A Special Section on Accountability
A Balanced School
An Alternative to High-Stakes Testing
The health of our public schools, Mr. Jones argues, depends on defining a
new model of accountability — one that is balanced and comprehensive.
And it needs be one that involves much more than test scores.
BY KEN JONES
OR SOME time now,
it has been apparent to
many in the education
community that state
and federal policies intended to develop greater school accountability for the learning of all
students have been terribly counterproductive. The use of high-stakes
testing of students has been fraught
with flawed assumptions, oversimplified understandings of school realities, undemocratic concentration
of power, undermining of the teaching profession, and predictably disastrous consequences for our most
vulnerable students. Far from the
noble ideal of leaving no child behind, current policies, if continued,
are bound to increase existing inequities, trivialize schooling, and misKEN JONES is the director of teacher education, University of Southern Maine, Gorham.
PHI DELTA KAPPAN
lead the public about the quality and promise of public education.
What is needed is a better means for evaluating
schools, an alternative to the present system of using
high-stakes testing for school accountability. A new
model, based on a different set of assumptions and
understandings about school realities and approaches
to power, is required. It must be focused on the needs
of learners and on the goals of having high expectations for all rather than on the prerequisites of a bureaucratic measurement system.
In the realm of student learning, the question of
outcomes has often been considered primary: what do
we want students to know and be able to do as a result of schooling? Once the desired outcomes have been
specified, school reform efforts have proceeded to address the thorny questions of how to attain them. Starting from desired outcomes is an important shift in how
to think about what does or does not make sense in
In the realm of school accountability, however, little
attention has been paid to corresponding outcome-related questions. It has simply been assumed that schools
should be accountable for improved student learning,
as measured by external test scores. It has been largely assumed by policy makers that external tests do, in
fact, adequately measure student learning. These and
other assumptions about school accountability must
be questioned if we are to develop a more successful accountability model. It would be well to start from basic
questions about the purposes and audiences of schools.
For what, to whom, and by what means should schools
be held accountable? The following answers to these
questions provide a set of premises on which a new
school accountability system can be based.
For what should schools be accountable? Schools
should be held accountable for at least the following:
• The physical and emotional well-being of students.
The caring aspect of school is essential to high-quality
education. Parents expect that their children will be
safe in schools and that adults in schools will tend to
their affective as well as cognitive needs. In addition,
we know that learning depends on a caring school climate that nurtures positive relationships.
• Student learning. Student learning is complex and
multifaceted. It includes acquiring not only knowledge of disciplinary subject matter but also the think-
ing skills and dispositions needed in a modern democratic society.
• Teacher learning. Having a knowledgeable and
skilled teacher is the most significant factor in student
learning and should be fostered in multiple ways, compatible with the principles of adult learning. Schools
must have sufficient time and funding to enable teachers to improve their own performance, according to
professional teaching standards.
• Equity and access. Given the history of inequity
with respect to minority and underserved student populations, schools must be accountable for placing a special emphasis on improving equity and access, providing fair opportunities for all to learn to high standards.
Our press for excellence must include a press for fairness.
• Improvement. Schools should be expected to function as learning organizations, continuously engaged
in self-assessment and adjustment in an effort to meet
the needs of their students. The capacity to do so must
be ensured and nurtured.
To whom should schools be accountable? Schools
should be held accountable to their primary clients:
students, parents, and the local community. Current
accountability systems make the state and federal governments the locus of power and decision making. But
the primary clients of schools should be empowered
to make decisions about the ends of education, not just
the means, provided there are checks to ensure equity
and access and adherence to professional standards for
By what means should schools be held accountable? To determine how well schools are fulfilling their
responsibilities, multiple measures should be used. Measures of school accountability should include both qualitative and quantitative approaches, taking into account
local contexts, responsiveness to student and community needs, and professional practices and standards.
Because schools are complex and unique institutions
that address multiple societal needs, there should also
be allowances for local measures, customized to meet
local needs and concerns. A standardized approach toward school accountability cannot work in a nation as
diverse as the U.S.
Given these premises, what are the proper roles of
a government-developed and publicly funded school
• It should serve to improve student learning and
school practices and to ensure equity and access, not
to reward or punish schools.
• It should provide guidance and information for
local decision making, not classify schools as successes
• It should reflect a democratic approach, including
a balance of responsibility and power among different
levels of government.
A BALANCED MODEL
An accountability framework called the “balanced
scorecard” is currently employed in the business world
and provides a useful perspective for schools.1 This framework consists of four areas that must be evaluated to
give a comprehensive view of the health of an organization. The premise is that both outcomes and operations must be measured if the feedback system is to be
used to improve the organization, not just monitor it. In
the business context, the four components of the framework are: 1) financial, 2) internal business, 3) customer,
and 4) innovation and learning.
Applying this four-part approach to education, we
can use the following aspects of school performance
as the components of a balanced school accountability model: 1) student learning; 2) opportunity to learn;
3) responsiveness to students, parents, and community; and 4) organizational capacity for improvement.
Each of these aspects must be attended to and fostered
by an evaluation system that has a sufficiently high
resolution to take into account the full complexity and
scope of modern-day schools.
1. Student learning. Principles of high-quality assessment have been well articulated by various organizations and should be followed.2 What is needed is a
• is primarily intended to improve student learning;
• aligns with local curricula;
• emphasizes applied learning and thinking skills,
not just declarative knowledge and basic skills;
• embodies the principle of multiple measures, including a variety of formats such as writing, open-response questions, and performance-based tasks; and
• is accessible to students with diverse learning styles,
intelligence profiles, exceptionalities, and cultural backgrounds.
Currently, there is a mismatch between what cognitive science and brain research have shown about human
learning and how schools and educational bureaucracies continue to measure learning.3 We now know that
human intellectual abilities are malleable and that people learn through a social and cultural process of con586
PHI DELTA KAPPAN
structing knowledge and understandings in given contexts. And yet we continue to conduct schooling and
assessment guided by the outdated beliefs that intelligence is fixed, that knowledge exists apart from culture
and context, and that learning is best induced through
the behaviorist model of stimulus/response.
Scientific measurement cannot truly “objectify” learning and rate it hierarchically. Accurate decisions about
the quality and depth of an individual’s learning must
be based on human judgment. While test scores and
other assessment data are useful and necessary sources
of information, a fair assessment of a person’s learning can be made only by other people, preferably by
those who know the person best in his or her own context. A reasonable process for determining the measure
of student learning could involve local panels of teachers, parents, and community members, who review data
about student performance and make decisions about
promotion, placement, graduation, and so on.
What is missing in most current accountability systems is not just a human adjudication system, but also
a local assessment component that addresses local curricula, contexts, and cultures. A large-scale external test
is not sufficient to determine a student’s achievement.
District, school, and classroom assessments must also
be developed as part of a comprehensive means of collecting data on student learning. The states of Maine
and Nebraska are currently developing just such systems.4
Most important, locally developed assessments depend on the knowledge and “assessment literacy” of
teachers.5 Most teachers have not been adequately trained
in assessment and need substantial and ongoing professional development to create valid and reliable tasks
and build effective classroom assessment repertoires.
This means that an investment must be made in teach-
er learning about assessment. The
T H E L O S A N G E L E S C O U N T Y O F F I C E O F E D U C AT I O N P R E S E N T S
value of such an investment is not
only in the promise of improved
PARENT EXPECTATIONS SUPPORT ACHIEVEMENT (PESA)
classroom instruction and measureFacilitator training for parent workshop leaders
ment. Research also shows that imHelp parents prepare their children for success.
proved classroom assessment results
Become a Certified PESA Facilitator and lead parent workshops at your school!
in improved student achievement on
PESA fulfills the requirement of providing parent involvement activities to improve student
academic achievement and school performance for the federal reform legislation of
Last, the need to determine the efthe No Child Left Behind Act of 2001 (Title I, Sec. 1118. Parent Involvement)
fectiveness of the larger state school
PESA facilitator workshops are available in English, Spanish, and Chinese.
system can either support or underPESA Facilitator Trainings are scheduled for:
mine such local efforts. If state or
Sept. 7-8, 2004
federal agencies require data to be
Nov. 30-Dec. 1, 2004
San Francisco, CA
aggregated from local to state levels,
Jan. 20-21, 2005
Feb. 1-2, 2005
local decision making is necessarily
Mar. 15-16, 2005
weakened, and an undue emphasis
is placed on standardized methods.
• The $300 registration fee includes the 2-day training, PESA Facilitator Manual,
If, however, the state and federal
instructional video, interaction wall chart, and refreshments.
agencies do not rely on local assess• Please call (800) 566-6651 for a Registration Form with locations.
ment systems to gauge the health
Discount for on-site PESA Facilitator Trainings.
To request a registration form or addtional information regarding
of the larger system, much may be
the TESA or PESA programs, please call (800)566-6651.
gained. In New Zealand, for example, a system of educational moniLook for the TESA training schedule on page 581 of this issue.
toring is in place that uses matrix
E-mail: firstname.lastname@example.org Website: www.lacoe.edu/PESA
sampling on tasks that include oneto-one videotaped interviews, team
tasks, and independent tasks. 7 No stakes are entailed
How should we define and put into practice our understanding of opportunity to learn? How will we measfor schools or students. The data are profiled and shared
with schools for the purpose of teacher professional de- ure it? How can an accountability system foster it?
velopment and as a means of developing model tasks
At a minimum, one might expect that schools and
for local assessments. Such a system supports rather than school systems will provide qualified teachers, adequate
undermines local assessment efforts.
instructional materials, and sound facilities. This is the
2. Opportunity to learn. How can students be ex- contention in a recent lawsuit, Williams v. State of Calipected to meet high standards if they are not given a
fornia, in which the plaintiffs argued for an accountfair opportunity to learn? This question has yet to be ability system that is reciprocal — that is, while schools
answered with respect to school accountability. Schools
are held accountable for performance, the state is held
should be accountable for providing equitable oppor- accountable for ensuring adequate resources.8
tunities for all students to learn, and we must develop
But there is more to this issue than just funding.
ways to determine how well they do so.
Jeannie Oakes describes a framework that includes opAt the heart of the matter is that the responsibility portunity-to-learn indicators for access to knowledge,
for opportunity to learn must be shared by the district professional teaching conditions, and “press for achieveand state. The inequitable funding of public schools, ment.”9 Linda Darling-Hammond stresses the “fair and
particularly the disparity between the schools of the humane treatment” of students in a set of standards
haves and those of the have-nots, places the schools of for professional practice.10
disadvantaged students in unjust and often horrifying
As such standards for opportunity to learn are arcircumstances. Over the past decade, there have been ticulated, the question arises as to how to monitor and
lawsuits in various states attempting to redress this im- report on them. Clearly, the degree of adherence to these
balance, which is largely a result of dependence on prop- standards cannot be determined through the proxy of
erty taxes for school funding. Yet not a great deal of testing. It is necessary to conduct observations in schools
progress has been made.
and classrooms and to evaluate the quality both of inAPRIL 2004
dividual teachers and of the school as a whole.
Teacher evaluation has received a great deal of criticism for being ineffective. The hit-and-run observations that principals typically conduct do little to determine whether teachers are meeting established professional teaching standards. Unions have been described
as more interested in protecting their membership than
in ensuring high-quality teaching. A promising development that has potential for breaking through this
impasse is the recent initiation of peer-review processes
by a number of teacher unions. Adam Urbanski, president of the Rochester Teachers Association and director of the Teacher Union Reform Network (TURN),
has been a leader in advocating for and implementing
such teacher evaluation processes. In a recent unpublished manuscript, he describes how the process should
• Some classroom observation by peers and supervisors, structured by a narrative instrument (not a
checklist) based on professional standards such as
those of the National Board for Professional Teaching Standards (NBPTS) and framed by the teacher’s
goals for the lesson/unit;
• Information from previous evaluations and feedback, such as structured references from colleagues
and other supervisors;
• Portfolios that might include examples of teaching syllabi, assignments made, feedback given to students and samples of student work, feedback received
from parents and students as well as colleagues, data
on student progress, teaching exhibitions such as videotaped teaching samples, professional development initiatives taken, and structured self-evaluation. All summative evaluation decisions about promotions or continued employment should be made by a specially
established committee of teachers and administrators.
Urbanski goes on to describe safeguards for due process
and for preventing malpractice. He also describes how
such a process could be used in conjunction with professional development for improving teaching and school
In order to evaluate the performance of a school as
a whole, a school review process will be necessary. Variations of inspectorates and school-quality reviews have
been developed in New York, Rhode Island, Maine, and
other states, as well as in Britain, New Zealand, Australia, and other countries.12 In order for such reviews
PHI DELTA KAPPAN
to serve the purpose of school improvement, the data
should be collected in a “critical friend” manner, through
a combination of school self-assessment and collegial
visitations. Findings from such a process should not be
employed in a bureaucratic and judgmental way but
rather should be given as descriptions to local councils
charged with evaluating school accountability. As with
all aspects of any school renewal initiative, the quality
and effectiveness of a review system will depend on the
time, resources, and institutional support given to it.
Who will ensure that adequate opportunities to learn
are present in schools? As described below, a system of
reciprocal accountability must be set up so that both
local accountability councils and the state itself serve to
“mind the store” for all students. The issue of equitable
funding will undoubtedly be resolved through the courts.
3. Responsiveness to students, parents, and community. Current accountability systems move power
and decision making away from the primary clients of
the education system and more and more toward state
and federal agencies. As high-stakes testing dictates the
curriculum, less and less choice is available for students.
Parent or community concerns about what is happening in the classroom and to the students have become
less important to schools than meeting state mandates.
As the primary stakeholders in the schools, parents
and communities must be made part of the effort to
hold schools accountable. There are many examples
of local community organizations, especially in urban
areas, that have taken on the task of insisting that schools
are responsive to the needs of children.13
To demonstrate responsiveness to students, parents,
and the community, schools must go beyond spon-
soring parent/teacher organizations or encouraging parent involvement as a means to gain support for existing
school practices. They must also do more than gather
survey information about stakeholders’ satisfaction.
True accountability to the primary clients for schools
entails shifting power relationships.
Local school-based councils must be created that
have real power to effect school change. These councils
would review accountability information from state and
local assessments as well as from school-quality review
processes and make recommendations to school boards
about school policies and priorities. They would hold
school boards accountable for the development and
implementation of school improvement plans. Phillip
Schlechty discusses how such councils might work:
Community leaders who are concerned about the
futures of their communities and their schools should
join together to create a nonprofit ...
Purchase answer to see full attachment