is 
arbitrary and unfair. That said, suggesting that the inclusion 
of those measures can improve accountability systems is 
different from guaranteeing that the inclusion of such 
components will improve the system.
    So let me focus, if I can, on the principles and conditions 
that will help to maximize impact. But before I do that, just a 
word or two on the peer review process. I think you know that 
the secretary assembled a quite diverse group of peer 
reviewers.
    It included representatives from state education agencies, 
local school districts, assessment experts and child advocates. 
We were a diverse group in every way, including politically.
    Despite that, however, we came to full agreement on every 
single substantive issue that we have faced. And in fact, the 
conversations among the peer reviewers were so substantive that 
we actually wrote up a public summary of that that we all 
signed in order to help states assemble future growth 
proposals.
    I do want to share with you, though, the two most 
fundamentally important findings from that group. First, 
despite perceptions that the current AYP system is 
unnecessarily rigid and crude, the current system actually gets 
it right most of the time.
    It turns out that schools that don't make AYP in a system 
that does not include a growth component actually most often 
don't make AYP in a system that does include it. In both of the 
states that we approved, fewer than 50 schools that didn't make 
AYP actually made AYP with a growth component included.
    Second, it is really important to understand that growth to 
proficiency in most cases is actually a higher bar, a tougher 
standard, than the current status system.
    In North Carolina, for example, if accountability systems 
decisions had been made entirely on growth, actually, fewer 
schools would have made AYP instead of more. Why is that? 
Because lots of schools that are now over the status bar are 
actually not growing their kids very much.
    So growth isn't an easier standard at all, but done right 
it can help. Let me talk about four or five principles for 
doing it right. No. 1, as Mr. Miller suggested, it is really 
important and you insist on growth to proficiency. Any old 
growth won't do.
    No. 2, once a student's growth trajectory is established 
with a goal of proficiency within 4 years, you don't want to 
allow it to be reset and reset again.
    Third, growth models need to set goals for proficient 
students as well. The reason for that is very simple. A lot of 
students who are proficient 1 year won't be proficient the next 
year or the year after unless they get some focused attention 
as well.
    Fourth, just as in the current law you have not allowed 
schools to average and mask the low performance of some kids 
with the high performance of others, it is also important that 
you not allow schools to mask the low growth of some students 
with the high growth of others.
    Finally, it is always important to ask the question growth 
to what. Just as averaging performance can mask the 
underachievement of some groups of kids, too low a standard can 
mask problems with the performance of students.
    We need to have high and rigorous standards if growth is 
something--is actually to matter at all. We need to be focused 
more on where kids are going than simply where they have been.
    One final comment, though. Improving NCLB's accountability 
system is an important undertaking. But we need to keep it in 
perspective. When you look overseas, what you find is most 
other countries sort of got about the business of standards and 
accountability fast and then moved on to improvement.
    We Americans, however, keep obsessing about how to get that 
accountability system a little bit better instead of focusing 
on what is really important, and that is improving teaching and 
learning, helping low performance schools get better, making 
sure we have quality teachers and that they are distributed 
fairly, and also providing the curriculum and benchmark 
assessment that teachers need to make sure they are doing the 
job right.
    In the end, that is what is going to get us to higher 
achievement, not simply a better accountability system. Thank 
you.
    [The prepared statement of Ms. Haycock follows:]

   Prepared Statement of Kati Haycock, Director, the Education Trust

    Mr. Chairman, Mr. Miller, and Members of the Committee, thank you 
for the opportunity to testify this morning regarding the potential of 
growth measures to improve school accountability determinations.
    As you know, I head the Education Trust, an independent, non-profit 
organization focused on improving achievement and closing gaps between 
groups, pre-kindergarten through college. In that role, I and my 
colleagues do a lot of work with both states and school districts, 
including frequent opportunities to advise on and observe the effects 
of accountability systems. I also served as one of ten members of the 
Peer Review Team assembled by Secretary Spellings to advise her on 
proposals for the Growth Model Pilot program.
    My simple answer to your overarching question is, yes, 
incorporating growth measures can improve accountability systems. They 
don't in and of themselves ensure improved education for all students, 
but growth measures can help.
    First, using growth measures can help education leaders to 
prioritize the schools that need the most help. Giving schools credit 
for growth in students' learning can help distinguish between schools 
where, though students are not yet proficient, they are on a trajectory 
to become proficient soon-and those where students are on a trajectory 
to nowhere. These more nuanced determinations can help to ensure that 
help and resources are targeted to the schools and students with the 
most acute needs.
    Second, establishing credit for growth can help to ensure that 
schools don't focus inordinate attention on students who are just below 
proficient, but rather seek to grow the knowledge and skills of 
students at all ends of the achievement spectrum--including the middle-
achieving low-income and minority students who could become high 
achievers with more challenging instruction.
    Third, the inclusion of growth measures can help education leaders 
to set stretch goals for all schools, including the higher achieving 
schools that Bill Sanders often calls the ``Slide and Glide'' Schools.
    Finally, establishing credit for growth can reduce the perception 
of some that the NCLB accountability system is arbitrary and unfair, 
especially to schools that receive large numbers of underprepared 
students.
    That said, suggesting that the inclusion of growth measures can 
improve accountability systems is different from guaranteeing that it 
will. So rather than spending my time with you extolling the virtues of 
growth measures, let me focus instead on what I think I've learned 
about the principles and conditions that must undergird these or any 
other changes in accountability systems.
    But first a word about the peer review process.
The Peer Review Process
    In November 2005, the U.S. Secretary of Education announced a 
growth model pilot program, with the stated goal of informing the 
reauthorization of NCLB. Several months later, she named a panel of ten 
peer reviewers.
    The peer reviewers represented a broad cross section of experts and 
stakeholders, including practitioners from state departments of 
education and school districts, as well as outside researchers, 
assessment experts and child advocates. The group included both 
liberals and conservatives.
    Despite our different perspectives, however, we reached consensus 
on every significant aspect in our review of the states' growth model 
proposals. Indeed, our conversations were so constructive and 
substantive that the peer review panel took the extraordinary step of 
publicly releasing a statement of principles-signed by every member of 
the team--to help inform subsequent discussions of these issues. I'll 
draw from that statement in just a moment.
    I do, however, want to share what were perhaps the two most 
important findings from our work:
     First, despite perceptions that it is unnecessarily rigid 
and crude, the current AYP system actually gets it right most of the 
time. In both states that were approved to implement growth models, 
most of the schools that didn't make AYP without a growth component 
also did not make AYP with a growth component. As it turns out, schools 
with low achievement also tend to be schools with low growth. In each 
state, fewer than 50 schools that made growth targets that had not made 
AYP under the current rules.
     Second, in most cases, growth to proficiency is actually a 
higher bar than the current status system. In North Carolina, for 
example, if accountability determinations were based solely on growth, 
fewer schools would have met their goals. Why? Because a lot of schools 
that are now over the status bar are actually not growing their 
students' knowledge and skills very much.
    So growth isn't an easier standard at all. But, done right, it can 
both enhance the fundamental fairness of the system and provide the 
more nuanced information necessary to help leaders target resources and 
assistance to the schools and students with the most acute needs.
Principles
    So what are the principles that should undergird this or any other 
change in the NCLB accountability system?
    First and foremost, we must be very clear that any old growth won't 
do. Congress must insist on growth to proficiency. It's important to 
recall why you focused the nation on the goal of student proficiency in 
the first place. The reality is that there are absolute standards 
against which students will be judged, whether they go right to work or 
into postsecondary education. In the real world, there won't be 
allowances based on family background or parents' education level.
    If public education is going to serve as an engine of upward 
mobility, then expectations need to be pegged not just to where 
students come from, but where they need to go. A growth model can 
provide incentives to focus on students at all levels of achievement, 
but the goal must still be proficiency for all students.
    Some argue that this is somehow unfair to schools that serve 
concentrations of poor children. Our collective responsibility, 
however, must be to the students themselves. And frankly it's unfair to 
them not to require schools and districts to take responsibility for 
student achievement.
    Second, once a student's growth trajectory is established, it's the 
school system's responsibility to catch the student up within the 
designated time frame. If the expectation is that a student will reach 
proficiency in three or four years, the targets should not be reset 
downwards nor should the time frame generally be extended. Otherwise, 
students will never actually be expected to be proficient.
    Third, as the peer review committee suggested, growth models should 
set goals for proficient students. This is especially important because 
many students who perform at the proficient level one year will not be 
proficient in subsequent years without explicit attention to their 
needs. This is one of the reasons why, in recent years, we've observed 
a pattern on both NAEP and state tests in which more students are 
proficient on elementary tests than in the middle or high school 
grades. Growth models should orient educators toward students' success 
in successive years.
    Fourth, any growth models should retain NCLB's historic focus on 
individual students. If, as some suggest, we look at the average growth 
for a whole school, high growth with some students will mask the 
stagnant or slow growth--and even the academic decline--of other 
students. One of NCLB's strengths is that it does not allow schools to 
compensate for the under-education of low-achieving students by having 
a greater number of advanced students. Likewise, schools should not 
compensate for some students' stagnant growth by showing greater growth 
with higher-achieving students.
    Looking at the growth of individual students over time requires 
assessments that are aligned from year to year and longitudinal data 
systems. Many states still do not have these in place, limiting their 
ability to implement growth models. The quality of assessments and data 
systems is critically important to the accuracy and validity of growth 
measures, so focusing on state capacity and resources in this area 
needs to be a priority.
    Finally, it is important to ask the question, ``growth to what?'' 
Just as averages can mask under-achievement by some groups of students, 
so too can standards that are not sufficiently rigorous. If schools can 
meet their goals not only based on students that are meeting standards, 
but also on growth toward these standards, it becomes even more 
important to have meaningful, high-level standards. As Congress 
considers allowing states to incorporate growth into accountability, it 
is important to revisit the hands-off approach that has ignored the 
rigor of state standards.
Accountability Alone Is Not Adequate
    Improving NCLB's accountability system is an important undertaking, 
and the inclusion of a growth component guided by these principles can 
help.
    But we need to keep all this in perspective. Even the best 
accountability system is essentially a signaling system, helping 
educators, policymakers, and the public-at-large to understand whether 
schools are meeting their goals.
    I'm the last person to dismiss the power of accountability systems 
to improve coherence and focus in systems of public education that 
often lack both. Certainly, there is evidence that standards and 
accountability help: This year's edition of Education Week's Quality 
Counts report concluded that states with stronger implementation of 
accountability have seen appreciably bigger gains on NAEP, bolstering 
earlier findings from a study by the RAND Corporation. And NCLB's focus 
on accountability for different groups of students has provided 
critically important leverage to get systems of public education 
responding to the needs of low-income and minority students.
    What's really important, however, is improving teaching and 
learning. No amount of tweaking the accountability system will solve 
the serious, systemic problems plaguing our public schools. Indeed, I 
worry that, unlike in other countries, where leaders adopt standards 
and accountability systems and then move quickly to improvement 
activities, we Americans keep focusing the bulk of our energy on making 
accountability systems ever better.
    There are some critical improvement issues that demand their own 
attention:
    Capacity to turn around struggling schools: As a country, we've 
made a policy decision to no longer tolerate widespread failure in 
public education. This marks a historic shift, and one that was long 
overdue. But the old state and district systems were built when low-
performing schools were considered acceptable and even inevitable, and 
weren't set up to diagnose problems and intervene in struggling 
schools.
    We need to rapidly expand the expertise and the resources focused 
on turning around persistently low-performing schools. The current 
budget proposal to dedicate $200 million to the school improvement fund 
is a good idea, in part because it sends a signal to the states that 
this is a priority. But right now states aren't investing enough of 
their own resources in this area, and demand far outpaces ability to 
respond.
    Teacher Quality: We cannot close achievement gaps without closing 
gaps in access to teacher quality. Recent research from Illinois 
documented that students who studied all the way through Calculus in 
schools with the lowest teacher quality learned less math than students 
who only went through Algebra 2 in schools with just average teacher 
quality. Yet Congress has continued to pour billions of dollars into 
systems, ostensibly to help educate poor kids, only to have systems 
provide these students with the least access to qualified teachers and 
high-quality teaching.
    Recently, for the first time in the four-year history of NCLB's 
implementation, the U.S. Department of Education required states to 
develop equity plans to ensure poor and minority students get their 
fair share of teacher talent. And Congress has encouraged innovation in 
teacher assignment, evaluation, and compensation by creating the 
Teacher Incentive Fund. These are important steps, but raising teacher 
quality and ensuring equal access to effective teachers must remain a 
bipartisan priority.
    Title 2 of NCLB is by far the biggest investment in raising teacher 
quality, especially in school districts with high proportions of low-
income students. With Title 2, Congress gave districts virtually 
unfettered discretion to use the money as they saw fit. The result has 
been general programs that diluted the targeted support envisioned by 
Congress, with no discernable impact on teacher quality distribution 
despite a $3 billion annual investment. Congress should conduct an 
intense inquiry into how Title 2 is being implemented and stipulate 
better targeting, but should not cut $300 million from the program, as 
is currently proposed.
    Curriculum development with aligned benchmark assessments: In the 
early days of standards-based reform, leaders thought that standards 
themselves would provide sufficient guidance to teachers about what to 
teach and to what level. They believed that teachers themselves would 
figure out how to get students to those standards.
    It turns out that most teachers neither want to develop their own 
curriculum nor have the skills to do it. Rather, they need coherent, 
well-designed lessons, units and assignments that they can use day to 
day. Fortunately, states and districts are doing more and more along 
these lines, in part to try to reach their accountability goals.
    They are also doing more with the kinds of regular, teacher-
friendly, ``benchmark'' assessments that teachers need to gauge their 
students' progress toward state standards.
    While states and districts have increased their activity and 
funding in this arena, the federal government could play an important 
role in bringing these practices to scale by targeting grants in this 
area to jurisdictions that have good data systems and want to submit 
these activities to rigorous evaluation, so that lessons can be learned 
and widely disseminated.
Conclusion
    I wish I could report to you that the culture of accountability and 
continuous improvement had permeated public education, so that you 
could hand back more discretion to the states to set accountability on 
their own terms. But nothing in our history or the current climate 
suggests that we have made enough progress on this front. In fact, some 
of the push for growth models is a ruse to distract attention from the 
stark reality that many of our schools themselves must grow a lot, and 
fast. Several of the states that clamored the loudest for growth models 
did not even apply for the pilot because they do not have the requisite 
assessments or data systems in place. For some of the most outspoken 
critics, the focus on growth amounts to little more than an attempt to 
diminish public support for meaningful accountability.
    Incorporating growth is a good idea and Congress should not be 
deterred because some of its boosters have mixed motives. But it does 
mean that you must be vigilant in scrutinizing proposals to ensure that 
core principles are preserved and strengthened.
    Strong accountability is the most important leverage we have to 
focus public education on continuous improvement and the quest for 
equal educational opportunity. The consequences of weakening 
accountability will reverberate in the nation's military preparedness, 
economic vitality, and social cohesion.
    Basing school accountability determinations on measures of 
individual students' growth over time can improve accountability, and 
those improvements can help ensure that public education targets its 
resources to the students who need the most help. I do not in any way 
want to diminish the importance of getting accountability systems to be 
as good as they can be.
    But we have got to get beyond this never-ending quest for the 
perfect accountability system and turn to the hard work of curriculum 
development, teacher professional development, and leadership training 
for principals.
                                 ______
                                 
    Chairman McKeon. Thank you.
    Dr. Sanders?

 STATEMENT OF WILLIAM L. SANDERS, PH.D., SENIOR MANAGER, VALUE-
       ADDED ASSESSMENT AND RESEARCH, SAS INSTITUTE, INC.

    Mr. Sanders. Thank you, Mr. Chairman, Mr. Miller, members 
of the committee. I am honored by your invitation.
    My letter of invitation asked me to comment on my 
experiences with value-added and growth models, how growth or 
value-added models might fit into state-developed 
accountability systems under No Child Left Behind, benefits and 
challenges of implementation, as well as some of our major 
research findings that we have accumulated over the past 24 
years.
    In my remarks I will refer to each of these requests. 
However, the total intent of my remarks will be to make the 
case to Congress that the addition of a properly constructed 
growth component to the adequate yearly progress measure will 
make NCLB fairer to schools and will provide positive benefits 
to a greater percentage of their student populations.
    My experiences--I am a statistician that fortuitously got 
involved with educational research 24 years ago. At that time 
in Tennessee, there was considerable discussion how to improve 
the effectiveness of public schooling for that state's 
populations.
    These discussions inevitably led to the question of how to 
quantitatively measure the impact of schooling on measures of 
academic performance, especially academic growth.
    In fact--and I think this is important for the committee to 
recognize--in fact, some of the same quantitative issues that 
are now being raised relative to the growth model discussions 
were indeed being raised in that era.
    After learning of these issues, and my personal knowledge 
of statistical mixed model theory and methodology, I felt then, 
and I certainly believe even more deeply today, that there 
exist solutions to many of the pertinent questions that are 
often cited as impediments to using student test data to 
provide quantitative, reliable, robust measures of schooling 
influences.
    Our work over the past 24 years has certainly been about 
refining the statistical methodology for application to the 
educational measurement arena.
    Now, next, Mr. Chairman, what I would like to do is to try 
to make a distinction between value-added models and growth 
models in the context of No Child Left Behind.
    Value-added models and growth models--and I often refer to 
them as projection models--all use longitudinal data, as 
Chancellor Klein has mentioned--in other words, following the 
progress of each student as an individual.
    In value-added models like the Tennessee value-added 
assessment system, the purpose is to essentially get a measure 
of the impact, a summative measure of the impact that that 
school is having on the rate of progress of all students.
    In the context of No Child Left Behind and in the context 
of what I would certainly recommend Congress consider, the 
projection models use longitudinal data but for a different 
purpose. The purpose is to essentially take the data for an 
individual child and develop a projection of whether or not 
that child is on a trajectory to meet a meaningful academic 
standard in the future.
    Let me give you a specific example. In Memphis, there is a 
school, and I could certainly point to many other schools, 
where the average entering 3rd-grader profiles at about the 
25th percentile point relative to all 3rd-graders in the state 
of Tennessee--very low-achieving kids entering that school.
    Yet by the end of 5th grade in that school, the average kid 
is scoring at the 45th percentile point relative to all kids in 
that school. That school in the past has failed AYP because 
under existing rules you have got to aggregate the 3rd-grade 
scores, the 4th-grade scores, the 5th-grade scores.
    In my view, that is not a failing school. That is an 
excellent, outstanding, effective school. So to me, the 
addition of the growth measure could enable schools like that 
to be recognized proudly for their effectiveness.
    Now, even with the additional augmentation of AYP, as two 
states have, as was mentioned, now been approved in a pilot 
project by the Department of Education, there is another 
consideration that Congress should give prior to 
reauthorization of No Child Left Behind. And basically, I will 
be restating what Chancellor Klein said.
    There are a lot of schools, and I certainly could show you 
many, that have presently passed AYP, presently passed adequate 
yearly progress, yet when you look in those schools you will 
often see that those students have a projection whereby that 
they will not be proficient in the future.
    And this notion that unwittingly we put this perverse 
incentive in too many places to teach to the bubble kids, the 
kids just below proficiency such that if we get them over the 
bar, then we will tend to raise our percent proficient.
    And often, this is at the exclusion of those children that 
are so far behind we are not going to get them over the bar 
this year, and also at the exclusion of those children that are 
already above the bar because they are going to be proficient 
anyway.
    This is the kind of thing that if there is a projection 
component included in the reauthorization of No Child Left 
Behind, I think some of that disincentive will be removed.
    Now, NCLB, in my view, is beginning to yield many positive 
results and raising the nation's academic achievement for a 
large segment of the student population. The suggested tweaking 
with addition of a projection component will be an improvement.
    And certainly our research accumulated over the past 24 
years certainly has documented that effective schooling 
sustained over time will trump socioeconomic influences.
    Thank you, Mr. Chairman.
    [The prepared statement of Mr. Sanders follows:]

Prepared Statement of William L. Sanders, Ph.D., Senior Manager, Value-
           Added Assessment and Research, SAS Institute, Inc.

    Thank you, Mr. Chairman. My name is William L. Sanders; I am 
presently Senior Manager, Value-added Research and Assessment, SAS 
Institute, Inc. Additionally, I hold the honorary title of Senior 
Research Fellow with the University of North Carolina. Previously, I 
was Professor and Director of the Value-Added Research and Assessment 
Center with the University of Tennessee.
    My letter of invitation asked me to comment on: my experience with 
value-added and growth models, how growth or value-added models might 
fit into State developed accountability systems under No Child Left 
Behind (NCLB), benefits and challenges of implementation, as well as 
some of our major research findings. In my remarks I will refer to each 
of these requests. However, the total intent of my remarks will be to 
make the case to Congress that the addition of a properly constructed 
growth component to the adequate yearly progress measure (AYP) will 
make NCLB fairer to schools and will provide positive benefits to a 
greater percentage of their student populations.
    My experiences. I am a statistician that fortuitously got involved 
with educational research 24 years ago. At that time in Tennessee, 
under the leadership of Governor Lamar Alexander, there was 
considerable discussion on how to improve the effectiveness of public 
schooling for that state's population of students. These discussions 
inevitably lead to the question of how to quantitatively measure the 
impact of schooling on measures of academic performance, especially 
measures of academic growth attributable to various schooling entities. 
In fact, some of the same quantitative issues that are now being raised 
relative to the growth model discussions were indeed being raised in 
that era. After learning of these issues, and being knowledgeable of 
statistical mixed model theory and methodology, I felt that there 
existed solutions to many of the pertinent questions being cited as 
impediments to using student test data to provide quantitative, 
reliable robust measures of schooling influences on the rate of 
academic progress of student populations.
    Using this methodology, my colleagues and I built the quantitative 
system on which the Tennessee Value-Added Assessment System is based. 
Perhaps I was not the first, but one of the first to apply the term 
``value added assessment'' to measurement of educational outcome. 
Value-added assessment provides measures of the influence that 
educational entities, (i.e. districts, schools and classrooms) have on 
the rate of student academic progress. All value-added procedures use 
longitudinal data (i.e. follow the progress of individual students over 
grades) to get measures of these influences. These measures provide 
information as to the effectiveness schools or districts in providing 
the opportunity for academic progress for all students. Tennessee has 
had value-added measures as part of its accountability system since 
1993.
    However, all value-added modeling efforts do not give equivalent 
results. Some of the more simplistic value-added approaches should be 
rejected because of serious biases and/or unreliable estimates that 
they provide. If value-added models are to be used as part of an 
accountability system, then there are some minimal criteria that must 
be required. To dampen the error of measurement associated with a 
single test score for an individual student, all test data over grades 
and subjects for each individual student must be used in the analysis. 
However, all students do not have the same quantity of test data. 
Disproportionately low scoring students have more missing longitudinal 
data than higher scoring students. Thus, any value-added model approach 
must be sophisticated enough to provide unbiased, reliable measures 
using all data for each student no matter how sparse or complete. 
Simple posttest minus pretest averages and simple regression 
approaches, which use only the previous year's score as a predictor 
variable, are examples of value-added attempts that should not be used.
    Next, I would like to make a distinction between the use of value-
added models (like the Tennessee Value-Added Assessment System) and 
growth, or projection models to be used as part of NCLB. In 
accountability systems, value-added models use longitudinal data to 
provide a summative measure of the aggregate progress of all students 
attending a school. The projection (growth) model recently approved for 
Tennessee to augment AYP uses longitudinal data to ascertain if a 
student is on a trajectory to reach a proficiency standard three years 
in the future. The same data structure is used for two different 
purposes. However, the same statistical issues (fractured student 
records, using all data for each student, etc.) are just as important 
and must be accommodated in the projection (growth) models as with the 
value-added models. Again some of the more simplistic approaches to 
measurement of growth should not be used because of the resulting 
innate biases and greater unreliability. On this topic, I concur with 
the U.S. Department of Education's peer review team's comment that all 
of each student's prior data should be used; not just two data points.
    Why should NCLB be augmented to allow projection (growth) models? 
Students enter a school with a wide range of achievement levels. Under 
existing rules if a school has an entering population whose achievement 
level is extremely low, then regardless of the magnitude of progress of 
these students, it is most difficult for the school to make its AYP 
targets. For those schools which are eliciting superior academic growth 
for its student population, this additional measure can clearly 
differentiate these schools and enable them to be recognized proudly 
for their effectiveness.
    Even with the additional augmentation of AYP, as two states have 
now been approved in a pilot project by the Department of Education, 
there is another consideration that Congress should give prior to 
reauthorization of NCLB. With the existing AYP rules, schools can now 
be meeting their AYP targets, yet within those schools the progress 
rates of students who are currently proficient can be so relatively 
modest that their likelihood of not meeting proficiency in the future 
is greatly enhanced. Unfortunately, it appears that in too many 
schools, which are in jeopardy of not meeting their AYP targets, more 
instructional effort is focused on the ``bubble kids'', (i.e. those 
kids who are perceived to be near proficiency) with less effort 
extended for other students. The mistaken belief is that some students 
are so far behind that regardless of effort they will not reach 
proficiency this year, while other students are going to meet the 
proficiency requirements without much curricular attention. The focus 
on ``bubble'' students leaves two groups of students vulnerable, those 
most behind academically and those barely proficient.
    To provide a disincentive for this practice, I would recommend that 
Congress give serious consideration to replacing the existing ``safe 
harbor'' component of the AYP rules with a projection component so that 
schools will receive credit for keeping all students on trajectories to 
be proficient in the future. This should tend to alleviate the problem 
of not focusing appropriate instructional effort on all students. All 
states will now have the annual testing in place to allow for the 
projection approach to be applied. At a minimum, states should be 
allowed to substitute this better approach for the existing ``safe 
harbor'' rules.
    In summary, the impact of NCLB is beginning to yield many positive 
results in raising the academic achievement level for a large segment 
of the nation's student population. The suggested tweaking with the 
addition of a projection (growth) component will be an improvement. Our 
research accumulated on the past 24 years, certainly has documented 
that effective schooling will trump socio-economic influences if 
effective schooling is sustained over time for each student. The data 
resulting from the implementation of NCLB, if wisely used, can 
certainly lay the information base for insuring that all students will 
have the opportunity to learn, consistent with their achievement level.
                                 ______
                                 
    Chairman McKeon. Well, thank you very much.
    This has been very interesting. I was listening to Mr. 
Miller's opening comments and my opening comments, and there is 
not a lot of difference.
    I think one thing that is exciting to me as we move 
forward--we have had now--this is the fourth hearing on getting 
ready for the reauthorization. Each one of them has been good. 
This one has been excellent.
    I have no agenda. It doesn't sound like Mr. Miller has any 
agenda, other than improving on what we did when we wrote the 
No Child Left Behind. We are not trying to say this is 
something we want to emphasize.
    We want to listen to experts like you find out what can be 
done to make it better. And as we go through this process, we 
are going to be listening very carefully. That sounds like the 
ultimate goal, and I am hopeful that we are going to be able to 
continue to work in a bipartisan way to really make this 
happen, because it is too important to not do that.
    I forget now which one of you mentioned--it was probably 
Chancellor Klein--about ``gotcha.'' You know, I have some 
concerns about government at all levels--or not just 
government, but people that have that attitude, of trying to 
find people doing something wrong rather than trying to 
encourage people to do what is right.
    And I hope that we don't have that attitude, that all we 
want to do is find something wrong. And I think that is a poor 
management model. Better to find what people are doing right 
and focus on that. And government's role should be how can we 
help you do better, not can we find what you are doing wrong.
    And I think several of you mentioned the importance of 
principals, assistance principals, teachers, parents, because 
we are all in this together. There has been a lot of focus on 
teachers, but I know I go to school sometimes and I can see--
just as you walk on the campus, you can feel the influence of a 
great principal.
    You can feel the influence of a great superintendent. You 
can feel the influence of an assistant principal. You go into 
the classroom, you can feel the influence of a great teacher. 
You can also feel the influence of a teacher that is not 
measuring up.
    And I am hopeful that from your testimony today--and I 
think this is important. We have to do this. But I would rather 
get away from this kind of a setting and sit around in a circle 
and really have some give and take and some real meaningful 
discussion about--you are all experts.
    Really, you can teach us a lot. And maybe we can also do 
some of those kind of settings as we move forward, just have 
some real round-table discussions and go into some of this 
stuff in depth.
    One thing, though. I think one of the problems we have had 
with No Child Left Behind is people have put, you know, 
different labels on things, and sometimes we have gotten blamed 
for things that are part of state government. We say that 
states have standards, and then we get blamed for the standards 
the states set.
    Dr. Weaver, in the conclusion of your testimony you told 
the story about this young student, and I think that grabbed 
all of us. He was devastated when he didn't pass an exam and he 
was labeled an unsatisfactory learner.
    What responsibility do you think the teachers have for 
ensuring that students really understand the role of state-
developed assessments and the fact that NCLB judges the 
progress of schools, not individual students?
    And do teachers have a role in ensuring the public 
understands that nothing in Federal law labels schools as 
failing?
    Mr. Weaver. I think that if we understand the rules, if we 
understand the law, if we understand what it is we are supposed 
to do, then I think that we can be good messengers.
    However, there have been so many mixed messages. There have 
been many stops and many starts. And so it has left a number of 
people, not only teachers--but it has left a number of people 
uncertain as to what it is that we are supposed to do.
    Also, when it comes to the involvement of people, teachers, 
regardless of what position, involvement of those individuals 
in implementation, it makes it much easier if, in fact, we have 
been involved or included in terms of determining what that 
outcome is going to be.
    However, in many instances, we have had decisions made for 
us, without us, in spite of us, without even talking with us. 
And then the expectation is that we implement it. That is not a 
good way to encourage good implementation. I don't think it 
works at our level. I don't think it works at your level. I 
don't think it works at any level.
    So the more that we are----
    Chairman McKeon. Or any business.
    Mr. Weaver. Absolutely. The more that we are involved in 
helping to shape the circumstances that impact us, I think the 
better off we all will be.
    Chairman McKeon. Thank you. I have a whole lot of other 
stuff I would like to talk about, but my time is up.
    Mr. Miller?
    Mr. Miller. Thank you very much.
    Thank you for participating this morning. We look forward 
to spending a lot of time with you all as we move forward in 
this process. I want to make sort of a comment and then maybe 
get your reaction, because I won't have time for individual 
questions this morning.
    But back in the days of yesterday when I was a kid, I 
brought home a report card at the end of my 4th-grade year, and 
it said either 4th-grade reading, 4th-grade math, 4.5, 3.5, 
whatever, so my parents said, OK, you are doing all right.
    And if I brought home the next one and it said--at the end 
of my 5th grade and it said 4.5, my parents said, you are 
falling behind here, you are going to go to summer school, you 
are going to do something like that.
    I won't even go to the deportment questions, but there was 
an implied bargain. You put your kid in the school, and you get 
a year's growth for a year's schooling, that they would be at 
grade level. That was sort of the idea.
    That is where you were moving, and that is--I think when we 
talk about proficiency, we all understood that we were talking 
about grade level.
    And, Mr. Klein, you talked about, you know, the mass of 
investments you are making in data, and one of the things as we 
kind of look around the country--districts that have decided 
that they are going to--and states that decide they are going 
to embrace the data seem--and, Kati, you referenced this in 
your discussion--and try to figure out accountability and use 
this data to get a road map seem to be doing somewhat better.
    But you have to--I think in this version of trying to close 
the gap, you have got to embrace the data. You have got to know 
something about your students. You have to know on a real-time 
basis what is happening in that classroom.
    And we didn't invent high-stakes testing. We didn't do any 
of that. We took the stakes as we found them. In fact, the 
president and I, one of the few times we have ever been in 
unison, talked about diagnostic data. We really wanted real-
time interventions, hopefully small interventions that could 
make a large difference for students.
    But as we now start to look at the data, another thing is 
happening. Some of the urban myths and legends are falling 
apart: But for those English learners, we would have made AYP. 
The data from the Aspen Institute starts to suggest that is not 
the case.
    If we had only had a growth model, we would have made AYP. 
Your data suggested in some cases that is not the case.
    And I think what we need to understand here is that in 
achieving the growth and the proficiency that we want, there is 
a lot of work that needs to be done. It is not just changing 
the standard.
    And I think you sort of say that, Kati, in your testimony.
    You know, Dr. Sanders, you are kind of a legend, you know, 
to me, in the sense that how you took this problem apart of 
value-added.
    But it starts to suggest that if we really make the 
investment in our teacher corps, if we really, you know, 
understand the need for that core investment, we can, in fact, 
get these years of growth in succession and not have the 
student fall back.
    So the data becomes--it is troublesome, because it 
describes the magnitude of what we have to do and the resources 
that are going to have to be necessary. Some states don't have 
a clue what is going on inside that classroom. Therefore, they 
can't make changes that help those students.
    And I just--I don't know if there is a comment there for 
you to make, but I am concerned that we think if we just tinker 
with this, all will be right, because I have gone into I don't 
know how many school rooms in how many schools all across this 
nation, and I have always been told but for those English 
learners we could have--just two students missed the mark, we 
were going to be a great school.
    Now, I don't know if AYP would have told me that was a 
great school or not, because we understand that it is not a 
fair measurement in some instances. But that is not the 
problem, I don't think.
    It is what we have--we have to hold that as a standard, but 
getting to those high standards, getting to those aligned 
curriculums, getting to tests that are useful and impart 
information on a timely basis is also a big part of the 
problem, and I don't--you know, goes to some question of 
resources.
    That is my comment here, because--if you want to comment, I 
would certainly appreciate it. If you can't decipher it, send 
me a note 10 days from now. But I just want to--I want us to 
understand the magnitude of the problem we are engaging here.
    I mean, there has been some alignment, and I think we have 
got a lot of people headed in the right direction over the last 
5 years, but there is a big unfinished part of this painting 
here.
    Mr. Klein. Mr. Miller, I would like to comment. I will also 
send you a letter, but I would like to comment.
    I think what you say is absolutely true, and I don't think 
anybody on the panel would disagree, that if you--you can have 
a perfect accountability system, but if you don't have the 
proper investments in terms of your human resources, your 
talent and so forth, you are not going to get the job done.
    However, it seems to me accountability, which is a core 
thing, and trying to change the culture of a school system in 
which you are saying to your folks, which I am saying all the 
time, we are responsible for the educate of our kids--the days 
of excuses are over--it is the kids, it is poverty, it is the 
fact that they are immigrants. I don't want to hear that.
    What I want to hear is how we are going to educate each and 
every one of our kids. And accountability is critical to that 
equation. And so what I am trying to say to you today is the 
message that I try to send to my school system is this is our 
collective responsibility in the city of New York.
    For far too long, far too many of our kids have not gotten 
the education that they need to succeed. But I need to be able 
to convince those people that the measurements we use are 
really very powerful and fair.
    And I agree with you. The fact that there is a growth model 
is going to lead to schools that don't pass the average yearly 
progress. There is no question about it.
    And I think Bill Sanders' point is a very important one. 
Some of those schools are going to be schools that people now 
call good schools, because they are not moving their kids. And 
what you want to see is really almost a kind of kinetic motion 
in the system of all the kids moving, not a strategic motion.
    I walked into a school not far from Major Owens' district 
and the principal said--I said you had such great results, you 
went up 20 points last year, how did you do it.
    You know what he did? He took out a little chart, and he 
said here are all my kids. I looked at all the ones that were 
closest to level three and just above level three.
    The ones above level three I wanted to make sure they don't 
fall below level three, so I focused on them. The ones below 
level three, I boosted them up level three, and we raised it 
then by 20 points. I said what happened to the other kids. He 
said they probably did OK.
    And I think it is very important that we don't breed 
cynicism about accountability, and that is why I think what Dr. 
Sanders has done has really enabled us to move it to a 
different level.
    I agree with Kati. In the end, the solution is not to 
simply grow to no end. It is to grow to proficiency, but I 
would say even beyond proficiency. Proficiency is a minimalist 
standard.
    Our nation isn't going to succeed if our kids are not 
proficient--many of them--at an entirely higher level. So I 
think what you say is right, but I wouldn't discount the power 
of what we measure in terms of the behaviors in the school 
system like mine.
    Mr. Miller. I was just going to say I don't want to be 
unfair to my colleagues who are waiting here for questions, so 
I will take my answers on the air later, because I am afraid I 
am going to run over time, and you had better recognize that.
    But I am sorry, Reg. We will talk about this. I don't know 
what to do. I took too much time with the question, is the 
problem.
    Mrs. Biggert [presiding]. Thank you.
    The gentleman from Delaware, Mr. Castle, is recognized.
    Mr. Castle. Thank you very much.
    I think this is an extremely important hearing, which I 
have been pushing for, and it is an extremely important but 
very difficult subject. I agree with chairman's suggestion that 
we have round table discussions at some point.
    I don't know if a hearing, 5 minutes of testimony and a few 
questions, lends itself to what we need. Plus, this happens to 
be an extraordinarily busy morning, which is unfortunate, too.
    You know, this is an aside from everything, but in dealing 
with education over many years now, I just have learned that 
you need sort of a systemic component--and Dr. Shaul is 
retiring, and I say this to her now, but you need a systemic 
component to all this.
    I see so many charismatic type programs and individuals who 
do things in education, but it doesn't translate to other 
people being able to do it.
    I think any time we have a report, any time we have a 
suggestion for something, you need to have some sort of a 
spelling out of how it could be done systemically and not just 
by those individuals who sort of embrace it, or it is what they 
do, or how they get their foundations of whatever the heck it 
may be. Some of these things just fall when you get into that.
    I also worry about what Ms. Haycock said, that we worry too 
much about accountability definition and not improvement. I am 
just seeing way too much of that. I am from Delaware, and I 
have seen just incredible improvement in these scores.
    And I have been in those schools, and I have seen what they 
are doing, and they are really breaking it down and really 
analyzing it, and that is systemic. That is something I think 
would translate to other schools.
    But I have been in other schools where everybody is 
throwing their hands up and saying oh, we can't do this, we 
need something else, or whatever it may be, No Child Left 
Behind is no good, and that is just not teachers. That is 
administrators, even some parents in some cases, even state 
officials in some cases, as we have seen with lawsuits or 
whatever.
    So that concerns me a lot, too. We need to really focus on 
getting this job done, and I think it can be done.
    I also worry about the different standards and assessments, 
and I realize that all 50 states have gone through the 
Department of Education, but I still think there is a 
tremendous variance between standards and assessments.
    And I am not about to suggest national standards and 
assessments, but it goes through my mind every now and then. It 
would sure straighten out some of those problems that exist out 
there. You know, maybe it is worthy of discussion at some 
point.
    But I think this is important. I think this is the most 
important aspect of No Child Left Behind, what we are talking 
about today, in terms of what may change in the next 
reiteration of this particular legislation.
    And I am not suggesting I understand it, and I do 
appreciate the study, which is just being issued now, and which 
I haven't had a chance to really review myself, so I have a 
long ways to go, too.
    But my question, if anyone wants to take a stab at it 
before the red light goes on, is what are we talking about here 
in terms of the--is this an either/or circumstance, or is it a 
blended circumstance?
    I think, Chancellor Klein, you just made a statement--I 
think I got it correct--that the growth model will lead to 
schools that don't make AYP. Well, that would mean to me that 
it is not either/or. That would lead me to believe that it is 
some sort of a blended, you know, situation.
    And perhaps it should be. Why shouldn't a school that is 
doing very well also be able to show growth as well as the 
schools doing poorly, which, by showing growth, can get to AYP?
    I am not sure what the answer to that is, but I am 
interested in any quick comments you might have about the--you 
know, whether we are dealing with an either/or or blend 
situation with respect to growth models or the pure testing.
    Ms. Haycock. Well, I can tell you, Mr. Castle, that the two 
states that got approval for a growth model are, in fact, 
blended systems. About each school there are essentially three 
questions asked.
    First, do they make AYP under the straight status model. If 
the answer to that is yes, then they make AYP. If the answer is 
no, though, then the question is did you make it under the safe 
harbor or sort of improvement model. If the answer to that is 
yes, they make AYP. If the answer to that is no, then the 
question is do they make it under a growth model.
    So in both cases it is essentially a blended system. You 
essentially have several options in reauthorization. You could 
ask for blended systems, you could allow states to go with a 
growth only system, or any combination thereof.
    But the two experiments at least that you have under way 
now are blended models.
    Mr. Castle. Dr. Sanders?
    Mr. Sanders. I would just make one additional comment on 
that. Based upon all of the analysis that we have done, I would 
certainly recommend a blended system.
    I would allow, just as Kati has pointed out, that--schools 
to apply the existing AYP rules. What I would do is substitute 
the projection growth model, if you would, component for the 
safe harbor part that is present.
    In other words, instead of going the way it is right now, 
even for the two states that were approved, this was 
essentially a pilot to augment what existed, so you remember 
what Chancellor Klein mentioned and what I attempted to 
mention.
    One of the things that concerns me the most is particularly 
schools serving a disproportionate number of poor and minority 
kids. That is when those early above-average kids too often are 
allowed to slide.
    That is, to me, one of the biggest negatives associated 
with the present system. And that has not been that--it is not 
recognized under the existing system. And that is the sliding 
of above-average kids.
    So if you substituted a growth component for the existing 
safe harbor, that would be my preference.
    Mr. Castle. But just in closing, I hope when it comes time 
to write all this that you all will be around to help us with 
it, because it sounds to me like it is going to be complicated.
    I yield back the balance of my time.
    Mrs. Biggert. The gentleman yields back.
    The gentlelady from California, Ms. Woolsey, is recognized.
    Ms. Woolsey. Thank you, Madam Chairwoman.
    First of all, I agree with Chancellor Klein. No Child Left 
Behind is extremely valuable and important to us, but it does 
need to be improved.
    What I say is it needs to be fixed, and it can be fixed 
without giving up accountability, improving teaching and 
learning, and without giving up on high-quality schools and 
their teachers.
    So I am not going to say any more than that. I am going to 
ask two questions. I have done quite a bit of research and 
discussion with my local educators and with national education 
groups over the last year, and two major questions that 
continue to come up fit right in with what we are talking about 
today.
    One, on growth. The question of can we educate our most 
challenged students--English learners, those that are 
economically disadvantaged--within a growth model that, over 
time, will not leave them behind.
    And the second one is--my second question is accountability 
based solely on test scores is not very popular in my area. 
What would you recommend? What would you recommend that would 
include teaching the whole child?
    Can we provide growth models that include, yes, math and 
reading, of course, but also P.E., history, art, and music, and 
all of the other things that kids need to be whole people when 
they are finished with school?
    So I am going to leave it there and answer as you wish. 
Let's start with Dr. Sanders. He looks like he is ready.
    Mr. Sanders. I am sorry, I couldn't hear the last point.
    Ms. Woolsey. Pardon?
    Mr. Sanders. I am sorry, I missed your last comment. I am 
sorry.
    Ms. Woolsey. Well, the last comment was a growth model that 
includes educating the whole child, meaning in addition to math 
and reading.
    Mr. Sanders. Well, first of all, let me refer to some 
comments that Dr. Weaver made earlier. To me, one of the really 
huge advantages of states creating the longitudinal data 
structure--the accountability part is important, but what is 
even more important is the diagnostic information that is now 
available.
    And when you begin to start looking at projections for 
every single kid as an individual to various standards in the 
future--if we have got a child that is very low achieving, the 
first standard is to get that kid--the first objective is to 
get that kid on a trajectory to meet the proficiency standards 
1 year or 2 years in the future.
    But once we have got a kid like that, then let's talk about 
getting that kid on a trajectory to meet high school graduation 
requirements. Once we have got kids on that trajectory, let's 
talk about getting those kids on trajectories to have more 
opportunities for more college majors.
    So in other words, once you begin to create the 
longitudinal data structure, you have got the opportunity to 
begin to start focusing and planning for the needs for a 
diversity of students.
    Now, we worked very closely with the Tennessee Department 
of Education in preparing their proposal that was--and we are 
very proud of that, that was approved to be one of the two 
states.
    Now, back to your thing of all children and all subgroups, 
in that proposal, all subgroups with regard to their 
projections have got to meet it or they don't get credit from 
this alternative approach--ESL kids, et cetera, et cetera, et 
cetera.
    Now, I will make one comment on ESL kids. We have had the 
opportunity to look at a lot of data, a lot of years.
    If you are a principal or a classroom teacher, looking at 
it from a value-added or growth project, you often want those 
kids in your classroom because as they accumulate more language 
skills, often their gain rates is higher, not lower. So 
consequently, the data strongly suggest that that is not a 
problem.
    Now, for kids that are significantly learning-impaired, 
what we believe and was included in that proposal--those 
students take an alternative assessment. You include that data 
in the overall computations, but you do that relative to the 
appropriate assessment for their needs.
    So consequently that has got to be incorporated in total. 
But it can all be done within the overriding spirit of No Child 
Left Behind.
    Ms. Woolsey. Thank you.
    Dr. Weaver?
    Mr. Weaver. I cannot tell you how pleased I am to be a part 
of this discussion. I cannot tell you how pleased I am.
    But you know what I want to do? I want to speak to you as a 
parent, and many parents with whom I have spoken with. And I 
also want to speak to you as a teacher representing many with 
whom I have spoken with.
    And I also want to speak to you as an African-American, 
many of whom I speak with consistently and whose children are 
not receiving what they consider to be the kind of education 
that they want.
    And when I talk with them and I tell them that I was a part 
of this panel, and we were talking about growth models, you 
know what they are going to say? They are going to look at me--
what is that? I could care less, because how is that going to 
impact what happens to my kid.
    I want my kid to be able to have the opportunity to have 
science and math and technology. I want my kid to be able to 
have a qualified and certified teacher. I want my kid to be 
able to go to a school that is safe and orderly.
    I want my kid to be able to have counselors. I want my kid 
to be able to have smaller class sizes, because I know that 
that is what it is going to take for my kid to be successful.
    And the growth model--I think it can lead to that, but the 
parents don't understand that. And when we go out and talk 
about well, we think that we need to have growth model added to 
No Child Left Behind, we understand that, but how do we get the 
parents to understand that?
    How do we get the average teacher, administrator ad other 
educator to understand that? So I am saying a lot of times we 
talk to ourselves in a code that the parents don't understand, 
the public doesn't understand, and as a result it is difficult 
to get their support.
    I go to these parents and I talk about--and teachers--well, 
I believe that no one test should be used to determine the 
future of a child. I believe that there needs to be multiple 
assessments, such as college entrance, such as retention--
graduation rates, such as portfolios.
    Those are the kinds of things that I do believe that we can 
begin to utilize in addition to having the test. A test should 
be a unit to measure, not the unit.
    And also, I believe that we need to look at how people view 
outputs and inputs. When we think about outputs, we are talking 
about a test score as if the test score is the only unit to 
determine whether or not a kid is successful.
    And we get focused on the output which is the test score, 
and we forget about the input, input such as class size, 
qualified and certified teachers, safe and orderly environment, 
parental involvement, because I believe that the input helps to 
determine the output.
    And so again, I am speaking for those individuals who are 
not here to speak for themselves.
    Mrs. Biggert. The gentlelady's time has expired.
    You might have noticed that there is rather a dearth of 
committee members over on this side of the aisle, and I know 
that they would really like to be here, but there was an 
important meeting called. I just wanted the witnesses to know 
that.
    I will recognize myself for 5 minutes.
    Dr. Shaul, in your GAO report you mention that 26 states 
were using growth models and another 22 states were either 
considering using them or in the process of implementing them, 
even though they are not under the--they are not part of the 
pilot but just on their own.
    Do you believe that all of these states are in a position 
to implement the growth models that comply with the goals of No 
Child Left Behind and the guidelines set by the Department of 
Education in their pilot program?
    Ms. Shaul. Not all the states are in a position to be able 
to implement the department's growth model pilot, which set out 
some very high standards, seven core principles that models 
needed to meet, that used individual student data.
    Of the states that we looked at, seven of the states had 
models that were based on individual student data. The rest 
were using school-level data. So they would not have been 
eligible.
    As you know, 20 states chose to apply. Only eight states 
were peer-reviewed as being close enough to actually having 
systems in place that would meet the department's goals.
    Mrs. Biggert. When you look at these growth models and we 
are looking at the proficiency and having to reach 100 percent 
by 2013--and in the growth models, if they don't go up, you 
know, at a--in other words, a line that goes up instead of 
going like this at the end, then it is going to have to go like 
this.
    Do you think that this will work?
    Ms. Shaul. I think that states can design growth models 
that set a trajectory that will allow all students to be 
proficient by 2014. But for some students, those trajectories 
will be pretty high if they are starting at lower levels.
    But I think it is possible. But then, of course, setting 
the accountability systems--it goes back to points other 
panelists were making. Then you would have to have tailored 
instruction, the other things in place that will allow those 
students to reach those high standards.
    Mrs. Biggert. Thank you.
    And I will yield back.
    Mr. Van Hollen from Maryland is recognized for 5 minutes.
    Mr. Van Hollen. Thank you. Thank you very much, Madam 
Chairman.
    And let me thank all of you for your testimony. As others 
have said, I think this is a very valuable hearing for all of 
us, and I think we are approaching this in a bipartisan spirit 
at this point as to what is best for the kids.
    And I think that all of you have made the case that the 
current system of AYP does create some of these perverse 
incentives in the program. You talked about the kids between--
you know, on the bubble between 2.9 and 3.1, and a lot of 
emphasis and focus is put on them to the exclusion of people 
who are way below and probably won't make it up to that 
standard, or people who are above that standard.
    Now, clearly, I think that we need to eliminate those kind 
of perverse incentives. And I think that we need to come up 
with something different. In doing that, we need to make sure 
we don't replace it with another standard that creates 
different perverse incentives.
    And Mr. Castle said he hopes you are all here while we do 
that. I do, too. But I also hope that that doesn't mean we come 
up with a Rube Goldberg-type machine with, you know, three 
different things that no one can understand and does also 
substitute a system for this one and creates other perverse 
incentives.
    And that is why the quality--and the quality of the data 
and the ability of these school systems to assemble the data 
necessary for the growth model is something that concerns me 
greatly.
    As you just said, Dr. Shaul, I mean, you looked at about 25 
states or whatever the number is that are currently using it 
somewhat--most for their own purposes, and whether they are in 
a position to even begin to implement a growth model.
    And it seems to me that it is clear that we are not there 
yet. And if we really want to move them there, we are going to 
have to put some resources and some Federal Government money 
and some guidelines and standards for how we move in this 
direction. And let me just ask for the starting point.
    And, Chancellor Klein, you mentioned the difference between 
spotlight and hammer. And whatever system we come up with in 
terms of what the standard is for measuring AYP, it seems to me 
collecting some of this data is important for the purposes of 
spotlight anyway.
    Let's put aside the consequences, which can create these 
perverse--would you all agree that it makes sense to put all 
these school systems in a position where, whether or not we go 
to a full growth model, that we have the data necessary to 
determine whether progress is being made?
    Would you all agree that we should at the very least do 
that?
    Mr. Klein. Absolutely. And it is not just the school 
system. It is the people who are teaching our children who need 
the information, because if you get the test score at the end 
of the year, almost invariably that student is in another class 
at another time.
    And what happens is you don't have the information to 
create the kind of positive feedback loop. Those schools that 
are using effective assessment strategies are able to intervene 
much more quickly on the ground at the time, and that is one of 
the key things.
    And I admit this is a huge technological and data 
challenge, but if you don't address this challenge, it is like 
that old Thorndike experiment at Yale. You can throw darts at a 
target forever. If you are blindfolded, you are not going to 
get any better at it.
    Now, we need the information to help the people who are 
teaching our kids do the work that they need to do to make the 
adjustments, because one thing that worries me enormously about 
assessing systems is you can always have different assessing 
systems, but in my school system when I got there, thousands 
and thousands of kids were in high school, unable to read.
    Now, let me tell you something. That is because the system 
was unprepared to be serious about what it is to educate a kid. 
And we just passed them along to inevitable failure.
    So to me, strong benchmarks along the way are one aspect--
and then rich data to the people teaching our kids, so that 
they can actually do the transformative work with the kids.
    Mr. Van Hollen. Right. At some point--and time is running 
out, and I have another question--but it seems to me we need to 
get some kind of idea about what resources would be required to 
bring these school systems up to a point where they can have 
the--I mean, because bad data is not--bad results.
    And perverse incentives to gather certain kinds of data is 
also a problem.
    Let me just as you, Dr. Sanders--what I am trying to--and I 
commend you for what you have done in Tennessee. As I 
understand the system, under the waiver that you have got, you 
can achieve AYP in one of three different ways, is that right?
    Mr. Sanders. That is correct.
    Mr. Van Hollen. And is that the kind of model you want to--
I guess my concern at making that national is that it would 
allow a school system to game which of the three it is going to 
go for.
    I am not saying that is happening in Tennessee, but just as 
a national model--in other words, we talked about the problems 
with the more static--the sort of status quo model, what we are 
using now.
    But as I understand it, in Tennessee, you could hit AYP 
using that model. And if you miss AYP using that model, you can 
use the other model. But as you very aptly pointed out, there 
are schools that--your slide and glide, or whatever, rule.
    I mean, there are schools that can make it under the 
current rules that would not be making it under the growth 
model. And if you allow sort of the school system to target 
one, two or three, you may get a situation where they go for 
the strategy to go for the one they are making now, but they 
would miss it under another.
    And it seems to me you don't want that kind of gamesmanship 
going on.
    Mr. Sanders. Working with the Tennessee Department of 
Education, for the proposal to be considered for the pilots, we 
perceived the guidelines to be to talk about an augmentation of 
the existing process.
    What I attempted to do in my previous remarks was basically 
say if I were doing this from scratch, I would leave the 
existing AYP in place but I would change the safe harbor part 
to essentially be the growth projection model.
    And then I want to very briefly say I didn't--it is in my 
written statements but not here. I want the committee to 
recognize that all of these growth models are not the same and 
do not yield the same rigor and the same properties.
    And so consequently some of these more simplistic 
approaches to measures of growth I think should be seriously 
frowned on, like, for example, merely subtracting last year's 
score from this year's score, et cetera, because those things 
are extremely unstable.
    And everybody in the room would agree that in schools 
serving high concentrations of low-scoring kids, you have got 
more missing scores. And so any of the--I don't want to spin 
off on a whole bunch of quantitative mumbo-jumbo here, but any 
of these growth models--and this is where I really appreciate 
what the peer team--you folks need to look very carefully at 
what the peer reviewers said about that.
    So if a state is proposing this, there should be some 
minimal quantitative standards placed. But your specific 
question, sir--I would blend it. I would leave exactly what you 
have got. Then if they want to go under safe harbor, I would 
definitely go to that.
    And then finally, the other comment--I would not hold all 
states to the same capacity as the minimum state. In other 
words, states that have invested and were further down the road 
with regard to capacity to do this, you should build in the 
legislation flexibility where those states could take advantage 
of it.
    In other words, don't wait till all states have this 
capacity. Allow the state--and actually create an incentive for 
the states, because as Dr. Weaver says, the big--actually, the 
byproducts could be greater than the product.
    The byproducts is this wealth of diagnostic information 
that is going to flow to principals and teachers and so forth 
relative to the progress and the trajectories of all these kids 
as individuals.
    Mr. Van Hollen. Thank you.
    Mrs. Biggert. The gentleman's time has expired.
    The gentlewoman from California, Ms. Davis, is recognized 
for 5 minutes.
    Mrs. Davis of California. Thank you. Thank you, Madam 
Chairwoman.
    And thank you to all of you for being here. I wanted to 
raise an issue that we haven't talked about this morning that I 
raised when we were discussing the end sizes a while back and 
how we structure or how we count children in specific 
subgroups.
    And that is the whole idea of AYP and whether, in fact, we 
have a number of students who aren't counted because they enter 
the school after the yearly progress counts have been 
designated. And I don't know across the country the extent of 
that problem.
    And we know that it is not that kids who move around from 
school to school aren't counted, but when it comes to AYP, in 
fact, they are not necessarily included in that data. So how do 
we get at that issue in the growth models? And how can we begin 
to think about that?
    You all have talked about the importance of, you know, 
intervening on the ground. It worries me that we have a number 
of students who, in fact, for a number of reasons do move 
around. Some school districts have made great accommodations so 
kids can move in and out within clusters fairly easily. But the 
reality is I think we are missing these kids.
    Dr. Sanders, you are smiling. Could you comment on that? 
And how do we use the growth models to support that so that we 
do have the data, we do understand what impact that has on 
them?
    Mr. Sanders. Let me attempt to answer your question, first 
of all, from a statistical theory methodological point of view. 
If I had the data to know what percentage of time this child 
was in that school, then with this longitudinal record you 
would certainly have a way to apportion a part of that kid's 
time according to the amount of time that the kid was in this 
school compared to another school, et cetera.
    We probably have constructed more longitudinal data 
structures to date than anybody. We presently have the data 
from at least one district from over 20 states. Presently, I 
have yet to see any data system that would have that kind of 
information that would enable one to go so practically, even 
though the theory is there.
    The analytical capacity is there. The data structures are 
not there to know that little Suzie was in School A for 38 days 
and in School B for 61 days, et cetera, et cetera.
    That is a concern. But it is something that I think, as 
data structures mature over time--there are kids falling 
through the cracks. There I don't think is any question about 
that, particularly kids that move around.
    Now, in our value-added modeling efforts, where we are 
doing the summative measures--this is in my written remarks--
you will see this is one of the very points that I am making.
    Any of these structures have got to be sophisticated--any 
of these accountability systems have got to be sophisticated 
enough to utilize the data for all kids, not just for the kids 
that are not moving around.
    So I am attempting to answer your question from a 
theoretical conceptual point of view, and then move to the 
practical reality under No Child Left Behind with regard to 
that.
    Right now, I have yet to see any of the data structures 
that will enable you to have those sorts of things that would 
enable you to do that. The percentage of the kids that are 
falling through the cracks I think perhaps is lower than some 
people would guess it to be, because we can do things like--we 
know how many kids in a large urban district took the test 
somewhere.
    Then we know how many kids that we had previous scores on 
somewhere. And then we can--in fact, that was one of the 
questions the peer review team had. How does those ratios of 
kids with prior data compare across their socioeconomic groups?
    There is a difference. Minority kids, ESL kids, will have a 
lower ratio, but it is not as big as some would have guessed it 
to be. I think Congressman Miller's comment earlier--sometimes 
the things that we have heard merely being horror stories, when 
you really go in and look at the data, are not as big as 
sometimes people think it is.
    Ms. Haycock. Ms. Davis, if I could just say so, I hope you 
will stick with this issue. While it may not be our topic here 
today, it is a very important question that you are raising.
    As you know, in an effort to be sort of fair to schools, 
your decision last go 'round was to take kids who were not in 
the school for a full academic year out of the equation, put 
them into the district, so somebody is responsible, but as you 
know, if it is not really the school, then that doesn't work 
very well.
    And yes, in fact, that may feel fairer to schools, but we 
have got to really ask the question, given the amount of 
mobility there is in many communities, like certainly yours, 
whether it is fair to kids.
    And we have got a number of places, like Kentucky, that 
actually had always had all kids in the accountability system 
and were essentially forced to take them out in order to become 
compliant with NCLB. And there is a lot of dissatisfaction in 
Kentucky with that right now.
    So coming back to that issue in reauthorization and ask the 
question what is fair to both schools and kids, and how do we 
really make sure that the mobile kids get the attention they 
need to advance, because there is too many mobile kids to shut 
out to the side.
    Mrs. Biggert. The gentlelady's time has----
    Mrs. Davis of California. Thank you. My time is up.
    Mrs. Biggert [continuing]. Expired.
    Mrs. Davis of California. Can I get Dr. Weaver to make a 
very brief comment? Would that acceptable?
    Mrs. Biggert. Well, no.
    Mrs. Davis of California. No? OK.
    Mrs. Biggert. I think that we have got so many members 
waiting. We will have votes at noon, so we would like to try 
and finish by then and make sure everybody can have an 
opportunity. Otherwise we could come back.
    But the gentleman from New York, Mr. Owens, is recognized.
    Mr. Owens. I want to begin by thanking Mr. Klein for his 
kind remarks.
    And I also want to salute my friend Reggie Weaver. He heads 
a collective bargaining unit that doesn't represent my city, 
but I regret that very much. He is very outspoken on the 
national macro issues in education as well as the micro issues 
that concern his members, and I salute him for that.
    This panel has been very informative, starting with 
Director Shaul. I have got a good idea how to easily clearly 
explain this to parents from your presentation.
    I also very much appreciate the remarks of Dr. Sanders. 
Standard school efforts--I mean sustained school efforts can 
trump socioeconomic influences. I think you said something like 
that.
    I think that links with something that Reggie Weaver said, 
qualified, diverse and stable staff are very important as part 
of this process of achieving the goals that we have set.
    You have been very disciplined, all of you, in keeping the 
discussion within the context of what I call a grand 
accountability hypocrisy that No Child Left Behind forces us 
all to operate within, or at least a grand incomplete 
accountability system, because it holds the teachers and the 
children accountable, but it does not hold the system--the 
decisionmakers who promulgate the budgets are not held 
accountable.
    And just a quick piece of history. Ronald Reagan and the 
Governors started the whole business of let's have national 
standards. Under Bush we codified it--national standards for 
curriculum, national standards--voluntary, by the way--national 
standards for curriculum, national standards for testing.
    This committee, Democrats on this committee, fought very 
hard to add another set of national standards, national 
standards for opportunities to learn under leadership of now 
Senator Jack Reed and the late Patsy Meek--they were the great 
troublemakers. I just held their coats.
    We went up against the Clinton administration and Secretary 
Riley because the Governors, both Democratic and Republican, 
didn't want national standards for accountability because 
national standards--national standards for opportunity to 
learn.
    National standards for opportunity to learn said you know, 
as you--before you measure how well the kids and the teachers 
are doing in their schools, measure what you are doing to 
provide them with the resources that they need.
    Are you doing the things that will allow us to have 
qualified teachers, you know, libraries that are sufficient, 
school laboratories, et cetera?
    I don't think you have very many physics teachers in your 
physics classes in high school. Last time I looked there were 
few actual physics teachers who majored in physics teaching 
physics in high school.
    In the junior high schools, we don't even have laboratories 
in most of our junior high school science departments in New 
York.
    I want to say that we cannot continue to run away from 
opportunity to learn standards as being a part of the 
accountability model. We have to have--address this issue, not 
just focus on one part of it. That is qualified teachers.
    We do deal with that sort of as a footnote in No Child Left 
Behind. There is money in there that got reduced first when 
they started making cuts. But the teachers are bitter. The 
administrators are bitter. I am sure New York City is not the 
only place where they are bitter, because they think they are 
being forced into a system where they are not provided with 
appropriate resources.
    They need the books. They need the libraries. They need 
modern technology. So we need to zero in on, let's just say, 
qualified, diverse and stable staff.
    How can we accomplish that within the context of the 
present system, where not enough money is going to be made 
available, say, for New York City, which operates in 
competition with the surrounding suburbs and the rest of the 
nation?
    We train qualified teachers and they go off to higher 
salaries and higher opportunities.
    I want to focus on you, Mr. Chancellor, and say is there 
one--is there a possible way that we can combat this loss of 
qualified, diverse and stable staff? Because I don't think much 
is going to happen in improving these models.
    You know, I think Secretary Spellings pointed out that New 
York City is one of the places where we have still a great 
problem of certified teachers--still a problem. And the least 
certified teachers, non-certified teachers, and the substitutes 
are concentrated in low-income communities.
    Community leaders told you that 40 years ago. Nothing has 
changed. Or maybe something has changed. I hope it has. But 
basically, that is a truth that you have to leave with. Without 
the qualified teachers, we are not going to make the progress. 
Growth model, whatever you say--it is not going to happen.
    So, Chancellor, do you have any plans--I know that your 
administration talks about a conglomerate system willing to 
experiment, diversify. In order to hold teachers, I would take 
any experiment--I won't take but one experiment off the table. 
We don't want vouchers.
    But charter schools----
    Mrs. Biggert. I am sorry.
    Mr. Owens [continuing]. Non-profit schools--I don't talk 
fast enough.
    Mrs. Biggert. No, you didn't. I am sorry. Your time has 
expired.
    Mr. Owens. I will talk to you later about combining some 
kind of model to hold the teachers----
    Mrs. Biggert. The gentleman from Texas is recognized.
    Mr. Owens [continuing]. Have a stake--have them have a 
stake in what happens in the community.
    Mrs. Biggert. The gentleman from Texas is recognized, Mr. 
Hinojosa.
    Mr. Hinojosa. Thank you, Madam Chair.
    I want to say that it is--this is one of the best panels I 
have ever seen come before Congress to address one of the 
concerns that I have in education.
    And seeing the audience, the size of the audience, the 
makeup of the audience, pleases me very much, because it shows 
me that there are people out there wanting as much as I do to 
hear from experts like you on how we should address a problem 
that was uncovered yesterday by the Government Accountability 
Office.
    And I want to thank Marlene Shaul for you and your staff 
responding to our letter asking to give us the accountability 
of how No Child Left Behind was affecting our minority 
students.
    And I say that I have lots of concerns about education for 
the last three or four decades that I have been involved as a 
policymaker. But I want to say that that top concern is the 
persistently low high school graduation rate for minority 
students.
    And I like all the stories that each one of the panelists 
have given.
    And, Reg, you certainly did captivate those of us here as 
congressmen and congresswomen when you told the story about 
Cesar, because that reminded me of another young man who came 
from a math and science academy where more than 40 percent of 
the makeup of that student population are on the free lunch 
program. Eighty percent are Hispanic.
    And we have a young man who graduated from that school and 
went on to graduate from the school of engineering at the 
University of Texas-Pan American at Edinburgh at the age of 18. 
Unheard of.
    But the point is there are many limited English proficient 
students of low-income families who have risen and been able to 
function in our system.
    But what I hear from all of you and this gentleman, Mr. 
Sanders, as a statistician, is that we seem to have testing 
that is like a square peg being forced through a round hole, 
and it doesn't fit.
    And those tests obviously do not measure the progress that 
the children have made. And there are many schools throughout 
the country like each one of you has pointed out that are 
showing that something is working with many of the minority 
students.
    And it just seems like we in Congress don't have the 
political will to really invest the amount of money that it 
takes to be able to reach such large populations and to be able 
to bring the parental involvement that is one of the secrets of 
those schools that you all mentioned.
    Every single one of those schools, like mine in south 
Texas--there is parental involvement at a much higher level 
than in most of the other schools. So how do you do that? How 
do you get the certified teachers and all of that?
    So, Reggie, I want to start with you. Where does real 
accountability for high school graduation rates fit into the 
growth model of accountability?
    Mr. Weaver. Well, I think the accountability fits, again, 
when you give the people the opportunity to participate in 
helping to shape the circumstances that impact their work 
environment.
    As it relates to the point that you made about dropouts, it 
is absolutely critical that we have some kind of legislation or 
something that speaks to high school dropouts, that speaks to 
literacy coaches, that enable us to be able to have a better 
understanding with what these young people need in order to be 
successful.
    If, in fact, you really want to have the teachers to remain 
in the profession and to come into the profession, then I 
believe that you need to make sure that they have the respect 
that they deserve, that they have the opportunity to 
participate in the decisionmaking process, that you pay them, 
that they are involved in an atmosphere that is conducive to 
good teaching and learning. The schools have to be safe.
    And so these are the kinds of things that I do know will 
attract and keep the kind of people that we want in our 
profession to be able to work with these young people.
    Mr. Hinojosa. Kati, I am sure that being here in 
Washington, D.C., you must have participated and heard about 
the National Association of Governors meeting here in 
Washington on one issue, and that was the high school dropout 
rate, addressing it, and they brought in Margaret Spellings and 
all the experts we could bring, and we addressed it.
    So the question to you is why do you think that Congress 
has not been able to address the need for the resources that 
are needed after seeing that 50 Governors came, all concerned 
about this problem, and yet a year has gone by and----
    Mrs. Biggert. The gentleman's time from Texas has expired.
    Mr. Hinojosa. Thank you, Madam Chair.
    Mrs. Biggert. The gentleman from Michigan, Mr. Kildee, is 
recognized.
    Mr. Kildee. Thank you very much, Madam Chair.
    First, I apologize for being tardy for the meeting. I was 
down at the White House for the signing of the Voting Rights 
Act, another very important milestone in our country's history.
    I have seen most of you, know most of you.
    Reg, I see you all over the country, including my home town 
of Flint. Good to see you again.
    You know, the chairman made a comment that it would be 
great to have you in a better forum, a better setting with us, 
because we really--as has been mentioned, you are the most 
knowledgeable, helpful panelists that we have had.
    And it would be kind of nice if we had you as adjunct 
members of this committee as we work our way through the 
reauthorization, because I think you would be very, very 
helpful.
    Let me ask you this question. Under some growth model 
proposals, a school could have declining scores in one group, 
say economically disadvantaged group, and still make AYP. In 
our current system of disaggregated data, tell us this.
    How can we implement growth models and assure that the 
additional resources triggered by our current AYP system are 
still made available to that subgroup that is struggling? And 
if I may add another question to that, how would a blended 
system take care of the needs of that subgroup?
    Dr. Sanders?
    Mr. Sanders. When you--work with the Tennessee Department 
of Education in the proposal is that first of all, you start 
looking at all of the students as individuals. You don't pay 
any attention to subgroups first.
    You start looking at all the kids as individuals. Then 
after you have done that--of whether or not all these 
individuals are on trajectories or not, then you aggregate it 
into all of the existing subgroups required under No Child Left 
Behind.
    And for a school, then, to get credit for having passed AYP 
with this projection approach, all subgroups have got to meet 
it. So consequently, there will be no reason why that if a 
school is not hitting the trajectory for its ESL kids or its 
free and reduced-price lunch children, et cetera--if that 
subgroup doesn't meet it, they don't pass AYP with regard to 
the projection approach.
    Ms. Haycock. Let me, if I could, add one more point to 
that. And I very much agree with Bill's response. Under the 
current system, schools essentially make AYP or they don't. And 
you don't really have really good information on the sort of 
gradations of problem below that.
    If you add a growth component to that, you can distinguish 
between two schools where--that didn't meet the status bar but 
one of them just barely didn't meet it, and the kids are on a 
trajectory that is quite clear within the next couple of years 
they are going to be proficient.
    That school doesn't need nearly the extra intervention and 
resources that the school that is below that but not on a 
trajectory needs, and what the addition of the growth 
information does is allow you to make essentially more nuanced 
choices, decisions both about schools and how to categorize 
them, but also about what kinds of interventions are necessary.
    Mr. Kildee. Back again to the blended system, how would a 
blended system take care of the needs of those subgroups that 
may have special needs?
    Ms. Haycock. Bill is exactly right. In the blended system, 
the test is still both an overall test and a test for each 
subgroup, so you must make either the status, the safe harbor 
or the growth for every single subgroup.
    So there is still very clear subgroup accountability. That 
doesn't change at all.
    Mr. Kildee. Well, thank you very much. This is a very 
helpful group. I know you will be helpful as we begin the 
process of reauthorizing this bill. Thank you very much.
    Mrs. Biggert. Thank you.
    The gentleman from Virginia, Mr. Scott, is recognized.
    Mr. Scott. Thank you.
    And I want to add my voice to the comments you have heard 
about this panel as being one of the most helpful we have had 
on a difficult subject.
    We have heard about the perverse incentives that occur when 
you don't align the rewards with the behavior that you want.
    And, Dr. Sanders, you mentioned focusing just on the bubble 
and ignoring everybody else. And that same perverse incentives 
works around dropouts. There is a perverse incentive not to 
reach after people who have dropped out.
    It was my understanding that when the bill passed we put a 
provision in there that punished you for a high dropout rate. 
That was the intent of the provision. Do you know what has 
happened to that? Anybody?
    Ms. Haycock. I do.
    Mr. Scott. OK. Thank you.
    Ms. Haycock. Yes. In the case of high school 
accountability, you actually included a requirement that there 
be a second measure beyond performance on the assessment and 
that that measure be 4-year high school completion rates.
    However, by contrast to what you did on the achievement 
side, where you actually required the bar to be set at certain 
levels and to grow over time, you made no such requirement for 
the dropout level. You essentially left to states the 
definition of what is adequate progress in reducing dropouts.
    And the truth of the matter is that most states decided 
that not falling backwards very far was good enough.
    Mr. Scott. Well, that is an area that we have to, 
obviously, deal with to get that perverse incentive back in 
line.
    There is another problem we have had with special ed 
students. If someone qualifies for special education, obviously 
they are going to have a harder time achieving, and we want to 
align good teaching with reality.
    Unfortunately, if you do it too well, there is a perverse 
incentive to over-identify people as special ed. You have got 
some students that aren't doing so hot, you just call them 
special ed and they all of a sudden don't have to achieve.
    How can we avoid the incentive to over-identify and give a 
reasonable measure of how to measure special ed students?
    Ms. Shaul. The flexibilities the department provided only 
allow a certain percentage of students who have disabilities to 
be counted or excluded from the calculations for AYP.
    Mr. Scott. So if you over-identify, you still have to--only 
a certain percentage get excluded?
    Ms. Shaul. The students, for example--with the severely 
cognitively disabled students, up to 1 percent of those--those 
students--when they take those tests, they can be--1 percent of 
those students can be excluded.
    But they can be included if they pass the test. One percent 
can be excluded. But if there are more students who do not pass 
the test, they are still included in the overall count.
    So for that subgroup, they would not count as passing.
    Mr. Scott. The growth model helps if you are having a--if 
you have a high school and the middle schools from which you 
receive students are failing schools, you have a tougher job 
than if the students you receive were not failing.
    And yet the incentive there is to try to talk the school 
board into redistricting, so you get some better middle 
schools. If you don't have a growth model, how do you deal with 
that? Because if you receive all failing students and get half 
of them to pass, that is better than somebody who just 
maintains the status quo.
    Without a growth model--would the growth model help deal 
with that problem?
    Mr. Weaver. I don't know whether a growth model--I would 
think that it would, but I--you know, you mentioned special 
education, and then there was mention of the dropout.
    I believe that what we need are programs that help these 
individual students to feel, No. 1, that they can be successful 
and that that leads to them staying in school. It leads to them 
believing that they can be successful and attain a college 
degree or a trade degree.
    And as we talk about growth models, I believe that I want 
to inject in here something that I think is crucial, and that 
is pre-K, early childhood education, because if we get them 
earlier--the earlier we get them, the much better chance we 
have of----
    Mr. Scott. Well, let me get to that again in a quick 
question. You mentioned input and output. Can you see how that 
works from--are we moving the achievement gap? If you have got 
a gap, don't you have to improve the input on the students for 
whom--who are victims of the gap?
    Mr. Weaver. Well, see, the input to me means things that 
are important to achieving the--closing the gap--qualified and 
certified teachers, class size, adequate and equitable funding, 
safe and orderly schools, parental involvement.
    Those are the kinds of inputs that I believe are important 
to help the----
    Mr. Scott. And if you have an achievement gap you would 
have to emphasize those inputs for that subgroup.
    Mr. Weaver. I believe you need to emphasize that regardless 
of what the population of students are. The question is will 
the key policymakers have the commitment to make it happen.
    Mrs. Biggert. Does the gentleman yield back?
    Mr. Miller. Just on that point, it would seem to me that if 
the growth model is constructed correctly, and assuming a 
decent level of resources, that the growth model should in some 
ways equate expanded opportunities for those students who are 
not on that trajectory that you want.
    I mean, you know, we--under the schools in need of 
improvement, theoretically additional resources would flow. In 
some cases, it has. In some cases, it hasn't.
    But in theory, if you had that information, and the earlier 
you had that information and that trajectory, you should be 
able to make an intervention, so that information in itself 
will help those teachers, that school principal and that 
district understand the deployment of those resources.
    I am assuming that you would get additional resources, but 
even with existing resources you may make a more efficient use 
of those resources than you otherwise would when you didn't 
have that information.
    Mr. Klein. Absolutely. But it is important to understand 
that--I think everyone said it, but I want to underscore it. It 
means that each kid has real value in the accountability 
system, and that there is no reason for every kid--special ed 
kid, and language proficient kid--there is no reason for every 
kid not to be moving forward in a steady, coherent way.
    And the theory of a growth model is to say now, if you are 
not moving those kids forward, whatever the resource 
constraints--there is not a person who doesn't want to see more 
resources.
    But whatever the constraints are, you are able to then both 
target the resources in terms of differentiated strategies and 
also to look, frankly, in terms of the professional development 
of your workforce, so that you can make sure--why are these 
kids who started in the same place in one class--they moved 
forward a full grade.
    These kids who started in the same place in another class 
did not. What is that about? And you need to be able to address 
that day in and day out in the work we do, because--and that is 
what is wrong with NCLB right now.
    I have got kids in so-called high-performance schools who 
are not moving forward. And nobody is going to bother them 
because they meet AYP and everyone sees them as high 
performing.
    So if you want the kind of precision you are looking for, 
Mr. Miller, then I think you need a heavy dose of growth.
    And I want to end this way. Everybody wants to see 
proficiency. But if you look at New York City public school 
system right now, we have vast numbers of kids, and we have 
had, for as far back as anybody can remember, vast numbers of 
kids who are not close to proficient.
    And we have got to devise a system that gets them there 
over the time that they are in the school system. And we can 
pretend we have a national standard that says they are all 
going to get there on a certain date.
    But the truth of the matter is if they don't get there 
incrementally, year by year, they are not getting there. And we 
can deal with that issue in a multitude of ways--changing the 
proficiency standard, all sorts of other goofy things.
    Or we can get serious about the fact that the dimensions of 
the problem require each and every kid growing all of the time 
in the 12 years, not kids who are in the 12th grade in New York 
City who can't read. Those kids didn't grow, period, in the 
time they were in the school system.
    Mrs. Biggert. And that would be the last word. Thank you.
    I remind members that they have 14 days to submit questions 
for written answers from the witnesses.
    I would like to thank the witnesses for their valuable time 
and testimony, and both the witnesses and members for their 
participation.
    If there is no further business, the committee stands 
adjourned.
    [Whereupon, at 12:01 p.m., the committee was adjourned.]