Commissioned by

I m p r o v i n g Le a d e r s h i p f o r Le a r n i n g

O 1 2 3 4 5 6
Data-Informed Leadership

Data-Informed
Leadership in
Education

Michael S. Knapp
Juli A. Swinnerton
Michael A. Copland
Jack Monpas-Huber
University of Washington

October 2006
 The Leadership Issue Project
State-of-the-Field Reports: Improving Leadership for Learning
This report is one of a series produced by a research team at the Center for the Study of Teaching and Policy, a
national research consortium home-based at the University of Washington. Developed with support from The Wal-
lace Foundation during the early stages of an initiative that explores central issues in the exercise of educational
leadership, the reports synthesize studies, conceptual work, and examples of current and emerging practice.
The reports are intended to clarify each leadership issue, while assembling what is known from
empirical studies. The information in these reports lays the groundwork for further study and practical
experimentation by leaders and reformers in states, districts, and schools.
The first report offers an overview of leadership and leadership support in relation to the overarch-
ing goal of improving learning. The remaining six explore in more detail particular issues within that terrain.

O
Overview
Leading, Learning, and Leadership Support
By Michael S. Knapp, Bradley S. Portin, Michael A. Copland, and Margaret L. Plecki.

1
Data-Informed Leadership
Data-Informed Leadership in Education
By Michael S. Knapp, Juli Ann Swinnerton, Michael A. Copland, and Jack Monpas-Huber

2
Resource Allocation
Allocating Resources and Creating Incentives to Improve Teaching and Learning
By Margaret L. Plecki, Christopher R. Alejano, Michael S. Knapp, and Chad Lochmiller

3
Redefining Leadership Roles
Redefining Roles, Responsibilities, and Authority of School Leaders
By Bradley S. Portin, Christopher R. Alejano, Michael S. Knapp, and Elizabeth Marzolf

4
Leadership Assessment
Purposes, Uses, and Practices of Leadership Assessment in Education
By Bradley S. Portin, Sue Feldman, and Michael S. Knapp

5
Improving Governance
Redefining and Improving School District Governance
By Margaret L. Plecki, Julie McCleery, and Michael S. Knapp

6
High School Transformation
Leadership for Transforming High Schools
By Michael A. Copland and Elizabeth Boatright

This document and the others within the series can be downloaded free of charge from the Center’s Web site,
www.ctpweb.org, and also from The Wallace Foundation’s Knowledge Center site, www.wallacefoundation.org.

The development of these reports was supported by a grant from The Wallace Foundation. Opinions represent those of the authors and not necessarily those of the Foundation.
1 Contents

Introduction: Data in Decisionmaking and Leadership.....................................................................2

• A Broader Focus: Data-Informed Leadership .....................................................................................5
• Scope of Discussion...........................................................................................................................9

Understanding Data-Informed Leadership...................................................................................... 11

• Anchors for Data-Informed Leadership............................................................................................. 11
• Cultures and Cycles of Inquiry..........................................................................................................15
• Relevant Conditions in the Policy Environment..................................................................................19
• The Elements of Data-Informed Leadership at Work.........................................................................21

Common Practices and Emerging Strategies in States, Districts, and Schools...........................23

• Efforts to Build Leaders’ Data Literacy and Expertise........................................................................23
• Efforts to Develop and Sustain Cultures of Inquiry . ..........................................................................24
• Engaging in Data-Informed Inquiry for Planning, Accountability, and Performance Tracking ..............28
• Efforts to Create Data Infrastructures and Data Systems .................................................................30
• Issues Arising in Patterns of Data Access and Use ..........................................................................34

Unanswered Questions and Enduring Dilemmas . .........................................................................39

• Unanswered Questions....................................................................................................................39
• Enduring Dilemmas and Ideological Tensions ..................................................................................42

References .......................................................................................................................................46

Endnotes............................................................................................................................................52

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 
 Introduction: Data in Decisionmaking and Leadership

An argument can be made that educational leaders have always had “data”
of some kind available to them when making decisions intended to improve
teaching and learning. Effective leaders gathered whatever information they
could readily access, and then drawing on accumulated experience, intuition,
and political acumen, they chose the wisest course of action to pursue. The
data they collected was likely impressionistic and rarely systematic, complete,
or sufficiently nuanced to carry the weight of important decisions.
Converging trends have shifted the basic terms of this equation, cre-
ating new possibilities for leaders to attain a deeper level of understanding
about the complexities of teaching and learning, and to learn how to maxi-
mize educators’ efforts to meet students’ needs. Consider this example from
a high school:

Staff members at Canyon View High School (pseudonym)

wanted to use their data to understand why more than half of the
school’s ninth grade students failed the state reading proficiency
examination. Working backward through the students’ education
experiences to determine the earliest occurrence of a characteristic
common to all students who had not passed the exam, the teachers
were shocked to see that most of these students had missed up
to 30 or 40 days in a 180-day school year when they were first
graders.

These ninth graders and the students in grades just below them
were already getting remedial reading help, but the new data pro-
vided an opportunity to save younger students from the same
fate. The district began more extensive screening of elementary
and middle school students who were likely to suffer academi-
cally because of high absenteeism in early years. Teachers, coun-
selors, and principals followed up by working closely with par-
ents—setting up telephone trees, for example, and in some cases

 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 making home visits—to make sure that the children got to school.
(Bernhardt, 2003, p. 29)

Various forces and conditions, both local and national, encouraged the
Canyon View staff to engage in data-based problem solving. Building on a
robust evaluation movement in the 1960s and 1970s, a variety of techniques
and strategies are now available for systematically evaluating the implemen-
tation, effects, and effectiveness of educational programs, policies, or initia-
tives. Standards-based reform has generated growing attention to outcomes
and results, with a corresponding lessening of interest in inputs. Since the late
1980s, the accountability movement associated with standards-based reform
has been steadily ratcheting up the demand for an evidence base to demon-
strate the effectiveness of student learning and educational programs. Finally,
the rapid increase in sophisticated technologies for handling digital informa-
tion makes the prospect of making educational decisions with a strong evi-
dentiary base more realistic, yet at the same time, more costly and complex.
In this context, forward thinking educators are beginning to envision a
future in which in-depth data analysis focused on student learning will be a
routine part of teachers’ and administrators’ daily work, and the most impor-
tant means for continuous professional learning. Consider this example from
a high-performing school district:

The Bodewell School District (pseudonym) is recognized region-

ally and nationally as a “high-performing” system, pursuing the
clear and specific goal of providing a college-preparatory educa-
tion for every student. For nine years, the district has realized
strong and steady growth in various measures of student achieve-
ment; however, data analysis now reveals a “plateau effect” in
key indicators. The district has stalled at about 80 percent of
students taking at least one Advanced Placement (AP) course,
and last year it saw a slight decline in state assessment scores in
mathematics.

Despite a systematic approach to curriculum, professional devel-

opment, and student support that extends to the earliest grade
levels, Superintendent Mark Rogers (pseudonym) has come to the
realization that the district has progressed as far as it can given

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 
 the current level of the knowledge and skills of its educators. He
notes, “What Ron Edmonds said years ago simply isn’t true—we
don’t have all the knowledge and skills we need to ensure every
student’s success. I’m saying, personally, that I’ve been at this
work for a long time, and I don’t know how to teach all kids in
a way that will guarantee their success.” Superintendent Rogers
believes that the only way to get continued improvement is to
turn the district into a real learning community, where exper-
imentation with new ideas and forms of instructional practice
and analysis of results of those experiments become the norm
for teachers’ work, rather than the heroic exception. He sees the
need for more sophisticated data structures that enable teachers
to gain even more clarity about the specific needs students bring
and learn from their ongoing efforts to improve learning. The
superintendent envisions a teacher community that is intimately
linked by the Internet, working to continually develop the range
and depth of the district’s curriculum, and using those same Web-
based structures to enable the sharing of knowledge about what
works, both inside and outside the district. He wonders how the
district can promote experimentation and harness new forms of
data and data use to break through the ceiling.1

The growing attention to questions of what counts as data, the devel-

opment of sophistication in understanding data, and the increase in tech-
nologies for manipulating data open up important possibilities for leaders
and the exercise of leadership throughout school, district, and state systems.
Coupled with support for continual professional and systemwide learning, as
suggested by the Bodewell case, the capacity for educational improvement
could increase significantly.
This paper explores these possibilities, both conceptually and in light
of research on the use of data in educational decision making. Based largely
on published accounts in the research literature and descriptive material on
established or emerging practices, the paper synthesizes and interprets ideas,
frameworks, beliefs, and activities concerning the availability, quality, and use
of data in the work of leaders at state and local levels related to the improve-
ment of teaching and learning. We concentrate primarily on those aspects of

 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
this emerging domain that have been systematically and empirically studied,
and therefore we have less to say about the burgeoning literature offering
advocacy, advice, or technical assistance to those who might make data a big-
ger part of leadership practice, or are already doing so (e.g., Holly, 2003; Hol-
comb, 1999; Leavitt, McDaniel, & Skogstad, 2004; Council of Chief State
School Officers, 2006). These latter bodies of work have much to offer lead-
ers, but they are not particularly helpful in gauging the nature or impact of
current practice, nor the conditions that enhance or limit that practice.
The paper unfolds as follows: First, following discussion of a broader
way of construing the use of data in leadership, we present ideas that can
help conceptualize the relation among the leaders’ access to data, the mean-
ings they give to it, and the uses to which they put this data, in the varying
settings in which leaders seek to improve teaching and learning. Next, we
briefly review the landscape of current practice, noting emerging strategies
that purport to improve the leaders’ access to, and use of, data for improve-
ment purposes, as well as conditions that limit or complicate leaders use of
data. Finally, we conclude with questions that emerge from the review and
beg to be answered through further experimentation and research, while not-
ing enduring tensions that will always be present in data-informed practice
and cannot be “resolved” through further effort or study.

A Broader Focus: Data-Informed Leadership

In the current context of accountability and school reform, data-driven deci-
sion making is increasingly seen as an essential part of the educational lead-
er’s repertoire, yet more is at stake—and more is possible—than this term, or
even the term data-based decision making, might imply. As might be inferred
from the vignettes, it is not just a question of laying out test scores, noting
areas of weakness, and mounting remedies that are clearly indicated by pat-
terns in the data. We find the term data-informed leadership a more useful
concept for considering what is, and might be, involved in this territory. The
term broadens the scope of thinking and action in two productive ways.
First, shifting to the concept of data-informed leadership escapes the
occasional deterministic implication of data “driving” action. Tempting as
it may be to imagine educational leaders’ actions as single-mindedly “driven”
by “bottom-line numbers,” wise leaders are likely to take more into account
as they frame a response to the challenges they face. While they can be fully

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 
 informed by data when they take action, they also bring core values and
insights into those aspects of practice for which there is not yet good data,
and may never be. Moving away from the potentially appealing rhetoric that
data can provide clear, indisputable direction for future action, the notion of
data-informed leadership captures the complex and often ambiguous nature
of data use in educational settings. According to Bernhardt (2004), “True
data-driven decision making is only partly about data. A clear and shared
vision and leadership play major parts in data-driven decision making” (p.
18). And Weiss (1995) reminds us that no matter how systematic and compre-
hensive the data gathering, several other factors are always likely to influence
decision making, including interests, ideologies, and institutional context.
Second, the concept presumes that data are useful for more in the prac-
tice of leadership than the making of decisions per se. For one thing, given the
inherent ambiguity and multiple meanings of much data in educational set-
tings (Coburn & Talbert, 2006; Honig & Coburn, 2005), data may prompt
questions and deliberation more than they point to specific decision options.
For example, certain data points (e.g., disaggregated state test scores) may
provide an awareness of a given situation, such as low scores for ninth grade
students in Canyon View High School, but the data do not necessarily indi-
cate how educators should address the issue at hand. In this example, assess-
ment data certainly inform conversation about possible actions, but they do
not necessarily “drive” decisions or provide information about how best to
address the issue of low performance. In fact, while the vignette described
district leaders’ attempts to prevent students in elementary grades from expe-
riencing similar struggles in high school, the data did not provide specific
direction for addressing the needs of current ninth grade students whose per-
formance was the basis for data analysis.
Because leaders do far more than make decisions, data can serve a
range of purposes in the leaders’ toolkit, as Table 1 suggests (e.g., Bernhardt,
1998; Boudett, City, & Murnane, 2005; Holcomb, 1999). Some of these sit-
uations imply internal, essentially “private” purposes, played out within a
leadership team or the inner circle of individuals with whom a leader works
most closely, while others imply audiences that are more public. Each implies
different ways of representing what the data say and communicating it to
intended audiences.

 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 Table 1: A Range of Ways that Educational Leaders Use Data

Type of leadership activity

(with and for internal or
external audiences) How data are used and what kinds of data are implied

Diagnosing or clarifying instructional State and local leaders seek to know whether, or to what extent, the learning that is
or organizational problems (primarily occurring for students in the local context matches those overarching expectations for
internal to the decision making group) learning (standards) established at the top of the system. Therefore, leaders would
seek out information such as disaggregated scores on criterion-referenced state
assessments that reflect one measure of student learning in particular content areas.

Weighing alternative courses of action State and local leaders use data to evaluate existing programs or curriculum
(primarily internal) approaches and (where they have relevant data) to judge their potential in comparison
with alternative programs, some of which may be implemented in pilot form.
Comparative implementation and outcome data would therefore be especially helpful
in such circumstances—e.g., to judge their relative contributions to a learning
improvement agenda.

Justifying chosen courses of action Data (e.g., concerning learner characteristics, learning outcomes, comparative
(primarily external) program benefits, school closure decisions) are used selectively to make a compelling
case for programs or courses of action that may or may not have been chosen on the
basis of the data.

Complying with external requests for State and local leaders are careful to generate information requested by external
information (external) agencies, authorities, or groups on which their funding or legitimacy depend—for
example, descriptions of how different learner groups are served on evaluations of
services to these groups.

Informing daily practice (internal) Data of various kinds are used by administrators and teachers to guide daily practice.
The data are often informal, gathered in mid-stream, and in a form that can be
immediately interpreted and used by a practitioner for refining teaching and learning.

Managing meaning, culture, and Data help leaders understand and guide the cultural aspects of the professional
motivation (internal) workplace, by representing to staff what the organization is accomplishing, how
people feel about their work, what matters in the work, and what professional
learning needs exist.

As the Table 1 categories make clear, not all of these leadership actions
imply specific decisions, but rather imply a range of actions (including the
investigation of new questions) in which data, appropriately interpreted, help
leaders understand what is happening in educational organizations, repre-
sent it to others inside or outside of schools, and fashion courses of action.
Furthermore, the policy and community environments in which educational
leaders work are likely to prompt a variety of uses of data, by

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 
 • Demanding information from the educational system about its per-
formance (as in accountability systems) or the effectiveness of par-
ticular programs (as in the evaluation requirements accompanying
categorical program funding).

• Offering sources of data or help with assembling or interpreting data

(as technical assistance centers, universities, or vendors may do).

• Creating occasions for inquiry (as when an influx of new immigrant

children raise questions about appropriate educational programs,
school assignments, and so on).

• Promulgating public images of the educational system’s functioning (as

in media accounts that beg for response, clarification, or refutation).

• Raising questions about the school system’s policies or responsiveness

to particular constituencies or needs (as in legislative debate about
support for teacher induction or school board debate about school
closures).

Converging conditions in the field bring all of these forces into play. The
federal No Child Left Behind Act (NCLB), for example, both demands infor-
mation about school system performance and creates occasions for inquiry
into the quality of its educational program. This legislation requires that all
schools receiving federal funds shall make available report cards that provide
specific data in three major areas: assessment, accountability, and teacher qual-
ity.2 These reporting requirements have prompted a flurry of activity related
to more fine-grained data collection, distribution, and analysis. Not only do
states and districts have to provide the data to the federal government, the
legislation also requires giving parents and other key stakeholders access to
these reports, thereby creating an obvious reference point for media accounts
and other public representations of school system performance, often nega-
tive. Most states have developed online Web portals that provide access to
NCLB-related data organized by state, district, and school level. Such public
availability of data has multiple implications for leaders as they interact with
those both internal and external to their organizations, among the implica-
tions are the continuing questions about the effectiveness of the system. The

 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
requirements have also stimulated the growth in the availability and sophis-
tication of data systems targeted to education, often made available through
private providers, which offer educational leaders a good deal of help (at a
price) in using data as a leadership tool (Burch, 2005).
These instances and many more reflect the range of ways that external
environments can prompt, support, or require educational leaders to make
use of data in support of improving teaching and learning, and often point
toward particular kinds of data that matter most to particular constituencies.
At the least, these events make it hard to ignore the need for data; at best, they
represent an opportunity to use data to strengthen the planning and execu-
tion of educational programs, as well as public support for them.

Scope of Discussion
To explore further what the concept of data-informed leadership might entail,
we need first to clarify what we mean by data and what leaders might be
using them. Here, we limit our attention to data implicated in what is argu-
ably the central function of educational leaders—to guide, direct, assess, and
support teaching and learning. For the purposes of this paper, we concentrate
on data as information that
1. Represents the content or conduct of instruction or its effects on
student learning and the student experience, as well as the factors
and conditions that most immediately affect these matters.

2. Is, or could be, used in leadership actions aimed directly at the im-
provement of instruction, learning, and the student experience, or the
organizational conditions that support instructional improvement.

A wide range of data, both quantitative and qualitative, fall within this
boundary. While leaders and their audiences may often use data that can be
quantified or averaged, such as grades, graduation rates, teachers’ experience
levels, and scores on state assessments—and they are likely to pay special
attention to test scores to which accountability consequences are attached—
there is clear evidence that many forms of qualitative evidence, such as captur-
ing the qualities of student work, teachers’ perceptions, or various features of
classroom-based assessment, have as important a role in improving teaching
and learning as their quantitative counterparts. As the boundary definition

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 
 makes clear, we are particularly interested in data that pertain most directly
to instruction—though other aspects of school system functioning may be
included as well. In other words, we are not focusing mainly on how leaders
use data for budgeting (i.e., dollar figures) or space utilization (square feet of
floor space in different buildings) or transportation planning (miles for bus
routes) and so on.
We also acknowledge that data are not the same as evidence. Put
another way, data by themselves are not evidence of anything, until users of
the data bring concepts, criteria, theories of action, and interpretive frames
of reference to the task of making sense of the data. In this regard, flooding
leadership practice with data is unlikely to bring about much improvement,
and even could get in the way, absent time and attention to the central issue of
making sense of the data. We will return to this matter in more detail as we
offer a framework for thinking about data-informed leadership.
Data and evidence are of potential importance to leaders working in
different places within the educational system. We are especially interested
in data use in four locations: (1) at the state level, among participants in the
policy community who deliberate policies related to instructional improve-
ment or who seek to implement such policies, e.g., agency officials and staff,
legislators and their staffs, professional associations, and advocacy groups;
(2) in district central offices, e.g., school board members, superintendents,
directors, and other staff who are involved in decision making focused on
instructional improvement; (3) in schools, e.g., principals, department heads,
teacher leaders, and others who take part in instructionally related inquiry,
and (4) in classrooms, as teachers themselves seek to improve their work or
as others, e.g., instructional leaders, work with teachers on various aspects
of their practice. All four are potentially engaged in data-informed leadership,
broadly construed, and, hence, our discussion concerns the ways that data are
or are not part of their daily practice.

10 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
Understanding Data-Informed Leadership

Three sets of ideas from recent lines of scholarly work help us understand
what is—or could be—happening in the case vignettes from Canyon View
High School and Bodewell School District, and in any settings in which data-
informed leadership practice is in place or being attempted. The first set con-
cerns conditions that anchor data-informed leadership—leadership focus, the
users’ beliefs and expertise, and the kinds of data available to the users. The
second set highlights the building of cultures of inquiry and engagement of
leaders and others in cycles of data-informed inquiry and action. The third
set directs attention to activities in the policy environment that prompt, guide,
and support leaders’ work with data, especially through investments in the
development of data infrastructures and leaders’ data literacy. All three work
together to shape whether and how leaders make use of data in the exercise
of leadership.

Anchors for Data-Informed Leadership

Before considering the ways that leaders make use of data, several conditions
can be identified that have enormous influence over their capacity to work
with data: what they are focused on, what they believe and know how to do,
what they are seeking to influence and how, and what data are either available
to them or can be readily generated.
Focus for data-informed leadership. Leaders are in a position to define
the focus for the data they might generate and use, reflecting their own lead-
ership priorities and their response to events that call for data and evidence.
Absent focus, data-informed leadership is an empty exercise, consuming time
and yielding little of consequence. While many foci are possible—such as the
number of students enrolled in Advanced Placement (AP) courses—we would
argue that a persistent, public focus on learning improvement offers an espe-
cially important reference point for the leaders’ use of data, with emphasis on
data concerning efforts to improve the quality of teaching and learning (e.g.,
Knapp, Copland, Talbert, 2003; Stoll, Fink, & Earl, 2003).3 In keeping with
this focus, data are a potentially useful resource for

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 11
 • Leadership that focuses attention and effort on improving student
learning. Both quantitative and qualitative data can help identify
what students know and can do, and they can help suggest aspects
of teaching that need to improve, e.g., through classroom assessment
for differentiating instruction and grouping by ability; by formative
assessment to refine instruction and enhance motivation; student self-
assessment; and so on.

• Leadership that guides the learning of individual professionals.

Quantitative and qualitative data about various aspects of professional
practice can stimulate productive conversation and problem-solving
by teachers and administrators. In the hands of a skilled leader, data
become a tool for focusing professional learning on the improvement
of daily practice.

• Leadership that guides what has been called “system learning”

(Knapp et al., 2003, p. 16). Various data can provide a picture of the
system’s functioning as a whole, documenting accomplishments and
helping to spot problems that need work.

This focus for leadership is not the only one that can be imagined, but it
prompts numerous possibilities for bringing data to bear on the improvement
of practice, while recognizing that the effort to improve practice entails more
than just student achievement scores.
Core values and theories of action. Whatever the leaders’ focus, data-
informed leadership rests on a foundation of values and strategic thinking that
guides the leaders’ reach for data, engagement in inquiry, meaning-making, and
subsequent actions. As noted in work on the moral dimensions of leadership
(e.g., Fullan, 2001; Sergiovanni, 1992), leaders’ work implies, and often is rooted
in, core values that concern the ultimate purposes of schooling, principles of
equity, and the justification for leadership strategies of all kinds. A number of
such values underlie efforts to focus on learning improvement; among them are
these five: ambitious standards for student learning, belief in human capacity,
commitment to equity, belief in professional support and responsibility, and
commitment to inquiry (Knapp, Copland, & Talbert, 2003). The latter value
highlights the use of evidence to plan, evaluate, and change practice, if not to
establish the scope and reach of the problems that the leaders hope to address.

12 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 Values such as these are implicated in the theories of action held by lead-
ers and in some instances shared more widely in the organizations they lead.
Treated as a set of assumptions about how the world works and a rationale
for how one can intervene to improve it (Argyris & Schon, 1978), a theory of
action is often implicit and may not be held by all parties in a given decision-
making situation, but it can almost always be discerned and represented as
the logic that connects the leaders’ initial framing of the problem, subsequent
leadership actions, consequences for teaching and learning (or the conditions
that support these matters), and the learning that participants experience
based on the results. Central to this aspect of the framework are two sets of
ideas that leaders hold (Fullan, 1999):
• Ideas about what constitutes “good teaching and learning” and how
it occurs, sometimes referred to as a “theory of education” or “theory
of instruction.”

• Ideas about what interventions by leaders and others will bring about
good teaching and learning, or at least improve existing practices so
that they come closer to a desired ideal, sometimes referred to as a
“theory of change.”
These ideas highlight certain actions, responses, and contextual condi-
tions—and the relations among them—that become the “variables” in educa-
tors’ inquiries into questions about practice and performance. In short, they
define what data leaders might wish to collect and how they might interpret
such data.
Leaders’ data literacy. If core values and theories of action establish a
sense of the ultimate purpose and rationale for engaging in inquiry, leaders’
expertise with data—what may be referred to as their data literacy (Earl &
Katz, 2002)—defines how much and what they are able to do with data. The
challenge is more than a technical one limited to the assembling and manipula-
tion of information, but rather it extends to what Fullan (2001) calls “knowl-
edge building,” the capacity to extract and share useful meaning from orga-
nizational experience. Thus subsuming the capacity of leaders and others to
assemble and use data responsibly in their daily practice, data literacy presumes
more than trial-and-error experience with data. It presumes an accumulating
facility with the interpretation of data, not to mention a familiarity with data
sources and creativity in assembling relevant data quickly and efficiently. As

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 13
 implied by work on cultures of inquiry (Copland, 2003), members of a school,
district, or other educational organization can become more literate in the use
of data and committed to this feature of their collective practice.
Available data and data sources. Given a focus on learning, leaders’
ability to bring data to bear on it is shaped in large measure by the actual data
they can find or generate with a reasonable investment of time and resources.
Some of this data, especially those which are likely to count in demonstrating
accountability to district, state, or federal audiences, reside in information
systems created through state policies and investments—such as those that
have created data warehouses, management information systems, or report-
ing systems. Other sources are more likely to be homegrown, derived from
the leaders’ own efforts to put together data that has meaning and usefulness
in the local situation, or from research and media accounts, or from other
efforts to represent what is going on in schools (Weiss, 1995).
Table 2, adapted from Bernhardt’s (1998) work, provides an overview
of the kinds of data educators may use as they engage in data-informed deci-
sion making, especially in information-rich environments.

Table 2: Types of Data Available to Educational Leaders in Information-Rich Environments

Data Category Sample Data Points

Student demographics Enrollment, attendance, dropout rate, ethnicity, gender, grade level (by school, district, etc.)

Perceptions Perceptions of learning environment, values and beliefs, attitudes, observations … (e.g.,
held by a school’s teachers, districtwide educators, or the local community)

Student learning Standardized tests, norm/criterion-referenced tests, teacher observations, authentic

assessments …

School processes Descriptions of programs, instructional strategies, classroom practices …

Teacher characteristics, Teacher assignment (grade, subject area, students served), qualifications, retention,
behavior, and participation in professional development …
professional learning

(adapted from Bernhardt, 1998)

From these raw materials, leaders who treat information as a useful lead-
ership tool may conduct various kinds of inquiries, including the use of simple
indicator systems that offer “warnings and hints” about system performance
such as the following seven indicators (Celio & Harvey, 2005): student achieve-

14 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
ment, trends in the achievement gap, student attraction (a school’s ability to
attract students), student engagement with school (e.g., attendance), student
retention/completion, teacher attraction and retention, and funding equity.

Cultures and Cycles of Inquiry

The capacity for data-informed leadership—embodied in leaders’ values,
expertise, theories of action, and the availability of data—sets the stage for
particular leadership activities that bring systematic information into con-
sideration by leaders and others. Specifically, educational leaders who are
so inclined build “cultures of inquiry” in their respective organizations and
engage, along with others, in cycles of data-informed inquiry and action. This
may mean being open to going beyond the initial boundaries of a given ques-
tion or problem, as was the case when Canyon View teachers followed a data
trail that began with ninth grade performance data and lead to actions at the
elementary level.
The creation of organizational cultures that enable and motivate
data-informed leadership. Data are only useful to the extent that leaders and
those who work with them ask questions that can be answered with the data.
Schools, districts, and other educational settings vary in the degree to which
they make data a prominent feature of deliberation about the myriad issues that
confront them on a daily basis. The literature is beginning to offer a number
of examples of educational organizations in which participants accept—even
hunger for—data, as they plan and implement their respective programs.
Such instances appear in descriptions of “reforming districts” (McLaughlin
& Talbert, 2002); schools engaged in “cycles of inquiry” (Copland, 2003);
schools in the midst of school improvement planning or “self-reflective renewal”
(Striefer, 2002; Portin, Beck, Knapp, & Murphy, 2003); and schools enacting,
or responding to, accountability systems (Spillane et al., 2002; Lemons, Luschei,
& Siskin, 2003).
In these cases, leaders have taken deliberate steps to build a culture that
supports inquiry into the pressing problems facing the organization. Such
a culture engenders trust and reduces the perceived risk from asking and
answering questions about practice and performance (Copland, 2003), and
it ultimately can support collective learning (Scribner, Cockrell, Cockrell, &
Valentine, 1999).

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 15
 Figure 1. Culture and Cycles of inquiry

Accessing or
searching

Problem
Sensemaking
(re)framing
CULTURE
OF
INQUIRY

Learning from action Taking action

Learning from action and communicating
Implementation
Building
and effects on
cultures of
student,
inquiry
professional, and
system learning

A central part of the culture of inquiry is that many players are partici-
pants in it, often implying that data-informed leadership is distributed, as are
other aspects of the exercise of leadership. In line with recent formulations of
the idea of distributed leadership (e.g., Elmore, 2000; Spillane, 2006), leaders
who find ways to stimulate and sustain inquiry into problems of practice con-
fronting a school, district, or state system invite others to share in the framing,
conduct, and interpretation of the inquiry and the subsequent actions based
on it. The participants often become co-leaders, and over time they develop
shared norms and expertise in data-informed problem solving. Such activities
emphasize expert over hierarchical authority, an essential attribute of distrib-
uted leadership arrangements (Bennett, Wise, Woods, & Harvey, 2003). Such
arrangements also recognize that the knowledge and skills necessary to shape
or exercise data-informed leadership may be located within a professional
community of practice more than in a particular individual (Wenger, 1998).
That said, leadership informed by data may not be shared equally among
participants, as research on committee deliberations about math performance
in a school indicates. When committee members held different beliefs about

16 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 what the data “said,” it was the leader with positional power whose framing
of the problem predominated (e.g., are we facing a curriculum problem or a
professional development problem?) and whose beliefs largely informed the
final decisions for action (Coburn, 2006).
Engaging in cycles of data-informed inquiry and action. Cultures of
inquiry develop from repeated attempts to bring data to bear on key problems
facing the school, district, or state system. In turn, having such a culture in
place supports leaders’ and colleagues’ efforts to ask questions about the prob-
lems of practice that can be answered with data within such settings. At least
five phases of activity, schematically represented in Figure 1, define this kind of
“inquiry in action,” work that connects data to learning improvement.
• Focusing and (re)framing problems for inquiry. In line with their ex-
pressed or implicit theories of action, leaders focus attention on prob-
lems of practice and frame them in terms that invite inquiry. Work that
highlights problem-framing ability (Cuban, 1998) and the capacity to
reframe problems from multiple vantage points or perspectives (Bol-
man & Deal, 1997; Copland, 2003) captures what leaders do, or can
do, to set inquiry in motion, thereby giving context for the use of data.

• Accessing or searching for data and evidence. In relation to problems

they choose to address, the leaders and their collaborators generate or
search for data using available inquiry tools, sources, and strategies, as
delineated in various works on “organizational learning” (e.g., Huber,
1991; Honig, 2006), or they simply access data that are already avail-
able. Leaders may or may not have the appropriate data and tools for
manipulating it at their fingertips (Heritage & Yeagley, 2005), which
has prompted rounds of developmental work in recent years, as schol-
ars and others develop tools to support data-based leadership (e.g.,
Holcomb, 1999; Leithwood, Aitken, & Janzi, 2001); in a similar vein,
vendors have been increasingly active in support of districts and school
data use (e.g., Burch, 2005; Colgan, 2004; Wayman, Stringfield, &
Yakimowski, 2004).

• Making sense of data and its implications for action. With data in
hand, leaders create occasions for making collective sense of the data
and probing the data for possible action implications. Here, drawing
on underlying frameworks concerning sensemaking in organizations

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 17
 (Coburn & Talbert, 2006; Weick, 1995), recent work has begun to
outline how leaders approach the sensemaking task (Spillane, Reiser,
& Reimer, 2002). The leap from data to action is not simple. How-
ever, scholarship that captures patterns of actual data use in school
districts, for example, notes how ambiguous the data often are, a fact
that can curtail the data’s perceived usefulness but which can also
stimulate deliberation about ways to serve student needs better (Ho-
nig & Coburn, 2005). In addition, individuals’ conceptions of what
counts as evidence, how evidence should be used, and how research
informs practice vary across systems, often informed by where an in-
dividual sits within an organization (Coburn & Talbert, 2006). Thus,
the same data may likely be interpreted differently and suggest different
courses of action depending on who is engaged in decision making.

• Taking action and communicating the action in different arenas. In-

formed by the sense they make of the data, and by other matters not
intrinsic to the data (e.g., the politics of the situation, basic values, re-
porting demands), leaders take action and communicate what the data
say to relevant audiences. Some actions take place out of the public
eye, but others are visible to relevant audiences and invite explicit com-
munication by leaders (Witherspoon, 1997). Data become an integral
part of the leaders’ actions and communications, and so a central part
of the leaders’ work is “making it public” in ways that are respectful
and politically astute (Holcomb, 1999).

• Learning from action through feedback and further inquiry. Inquiry-

oriented leaders construct feedback loops so that they and other par-
ticipants can learn about and from the implementation and effects of
their actions and can reframe leadership problems (Halverson, 2003).
Scholarship by cognitive scientists on short-term “quasi-repetitive
feedback cycles” supports the notion that regular feedback can be a
powerful influence on learning and, by implication, the learning of
leaders who receive such input (Schwartz, Bransford, & Sears, 2005).
Not surprisingly, syntheses of work on effective school leadership draw
attention to the role that feedback can play as an influence on leaders’
and others’ learning (e.g., Hattie, 1992, as cited in Marzano, Waters,
& McNulty, 2005).

18 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 Presented this way, leaders’ attempts to make use of data within cycles
of inquiry sound logical, rational, and orderly. In actual practice, these cycles
are likely to be more “messy,” and they are likely to differ considerably
depending on the participants’ experience and comfort with inquiry, as in
research that has identified schools exhibiting “novice,” “intermediate,” and
“advanced” cultures of inquiry (Copland, 2003), as well as on where data
users reside in relation to the organization (Coburn & Talbert, 2006). But the
underlying impulse is the same, regardless of the sophistication with data use:
to raise questions about practice and to develop insights into these problems
by considering what can be learned from data about practice.

Relevant Conditions in the Policy Environment

As noted earlier, events in the policy environment surrounding schools—
especially federal accountability pressures and related requirements from state
standards-based reform policies—compel leaders to use data in their daily
practice. But other environmental events affect the exercise of data-informed
leadership, in particular, two kinds of investments.
Investments in the development of leaders’ data literacy (e.g., through
leadership development or certification programs) coupled with ongoing sup-
port for leaders’ use of data (e.g., through relations with third-party groups
and vendors, or through in-house experts such as those who may reside in
district research-and-testing offices) are likely to increase the chances that
leaders learn what they need to know to work efficiently with data. However,
naïve overreliance on external support from external parties or software tools
may result in leaders bypassing important questions that grow out of core
values, institutional priorities, and local issues.
Investments in the development of data infrastructures. The design
of data systems by state or local agencies seeks to anticipate data elements
that will matter to leaders or their audiences. The extent to which they suc-
ceed in doing so has a lot to do with how useful leaders find them. Such data
systems can also be cumbersome, as they often involve large-scale, routine
data collection from sources such as district central offices. The quality and
timeliness of the data they collect vary, in part a reflection of how the sys-
tem attends to “data cleaning,” a prerequisite for maintaining data accuracy
(Mieles & Foley, 2005; Stringfield, Wayman, & Yakimowski, 2005). Based
on discussions of quantitative data infrastructures in the literature, the fol-

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 19
 lowing aspects of such systems are of particular concern to the practice of
data-informed leadership:
• The specific data elements that reside in the data infrastructure.

• The accuracy and completeness of the data, and whether it is updated

regularly.

• The timing and timeliness of data availability. Local educators, for

example, often lament the lag time between state assessment adminis-
tration and its availability to school and district audiences four to five
months later, often in the school year following the year the test was
administered.

• The architecture of the data storage and retrieval system, and whether
it enables easy, flexible, disaggregated queries that relate one data ele-
ment to others.

• The ease of access to the data system by a variety of users, with suf-
ficient safeguards to maintain confidentiality (where necessary) and
counter attempts at tampering.

• The cost of building and maintaining the data infrastructure.

Although it is tempting to treat infrastructure issues as solely or pri-
marily concerned with statewide quantitative databases, leaders may also
access or create local data sources, both quantitative and qualitative, that
are especially pertinent to the problems they face. In this regard, a variety
of leadership activities—such as “walk-throughs,” fast becoming a feature
of school and district instructional leadership work (e.g., Kerr, Marsh, Ike-
moto, & Barney, 2006), local action research projects of various kinds (e.g.,
Stringer, 2003), and local data collection for school improvement planning
(e.g., Striefer, 2002)—have an important role in providing immediate, often
qualitative, information to leaders about their improvement strategies and
effects.

The Elements of Data-Informed Leadership at Work

The elements just discussed bear a straightforward relation to one another, as
suggested in Figure 2 below. The anchors for data-informed leadership define
direction and capacity for this facet of leadership activity. Cultures and cycles

20 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
of inquiry bring participants together in the act of seeking, interpreting, and
acting on information they gather around problems they come to frame in
increasingly sophisticated terms, depending on their continual use of data.
Policy environments prompt the use of data in the first place and also provide
the wherewithal to support development of data literacy and the expansion of
data infrastructures.

Figure 2. Element of Data-Informed Leadership

State and local policy environments

Investment in
data infrastructure
and access Accessing or
searching

Problem
Sencemaking
(re)framing
CULTURE
OF
Investment in
developing leaders’ INQUIRY
expertise and data literacy

Learning from action Taking action

Learning from action and communicating
Implementation
Building
and effects on
cultures of
student,
inquiry
professional, and
system learning
Anchors for data-informed leadership
(leadership focus, core values, theories of action, data literacy, available data)

These elements may appear to suggest an idealized picture or model

of how data-informed leadership should work. That is not our intention,
although we have purposefully phrased the elements in terms that would cap-
ture the activities of educational systems in which data play a central, vital
role in leadership. But the elements are more properly understood as variables.
Cultures of inquiry, for example, can be developed to varying degrees; leader-
ship development to enhance data literacy may or may not be a focus of state
or local investment; and so on. This framework for data-informed leadership
presents ways of identifying and understanding what is—or is not—taking
place in particular state or local settings.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 21
22 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
Common Practices and Emerging Strategies in States,
Districts, and Schools

The categories of the previously described framework offer a useful organizer

for considering the evidence from published literature, much of it descrip-
tive, concerning the nature, extent, and effects of data-informed leadership in
contemporary practice. The literature sheds light on four primary streams of
activity at state and district levels, each of which relates to a central element
of the framework:
• Anchors for data-informed leadership: Efforts to increase leaders’
data literacy and expertise.

• Cultures of inquiry: Efforts to develop and sustain cultures of inquiry

in schools, districts, and state agencies.

• Cycles of inquiry: Efforts to use data for school improvement plan-

ning, demonstrating accountability, and managing district programs.

• Conditions in the policy environment: Efforts to create data infra-

structures and instructionally focused data systems.
The examples of emerging practices below highlight attempts to make
teaching and learning the central focus of data-informed leadership.4 These
examples are by no means exhaustive; they simply offer illustrations of what
is happening across the country. Understandably, there is little empirical
research to date on many of the more recent experiments aimed at promoting
or bolstering data-informed leadership.

Efforts to Build Leaders’ Data Literacy and Expertise

Not all leaders exhibit the same degree of data literacy. While numerous
scholars (e.g., Dembosky, Pane, Barney, & Christina, 2006; Earl & Fullan,
2003; Wayman & Stringfield, 2006) note the role that educators’ expertise
plays in using data to inform action, educators’ ability to interpret and apply
data has been described as “woefully inadequate and sometimes very wrong”
in some instances (Earl & Katz, 2002, p. 1,013). For many leaders, becoming

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 23
 data literate means developing new capacities for using data effectively. While
there have often been modest attempts to boost educational leaders’ ability to
understand data tables, interpret statistics, and present quantitative informa-
tion about performance more effectively, attempts to develop a deeper level of
data literacy are seldom reported in the literature.
One state is taking several steps in this direction. New Mexico’s Office
of Educational Accountability has developed an initiative based on the prem-
ise that educational leaders need skills in “accountability literacy” to use data
wisely to support students. To be literate in accountability, leaders need to be
competent in a number of areas, including having the ability to interpret data;
negotiate support for education in political, professional, and community set-
tings; and understand what data can and cannot tell about students. This
kind of deep understanding about appropriate data use goes far beyond the
ability to use technology tools; it “requires not only capturing and organizing
ideas but also turning the information into meaningful actions” (Senge, 1999,
as cited in Earl & Katz, 2002, p. 1,005). The efforts of educators at Canyon
View High School provide an example of this type of interpretative process
that led to meaningful action. Beginning with questions related to potential
causes of student performance, teachers and leaders explored relationships
captured by data and ultimately targeted efforts at the elementary level. As
in this case, leaders who understand how data, properly interpreted, fit into a
larger picture of leadership practice will likely be better equipped to leverage
such tools in the service of learning improvement.
Data literacy also includes the use of data with various stakeholders,
such as school staff, the media, and parents. Often, these contexts call for
different kinds of data-informed leadership. For example, leaders may frame
conversations about data differently with teachers seeking to identify instruc-
tional gaps than with community leaders interested in tracking the schools’
progress over time. Several states and districts, including New Mexico, are
engaging in targeted efforts to support leaders’ development of data literacy
for this range of potential uses.

Efforts to Develop and Sustain Cultures of Inquiry

The development of leaders’ data literacy does not by itself create a milieu in
which leaders or other educators make it a common practice to engage in such
use. As highlighted in our framework, a critical aspect of data-informed lead-

24 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 ership is the ability to prompt, support, and sustain conditions that encourage
other staff, along with the leaders, to turn to data, ask questions of the data,
reflect on the data’s meanings, and take action that references the data. This
stream of activity flourishes when an organizational culture has evolved that
encourages inquiry into problems of practice. Whether using school improve-
ment planning tools (e.g., Boudett, City, & Murnane, 2005; Celio & Harvey,
2005) or high-tech data warehouses, leaders in such cultures make it pos-
sible or even attractive to engage others in cycles of inquiry that maximize
the potential benefits of such tools. Such cultures develop over time through
repeated activity by many individuals, but data-oriented leadership is often a
“driving force” behind data use (Supovitz & Klein, 2003). In such instances,
leaders generally work from a theory of action that gives data a central place
in planning and problem solving. Consider what is taking place in the follow-
ing middle school:

A middle school uses biweekly, two-hour faculty study groups to

examine samples of student work for evidence of learning prog-
ress and areas of need, and to identify future instructional steps
to take with particular students or groups of students. In these
sessions, staff are learning about high-quality student work as
well as planning ways to make that happen in classrooms. Noth-
ing is allowed to interfere with this standing commitment. Each
study group posts a public record of their work, to which the
principal provides feedback, questions, and affirmation. In addi-
tion, study groups report their progress and evidence of student
learning improvement to the whole staff at staff meetings. Student
achievement is steadily increasing. (Knapp et al., 2003, p. 39)

Something is happening in this school that makes data common currency

in efforts to improve teaching practice. The school leaders have orchestrated
an ongoing, schoolwide conversation about the improvement of practice and
engaged the staff in considering systematically local data sources (e.g., stu-
dent work samples) that can inform their work. In so doing, the process of
inquiring into school practices presumes a kind of collective sharing of lead-
ership responsibility for the matter, punctuated by an internal accountability
mechanism (sharing study group results with the whole faculty). The ability

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 25
 of such a process to be helpful depends on the staff’s willingness and skill,
though in this instance both those conditions appear to pertain.
Leadership for creating and supporting cultures of inquiry does not
necessarily depend upon a single person such as a school principal, even
though principals are in a strong position to set the basic conditions support-
ing data use. Positions such as data coaches, often filled by teacher leaders,
have been established in some schools and districts to help build capacity for
using data to inform practice. Data coaches of this sort engage in a range of
inquiry-based activities such as helping teachers understand their students’
strengths and weaknesses and identifying instructional strategies, structures,
programs, or curricula to address identified needs (Killion & Harrison, 2005).
The coaches also “frequently facilitate data dialogues with teams of teachers.
Coaches often work with building administrators to identify necessary data
to examine ways to display the data so the analysis process with teachers is
effective and efficient” (Killion & Harrison, 2005, p.1).
Coaches may sometimes be put in place with a specific support mis-
sion, as in the case of the 76 regional “value-added specialists” hired by the
state of Ohio to help local educators understand and make use of the value-
added student assessment information that is becoming available in that state.
Elsewhere, coaches concentrate on a broader role, as in the Bay Area School
Reform Collaborative (BASRC), where “reform coaches” focus on helping
educators use data as part of a continuous data-based improvement process.
In this process, the cycle of inquiry, educators identify a high-priority student
achievement problem; pose questions about the causes of the problem; imple-
ment strategies to address the causes; and then analyze data to determine
the effectiveness of their strategy (McLaughlin & Talbert, 2002). Data use is
infused throughout the inquiry process, and as such it becomes an unavoid-
able part of how things are done in the school, as one high school teacher
involved with BASRC reform efforts noted:

Data in and of itself isn’t useful. It’s what you do with it. Before,
we had data. Probably we could have guessed that a lot of those
things were the case. But once you formalize it, that implies that
you have to do something. (pp. 5–11)

26 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 Leaders use different kinds of devices for establishing a culture of
inquiry. On one end of a continuum, the process may be informal, as in New
Mexico’s Aztec Municipal School District—the superintendent meets quarterly
with principals to discuss the following questions: How is business going? How
do you know? (L. Paul, personal communication, Dec. 6, 2005). Although
very simple, these questions allow for both flexibility and accountability, rather
than a standardization of effort. The structure of quarterly meetings ensures
that leaders are using more than annual test data to assess performance. A
parallel process, in place at the state educational agency (SEA) in New York,
required heads of the SEA’s departments to engage in a formal review on a
regular basis.

At the Chief State School Officer’s insistence, and resonating well

with business interests and other constituents, there is a strong
emphasis on data to demonstrate how school and system perfor-
mance measures up to the state’s strategic goals. This rhetoric is
regularly put into practice within the agency, as well as through-
out the state’s educational system. As the Chief put it, “Two and
a half years ago, it was not possible for superintendents in the
big urban districts to be in a room and have the data shared pub-
licly. Now it’s commonplace. Back about four weeks ago, I had
them all together and we looked at the data in detail for about
five hours and then together sketched a strategy, a very powerful
simple strategy to close the gap …”

This emphasis on data as a basis for practice extends to the state

education bureaucracy itself, which could be a predictable obsta-
cle to a comprehensive, multi-faceted reform strategy. Just the
opposite has happened. As a result, the state education agency
has become a fairly integrated, unified entity, pursuing the imple-
mentation of a teacher quality improvement initiative in an effi-
cient and single-minded manner. … In contrast to what took
place under the previous Chief, performance in each unit within
the agency is now reviewed quarterly in relation to the agency’s
Strategic Plan. Unit staff assemble and present data to the Chief
that sheds light on their work in relation to strategic goals. These
reviews engender some fear and trepidation; Agency staff are

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 27
 expected to jump—and they appear to do just that. The data-
based flavor of the agency’s work under this Chief—an explicit
feature of the state’s Strategic Plan—appears in various places,
including the otherwise empty main hallway of the agency’s main
office building, which now displays numerous large graphs and
other representations of performance indicators (all going up, by
the way). Some staff report that morale at the agency is also on
the way up, under the current Chief’s tenure.5

In this instance, data-informed leadership has played a central role in

building a culture within the state agency, and to an extent within the larger
state system, that connects agency action to strategic goals, both as a matter
of accountability and as an expectation of daily practice. While in some ways
uncomfortable, the exercise appears to provide direction for the agency’s
efforts at the same time that it affects the interaction between the state and
districts. Here, and elsewhere, data-informed leadership may also be creat-
ing a climate of fearful compliance as much as a culture of inquiry. Therein
lies another set of issues connected to data-informed leadership, concerning
cycles of inquiry that encourage participants to take ownership for a learn-
ing improvement agenda and generate or use information accordingly (as in
the middle school example above), as opposed to assuming a more defensive
posture in response to external accountability pressures.

Engaging in Data-Informed Inquiry for Planning, Accountability,

and Performance Tracking
What we know about actual use of data in states, districts, and schools comes
from a small but growing body of scholarly research (e.g., Booher-Jennings,
2005; Herman & Gribbons, 2001; Ingram et al., 2004; Kerr, Marsh, Ikemoto
& Barney, 2006; Massell, 2001; Wayman & Stringfield, 2006; Young, 2006).
Informed by this research and reports from practitioners, three categories of
data use in states, districts, and schools are especially common:
• School improvement planning.

• Responses to external accountability requirements.

• Public tracking of educational performance.

28 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 School improvement planning. School improvement planning, a cen-
tral focus of much writing in the field (Bernhardt, 1998; Holcomb, 1999;
Streifer, 2002), aligns not only with the current accountability context in U.S.
schools, but also reflects longer term efforts associated with the standards-
based reform movement and models of “continuous improvement” often
adopted from the business community (e.g., balanced scorecard, totally qual-
ity management).
Data can inform school improvement planning in several ways, by pro-
viding planners with a means to set organizational performance goals and
measure progress toward meeting them. Schmoker (1996) advocates the use
of data as a measure of school performance, but asserts that data analysis
must be a collaborative activity focused on clear, measurable, attainable goals.
Bernhardt’s (1998) conceptual framework identifies several types of data vari-
ables—demographics, school processes, perceptions, and student outcomes—
and a variety of sources to guide educators in their efforts to isolate problem
areas, set goals for improvement, and measure progress toward meeting the
goals. Both of these writers provide images of how data are used in the con-
text of school improvement planning, and how these activities connect to
the inquiry process, that take various forms in accounts of actual data use in
school improvement planning (e.g., Boudett, City, & Murnane, 2005; Cop-
land, 2000; Ingram et al., 2004).
Response to external accountability requirements. Clearly connected
to school improvement planning, a second category of data use reflects a
response to specific requirements, rather than voluntary participation in
reform activities. For example, schools participating in the federally funded
Reading First literacy program must provide assessment information three
times per year. Many schools use a standard assessment called Dynamic
Indicators of Basic Early Literacy Skills (DIBELS) to track individual student
progress in grades K–3 throughout the year. In some cases, schools showing
improvement receive additional money while those unable to meet identified
targets are at risk of losing funds.
There are numerous examples of data use in the context of increasing
external accountability requirements (Booher-Jennings, 2005; Lemons, Lus-
chei, & Siskin, 2003; Massell, 2001; Spillane et al., 2002). One investigation
of high schools in Ohio and Texas found a variety of ways that schools used
assessment data to respond to accountability pressures (Lemons et al., 2003).

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 29
 But given that the responsibility of interpreting and using assessment data to
guide school improvement in such high-stakes environments tended to fall
primarily on the principal, data use in these schools depended on the inter-
ests and skills of the individual cast in this role. The high-stakes environment
could also lead to unintended, and possibly counterproductive uses, as in a
different investigation, this time, concerning one elementary school in Texas
(Booher-Jennings, 2005). In this instance, principal(s) and teachers, rather
than helping those students who tended to score lowest on the state assess-
ments and thus needed the most assistance, directed their instructional efforts
toward “bubble kids”—those students whom they predicted were most likely
to score closest to the passing score on the tests.
Public tracking of educational performance. Data are also used as
means for comparing and ranking performance in a variety of ways (e.g.,
growth over time, among various groups of students, between particular
schools). Web-based tracking systems have been set up on a national basis,
such as Just for the Kids, sponsored by the National Center for Educational
Accountability, and School Matters, sponsored by Standard & Poor’s.6 Both
offer performance profiles of states, districts, and schools, all available to the
public via the Internet. Many states and districts also offer Web-based com-
parisons. For example, the Seattle Public Schools Research, Evaluation, and
Assessment site provides a variety of information on individual schools,7 rang-
ing from state test scores and value-added data to district writing scores and
school demographic reports. Although the kinds of comparisons provided by
these and many other Web sites are not new to educators, the public availabil-
ity of this information creates both opportunities and challenges for leaders.
While some leaders use tools like those listed above to “prove” the quality of
their school or district, others encounter new challenges given the sometimes
grim picture portrayed in charts, graphs, tables, and other data displays. In
some cases, the data are not accurate, or they may be misinterpreted by users
who may not understand the meaning behind the charts, graphs, and various
statistics (Earl, 1995).

Efforts to Create Data Infrastructures and Data Systems

State education systems have embraced data use increasingly, building on an
older tradition within federal categorical programs, which made extensive use
of testing data and other indicators to track and justify interventions (e.g., test-

30 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
ing in the Title I program). Over the years, states, districts and schools have
engaged in routine data collection of various kinds, though these informational
resources have generally not been used extensively for making decisions about
the improvement of teaching and learning, but rather have used these data
sources for compliance monitoring, tracking funds, and generating descriptive
profiles of educational programs in the state. It has been a natural outgrowth
of the state standards-based reform movement, in particular, as well as other
trends mentioned above, to engage in or encourage the following:
• Developing better data infrastructures, i.e., merging silo systems.

• Creating district and school profiles, as an informational device and

also part of an accountability system available to the public through
Web sites.

• Using data as a means for comparison, i.e., among schools and districts.

• Using data to inform instructional improvement.

• Creating partnerships to enhance technical and procedural support

for the use of data.

• Responding to federal reporting and performance demands.

Much of the activity at the state and district level revolves around the
creation of technical infrastructures, e.g., student information systems, Web
sites and data warehouses, and paying more attention to the usability of their
data systems (Snow-Renner & Torrence, 2002). However, according to a 2005
survey conducted by the National Center for Educational Accountability, no
state has all ten “essential elements of a robust longitudinal data system.8 In
fact, only 36 states reported having a unique student identifier that allows for
tracking student information from kindergarten through 12th grade.9
These systems contain various data elements—often collected annu-
ally—ranging from demographic information and student performance mea-
sures to data about instructional programs (Armstrong & Anthes, 2001).
States and districts typically issue periodic reports in a standardized format
that summarizes descriptively what is in the database—for example, test
score patterns by grade, student demographics, and so on. Besides providing
a basic descriptive record of important aspects of the educational system’s
functioning, these databases offer a possible tool for decision makers, though

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 31
 more often than not this may be wishful thinking, as state databases are
not always easy to access, are not set up for queries, may not be particularly
timely, and the data are not always accurate. Data “cleaning,” e.g., correct-
ing errors, omissions, and redundancies, is a critical process that happens as
states and districts implement new data systems (Mieles & Foley, 2005; Way-
man, Stringfield, & Yakimowski, 2004).
Although NCLB does not require a state or district to implement a tech-
nology solution for data management, most systems have taken advantage of
technology tools for data collection, management, and distribution. Accord-
ing to Wayman (2005), these systems fall into three general categories: stu-
dent information systems, assessment systems, and data warehousing systems.
Although related, each system performs different, nearly mutually exclusive
functions, ranging from real-time tracking of attendance and performance
data to dynamic systems designed for data manipulation and analysis. For
example, The Kentucky Department of Education, the Education Professional
Standards Board, and the Council on Postsecondary Education are all working
together to develop the MAX Education Data Warehouse.10 The system offers
leaders, teachers, parents, and policymakers various ways to access informa-
tion about school and district profiles, financial data, and assessment results.
School districts are also investing in their own data infrastructures,
as in Jefferson County (KY) Public Schools (JCPS), which recently invested
in a major upgrade to their data warehouse, growing from an initial eight
gigabytes to the new system with a 1.5 terabyte capacity. According to the
vendor,

[s]chool systems of JCPS’s size are not much different than

Global 1,000 businesses. In such large enterprises, optimum per-
formance and productivity are critical to meeting strategic goals
and satisfying stakeholders. The warehouse supports more than
1,000 discrete users, generating as many as 500 queries per day.11
(News Release, 2006)

Currently at least 25 to 30 different corporate vendors offer data

warehouse solutions to K–12 systems, including such major market players
as TetraData, EDmin, Cognos, and SchoolNet. Common features include
traditional storage capabilities, customized queries, visual displays of infor-

32 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 mation, and interactive statistical analyses of data (Sarmiento, 2004; Stein,
2003; Wayman, Stringfield, & Yakimowski, 2004). Although these systems
are becoming increasingly more robust and user-friendly, educators are the
ones who ultimately determine the value of these tools in supporting and
improving student learning.
While student information systems and data warehouses, like those
described above, hold some utility for school leaders and teachers, other sys-
tems have been created to more specifically target instructional decision mak-
ing. Project SOAR (Schools’ On-line Achievement Reports) in Ohio and Geor-
gia’s “Balanced Scorecard” system are two examples (Project SOAR, 2005).
The former, Project SOAR, claims to provide users with easy-to-understand
Web-based data charts that provide school comparisons and measurements
of student progress over time, in an effort to get at the “value-added” effects
of schooling on students learning. While there is debate about the technical
issues associated with value-added analyses (McCaffrey, Koretz, Lockwood,
& Hamilton, 2004) this system puts a relatively simple and accessible form of
this analysis at the fingertips of a large number of users. Participants can view
district, building, grade, and student-level data.
Other state supported data-systems try to provide the user with a
greater variety of data that could be useful to the decisions leaders have to
make, as in Georgia’s “Balanced Scorecard (BSC) approach to data-informed
leadership practice.12 As opposed to a single focus on lagging indicators such
as student achievement data, the BSC focuses on accurate, relevant, lead-
ing indicators of success that drive continuous improvement. As an official
involved in creating this system puts it, “Like business enterprises, schools
are not stagnant organizations …. The same BSC benefits of alignment, com-
munication, and strategic planning will be realized by the managers in the
education sector” (Duffy, 2005, p. 4). The BSC includes specific performance
objectives, performance measures, the units of measure, and rating informa-
tion. Under this arrangement, districts are also encouraged to determine, at
the local level, what measures matter most for inclusion in their “scorecard,”
leading to some variability in the kinds of data that are included by each
local district. In addition to this tool, the Georgia Department of Education
offers professional development for school leaders and leadership teams on
data use, targeting eight roles of leadership,13 including what it means to be a
“data analysis leader.” In addition, Georgia’s Leadership Institute for School

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 33
 Improvement specifically sponsors BSC training and supports professional
learning for district leadership teams.
Locally developed data infrastructures may also seek to inform class-
room instruction directly. For example, Boston Public Schools used the Fast-
Track system along with the MyBPS Web portal to support one of their six
“essential” strategies for school improvement:14 to examine student work and
data to drive instruction and professional development. Together, these sys-
tems provide educators with classroom-level analysis of various data elements,
including test scores, grades, attendance, and schedule information. Teachers
can access the text of every question on the state exam as well as individual
student responses. Principals and other leaders can access data associated
with customized groups of students not normally associated with traditional
data reports (Mieles & Foley, 2005).

Issues Arising in Patterns of Data Access and Use

A number of the examples reviewed above, and much of the case study research,
concentrates on relatively atypical cases in which leaders and others are tak-
ing active steps to make the use of data a regular part of educational practice.
In this sense, these cases may offer images of possibility, but they may also
gloss over problematic aspects of basing leadership more squarely on system-
atic data, as the following case suggests.
Shifting accountability systems in one district. While some districts
may call for, and in various ways support, the kinds of inquiry-oriented prac-
tice noted earlier (see for example, examples from Colorado in Massell, 2001),
the dynamics of data-informed leadership at the district level are likely to be
more complicated and often problematic. The case of Midvale School District
(pseudonym) and its evolving accountability policies throughout the decade
from 1995 – 2004 is instructive (Burch, 2005). A large urban district serving
a population exceeding 100,000 students, Midvale engaged in a first phase
of accountability reform that featured a low-stakes “balanced” accountabil-
ity system with multiple measures including classroom-based assessment,
systemwide performance measures, and standardized norm-referenced tests.
Aided by an external assessment firm, the district developed and fine-tuned
the system and provided systematic training and supports for teachers to help
them make use of the array of assessment information that could inform
their teaching. This phase of activity represented a significant step forwards

34 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
toward a richer environment of data, which could inform efforts to educate
the students.
The picture changed in the year 2000 in anticipation of NCLB, as dis-
trict leaders shifted gears toward a more high-stakes assessment and account-
ability system that emphasized primarily norm-referenced tests. To manage
the increasing data demands, the district contracted with another outside firm,
subsequently subsumed within a larger national firm, that created sophisti-
cated new software for the district’s performance information system and
new rules affecting who could access the data, while offering relatively little
useful technical assistance. The upshot of these changes balanced the follow-
ing signs of progress towards data-informed leadership with developments
that worked against a goal of improving the quality of instruction for all
(Burch, 2005).
• Greater use of data by district-level administrators, in their efforts
to address instructional issues and in their interactions with specific
schools.

• Diminished access of teachers to the new data system, and greater

inequities in data access and use by school staff, reflecting restrictions
on teachers’ access to the system, and considerable variation in school
administrators’ abilities to access and use data.

• More consistent attention to results in instructional improvement

planning.

• More limited measures for representing the quality of teaching and

learning.

• Greater investment of district resources in data systems and related

management tools and services, primarily through contracts to out-
side vendors.

• Less direct and personal technical assistance and support to various

users such as teachers and school staff.
The reasons for the mixed results are complex, but the case serves to
underscore some important dimensions of the data-informed leadership story,
among them, the power of external forces—both the strict accountability
requirements and the nature of the outside vendors—to influence what data

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 35
 were attended to, how they were used, and by whom. Furthermore, the case
vignette underscores the delicate dynamic across levels within the district (dis-
trict, school, and classroom), indicating that enhancing the use of data for
users at one level does not necessarily help users at another level, and hints at
the crucial role of support for data use. Finally, this case begs further ques-
tions about how, given the unevenness in school capacity and preferences
within most or all districts, one can stimulate and support equitable data-
informed leadership across a system of schools.
What complicates or limits leaders’ use of data. The Midvale case
points to a number of conditions and responses that are likely to discourage
continual data use. Consider the following: fears about the consequences of
systematic data; the belief that data representing what matters most in edu-
cation cannot be secured; unwillingness to expose one’s ignorance about the
manipulation and interpretation of data; lack of interest in the prospect of
changes in practice that might be suggested by data; limited knowledge of
what is possible to do with data; and lack of resources to invest in data-based
inquiry into problems of practice. These constitute major barriers to data-
informed leadership, and they are not simple to overcome, in part, because
there are sometimes sensible reasons for these responses to the prospect of
data-informed leadership.
Take, for example, educators’ fears concerning how data about their
practice or performance may be used. There are enough examples in most
educators’ working experience of information “being used against them and
their colleagues” (in fact or in perception) to prompt a healthy mistrust of
data use, especially in the context of high-stakes accountability (Heritage
& Yeagley, 2005). It is no wonder that in some instances of sustained data
use within high-performing districts, leaders went to considerable lengths to
“make it safe” for participants to consider data and their implications (Togn-
eri & Anderson, 2003). Wayman & Stringfield (2006) described leaders who
used a “non-threatening triangulation of data” approach, which relied on
using multiple sources to ensure that teachers felt supported and empowered
by data rather than threatened. And while leaders were able to prevail in these
cases, elsewhere the struggle is not so successful. And in situations where
accountability systems do lead to punitive action, there are clearly good rea-
sons to be cautious about how one represents one’s performance or practice.

36 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
 Lack of capacity to engage in data-informed inquiry can work in simi-
lar ways to suppress the amount of data-informed leadership. Despite pres-
sure to increase test scores, schools seem to vary considerably in their levels
and types of assessment data use (Massell, 2001). Researchers have observed
that educators are generally not skilled in interpreting systematically col-
lected data (Baker & Linn, 2004; Bernhardt, 1998; Heritage & Chen, 2005;
Schmoker, 1996). This lack of technical skills likely hinders most educators’
abilities to both physically work with data and make valid interpretations of
data. Ingram et al. (2004) found that teachers have their own metric for judg-
ing teaching effectiveness. Many choose to base their decisions on experience,
intuition, and anecdote rather than systematically collected data. It is plau-
sible to believe leaders approach data use in similar ways.
But even with increasing levels of data-informed leadership apparent at
school and district level, especially strong evidence concerning the capacity
of such activities to promote instructional improvement is scarce (Honig &
Coburn, 2005; Kerr et al., 2006). What evidence there is of such a connection
resides largely in the presence, in relatively small numbers of “high perform-
ing” schools or districts, of well-established routines for the use of data in
decision making. Reports of “instructionally effective school districts” (Mur-
phy & Hallinger, 1988) or “beating the odds” districts (e.g., Snipes, Doolittle,
& Herlihy, 2002) are among those that display such correlations. While it is
not possible from such research to isolate the effect of data-based decision
making on learning outcomes, it is clearly plausible that, as part of a syn-
drome of learning-focused leadership activity, this facet of leadership makes a
contribution to the improvement of performance.
One thing is crystal clear from the research to date: where it occurs, data-
informed leadership is a direct reflection of aggressive, persistent attempts by
leaders to create conditions in which this way of approaching the business of
schooling can flourish. Several scholars put it this way:

Marshalling the school’s community to a collective sense of pur-

pose about improving student learning, accepting that data use
can and will improve learning, aligning data use to school plan-
ning and decisionmaking processes, and creating time for data
analysis are key elements of leadership in creating a culture for
data use. (Heritage &Yeagley, 2005, p. 335)

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 37
 Others studying this phenomenon have come to similar conclusions
about the centrality of leadership in bringing data into play (e.g., Kerr et al.,
2006; Supovitz & Klein, 2003; Mason, 2002). Absent this kind of advocacy
for the use of data, the natural forces and resistance are generally too strong
to make data-informed leadership a feature of everyday practice.
Whether or not leaders are sharply attuned to the possible benefits of
data, the drumbeat signaling the need for data has become more insistent in
recent years, and is likely to continue. There is an “increasingly institutional-
ized assumption that data can and should be used to give credence to one’s
stated purposes, proposals, problems, and solutions” (Young, 2006, p.2).
And in a world of persistent accountability expectations, leaders are increas-
ingly in the position of “learning to live with data and like it” (Earl & Katz,
2002, p. 2). This assumption about data use drives many of the initiatives
described in this paper. It is in this context that data-informed leadership
may be a powerful bridge between desired improvement and educators’ cur-
rent capacity, but it also raises important issues concerning the way policy
environments promote and discourage activities aimed at improvement. The
persistent demand for performance data in the context of accountability can
limit, as well as enable, inquiry into local educational problems and how to
solve them.

38 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
Unanswered Questions and Enduring Dilemmas

The concept of data-informed leadership encompasses a wide range of issues

and raises numerous questions that will need to be pursued, both by those
who will be creating new approaches to the challenges this issue area presents
and by researchers who wish to study it. Currently, the leadership literature
does not provide a well-grounded conception of data literacy in the context
of school, district, and state systems. In addition, this paper sets the stage to
broaden and add cautions to the current call to use student performance data
to drive decisions. Rather, our hope is to reframe the conversation among
scholars and practitioners towards conceptions of how data inform leadership
and professional practice, especially in the current accountability context.

Unanswered Questions
The framework and examples reviewed above suggest questions that beg for
more complete answers from continued experimentation and related research.
While there are many such questions, the following six, related to key ele-
ments in the framework, are arguably very important to answer.
A first question arises in relation to one of the main anchors for data-
informed leadership: leaders’ expertise in accessing, generating, managing,
interpreting, and acting on data. This expertise concerns not only what a
leader or leadership team might do with the data, but also what they might
do to facilitate the process of using data as a primary reference point for their
planning and practice.
1. The development of data literacy among leaders.

a. What does data literacy consist of for practicing leaders in

schools, districts, and state agencies? What knowledge, skills,
dispositions, and beliefs enable and motivate leaders to bring
data to bear on the challenges of improving teaching and learn-
ing? What balance of technical know-how, political savvy, and
cultural understanding makes the leader fully literate in the
practice of data-informed leadership?

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 39
 b. How do leaders acquire data literacy? In what ways do events
or conditions in the community, larger policy environment, or
organizational setting support leaders’ acquisition of data liter-
acy—and explain differential acquisition of data literacy among
leaders? Where do leaders go to get help, and how can all lead-
ers be assured of the help they need, regardless of location, prior
skills, etc.?

A second question zeroes in on cultures and cycles of inquiry, and what it

may take to establish such cultures when policy reform pressures are acute:
2. Building robust cultures of inquiry in the context of high account-
ability. What conditions and support strategies are most likely to
build organizational cultures that support inquiry and data use, es-
pecially in situations where accountability pressures are most keenly
felt, e.g., low performing schools or districts? Conversely, how does
continued engagement in data-based inquiry influence the organiza-
tion’s culture over time? How can it enable productive responses to
accountability requirements?

A third and fourth question concerns activity in the policy environ-

ment that seeks to bolster data-informed practice through the creation of data
infrastructures, while pushing leadership practice to focus single-mindedly on
student achievement. The questions acknowledge that states and other enti-
ties are making increasing investments in support of data-informed leadership,
but these investments are not always informed by the perspectives of users.
3. The usefulness of increased investment in resources and supports
for data-informed leadership. To what extent, if at all, does the
combination of state and local policies and investments enhance
leaders’ access to data they consider useful? For what range of deci-
sions do leaders consider the data useful, and why?

A fourth question is prompted by the intense focus by current federal

and state policy on student achievement, and addresses the clear tempta-
tion for leaders under many accountability systems to take student achieve-
ment levels as a sole and unambiguous measure of the worth of anything—a
school program, a particular leader’s tenure in the school or district—without

40 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
attempting to understand what the achievement numbers reflect in a more
nuanced way. Hence the question:
4. Responsible treatment of student performance data. How can lead-
ers and their audiences be helped to interpret student performance
data in light of other relevant information, e.g., student character-
istics, community conditions, teaching, and the teacher workforce,
so that premature conclusions about the value of programs or per-
sonnel are avoided?

Two final questions concern the ultimate impact of this activity and its
links to other aspects of leadership. Regarding the underlying concern that
data-informed leadership contribute in some demonstrable way to improving
teaching and learning, one can ask:
5. The impact of data-informed leadership on teaching and learning.
What evidence suggests a direct connection between the practice of
data-informed leadership and (a) the (re)allocation of resources to
support specific instructional improvements; (b) teachers’ attempts
to engage in these forms of improved practice, and (c) students’
learning gains that are attributable to these practices?

Finally, recognizing that data-informed leadership is not the only aspect

of a leaders’ work that matters, the issue arises about how to integrate data-
informed leadership with other leadership activities aimed at improving teach-
ing and learning:
6. Integration of data-informed leadership with other aspects of learn-
ing-focused leadership. In what ways can data-informed leadership
be effectively aligned with other aspects of a leadership approach
that places priority on the improvement of teaching and learning?

Developing answers to these questions will require various approaches,

some through continued experimentation with data-informed leadership
arrangements that include a feedback loop or other means to learn from the
experience. Some can be addressed by formal research studies.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 41
 Enduring Dilemmas and Ideological Tensions
As educators and scholars seek answers to the questions just posed, it is well
to keep in mind that some things about the use of data in educational leader-
ship have not changed, and are unlikely to in the future. These matters reflect
dilemmas or tensions that are always present in the act of using data within
an organized setting, and cannot be eliminated by better technical solutions,
more training, greater commitment to data use, etc. We see three such ten-
sions among (1) state (or national) policy and local response, (2) the need for
immediate feedback to inform practice current and longer-term documenta-
tion of performance, and (3) what is technically desirable and what is politi-
cally or culturally feasible.
The tension between state assessment policy and local response, in
the context of shifting state politics and uncertain funding. To the extent
that large-scale assessment data are intended to figure prominently in district
and schools’ data-informed leadership, the role of the state in shaping coher-
ent assessment policy and building capacity to use data is critical. However,
the vagaries of state politics and funding may challenge states’ abilities to
fund and provide statewide systems to support data-informed leadership or
serve to obscure those aspects of school or district activity that most centrally
concern sustainable improvements in teaching and learning. How are local
districts to respond to changing policy from the state? Furthermore, whatever
the state does and no matter how coherent its approach to reform may be, the
fact that state reforms are externally imposed means that they will bump up
against the local cultures of a district or school, which have much to do with
whether and how reforms are implemented (Cuban, 1998). To the extent that
data-informed leadership is a top-down mandate or is externally driven—or
is seen as a tool of external control (as is often the case with accountability
systems)—it will always be subject to the “bottom-up” reinterpretation, and
even subversion, by local educators who do not wish to have their autonomy
compromised.
The tension between the need for immediate feedback to inform cur-
rent practice and the longer-term documentation of performance. Inevitably,
data can be used to answer questions about the merit or worth of an activity,
thereby helping certain audiences pass judgment on it, or it can be used to
diagnose the ongoing qualities of work and areas in which the work can be
altered, thereby contributing to its improvement. These summative and for-

42 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
mative purposes are not totally unrelated but they lead to different kinds of
actions or decisions, and they can easily get in each other’s way, especially if
users do not understand the underlying purposes for each kind of assessment.
Efforts to craft summative judgments, for example, about the nature of stu-
dent achievement from annual state assessments, particularly when there are
high stakes consequences attached to poor performance, may drive behav-
iors in the system to improve scores through whatever means possible: heavy
emphasis on test prep strategies to the exclusion of other existing curricula;
targeted teaching to students who are “on the bubble” to the exclusion of
those far below grade level, etc. Formative uses of such data that help districts
and schools to determine particular areas of instructional need may be lost
or downplayed in the process. On the other hand, relying solely on forma-
tive data that is critical for instructional decision making (National Research
Council, 2000) will not give leaders a systemwide perspective on achievement,
especially in terms of performance trends over time.
The tension between what is technically desirable and what is politi-
cally or culturally possible. Recent research on schools which practice data-
based decision making as part of their commitment to continuous improve-
ment and organizational learning identifies barriers to data use that stem
from natural and enduring dilemmas in the way technically advanced prac-
tices enter into the politics and culture of schools (Ingram, Louis, & Schro-
eder, 2004). As Table 3 emphasizes, these challenges are not easily resolved,
and may ultimately not be fully resolvable.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 43
 Table 3. Cultural, Technical, and Political Challenges to Data-Based Decision Making in Schools
(based on Ingram et al., 2004)

• Cultural challenges

1. Teachers have their own metric for judging teaching effectiveness.

2. Many teachers and administrators choose to base their decisions on experience, intuition, and
anecdote rather than systematically collected data.
3. Consensus about which outcomes are most important, and what data are most meaningful is lacking.
4. Some teachers disassociate their own performance from that of students. Entrenched cultural norms
of teacher isolation and unwillingness to measure or define their teaching effectiveness in terms of
student outcomes works against data-based decision making.
• Technical challenges
5. Data that teachers want—“really important outcomes”—are rarely available and hard to measure. It
is also difficult for many educators to see the connection between cause and effect in the data that
they have.
6. Schools rarely provide the time needed to collect and analyze data.

• Political challenges
7. Data have often been used politically, leading to mistrust of data and data avoidance. The perception
that others misuse data will make educators reluctant to trust data presented by others and to use
data themselves.

Beneath the tensions just described, and throughout the use of data in
decision making are ideological matters: competing belief systems about what
is desirable and about how the world works or could be improved. These
belief systems come to the fore in any large-scale system and the upshot
reflects whichever set of beliefs prevails, and for how long. Also, because
these beliefs rest on values more than empirical evidence, they are unlikely
to disappear—that is, be dismissed by evidence. Consequently they and the
competition among them will always be present.
For example, one continuum of ideology along which observers may
stake positions and oppose each other might be views on the strength of
accountability or the use of standardized testing as a lever for school change.
At one end might stand advocates of strong accountability who see a need for
incentives and sanctions (Hess, 2003). At the other end might stand writers
who view strong accountability systems with considerable skepticism, on the
grounds that such systems can do harm to certain groups of students (McNeil,
2000). The two ends of this ideological spectrum view data and how the data
should be used quite differently. Educators may also advocate or oppose data-

44 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
informed leadership based on philosophical views of good ways to do educa-
tional business. Some educators may oppose data, especially if it is quantita-
tive, on the ground that it threatens or misrepresents relationships they view
as central to the educational process. This conflict may mirror conflicts in
academic circles between quantitative and qualitative modes of inquiry.
Finally, there are inherent and longstanding tensions in people’s assump-
tions about good teaching and the assessment of learning, rooted in behav-
iorist and constructivist theories. Lorrie Shepard (2000) may speak for many
educators in her description of standardized testing as “crypto-behaviorism.”
Many educators may view data-informed leadership as little more than an
extension of standardized testing. Other educators (e.g., Popham, 1987) see
value in inquiry based on different kinds of data, including standardized test
scores. No study or experimentation will fully bridge this divide. One quickly
reaches the limit of what data can help leaders understand or do when basic
differences in belief systems of this sort come into play.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 45
 References

Argyris, C. & Schon, D. A. (1978). Organizational learning: A theory of action perspective.

Reading, MA: Addison-Wesley.

Armstrong, J., & Anthes, K. (2001). How data can help: Putting information to work to
raise student achievement. American School Board Journal, 188(11), 38–41.

Baker, E. L., & Linn, R. L. (2004). Validity issues for accountability systems. In S. H.
Fuhrman & R. F. Elmore (Eds.), Redesigning accountability systems for education
(pp. 44–72). New York: Teachers College Press.

Bennett, N., Wise, C., Woods, P., & Harvey, J.(2003). Distributed leadership (full report).
Oxford, UK: National College for School Leadership.

Bernhardt, V. L. (1998). Data analysis for comprehensive schoolwide improvement. Larch-

mont, NY: Eye on Education.

Bernhardt, V. L. (2003). Using data to improve student achievement. Educational Leader-

ship, 60(5), 26–30.

Bernhardt, V. L. (2004). Continuous improvement: It takes more than test scores: Analyz-
ing state assessment results is only the beginning of effective data-driven decision
making. Leadership, 34, 16–19.

Bolman, L., & Deal, T. E. (1997). Reframing organizations: Artistry, choice, and leadership
(2nd Ed.). San Francisco: Jossey-Bass.

Booher-Jennings, J. (2005). Below the bubble: ‘Educational triage’ and the Texas account-
ability system. American Educational Research Journal, 42, 231–268.

Boudett, K. P., City, E. A., & Murnane, R. J. (2005). Data wise: A step-by-step guide to us-
ing assessment results to improve teaching and learning. Cambridge, MA: Harvard
Education Press.

Burch, P. (2005). The new educational privatization: Educational contracting and high-
stakes accountability. Teachers College Record (online), at http://www.tcrecord.org/
content.asp?contentid=12259. Retrieved August 16, 2006.

Celio, M. B., & Harvey, J. (2005). Buried treasure: Developing a management guide from
mountains of school data. Seattle, WA: Center on Reinventing Public Education,
University of Washington.

Coburn, C. E. (2006). District evidence use: An analysis of instructional decision making.

Paper presented at the American Educational Research Association, San Francisco,
April 7–11.

Coburn, C. E., & Talbert, J. E. (2006). Conceptions of evidence use in school districts:
Mapping the terrain. American Journal of Education 112(4), 469–495.

Colgan, C. (2004). Data on demand. American School Board Journal, 191(09), 36–39.
Washington, D.C.: National School Board Association.

46 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
Copland, M. (2000). Leadership of inquiry: Building and sustaining capacity for school
improvement. Educational Evaluation and Policy Analysis, 25, 375–395.

Copland, M. A. (2003). The Bay Area School Collaborative: Building the capacity to lead.
In Murphy, J., & Datnow, A. (Eds.), Leadership lessons from comprehensive school
reform (pp. 159–184). Thousand Oaks, CA: Corwin Press.

Council of Chief State School Officers. (2006). Validity threats: Detection and control prac-
tices for state and local educational officials. Alexandria VA: Author.

Cuban, L. (1998). How schools change reforms: redefining reform success and failure.
Teachers College Record, 99, 453–477.

Dembosky, J. W., Pane, J. F., Barney, H., & Christina, R. (2006). Data driven decisionmak-
ing in southwestern Pennsylvania school districts: RAND Corporation.

Duffy, A. (2005). Balanced scorecards for public schools: Accountability beyond test scores.
Atlanta: Georgia State University.

Earl, L. (1995). Moving from the political to the practical: A hard look at assessment and
accountability. Orbit, 26(2), 61–63.

Earl, L., & Fullan, M. (2003). Using data in leadership for learning. Cambridge Journal of
Education, 33(3), 383–394.

Earl, L., & Katz, S. (2002). Leading schools in a data-rich world. In K. Leithwood & P.
Hallinger (Eds.), Second international handbook of educational leadership and
administration. Dordrecht: Kluwer Academic Publishers.

Elmore, R. (2000). Building a new structure for school leadership. New York: The Albert
Shanker Institute.

Fullan, M. (1999). Change forces: The sequel. London: Falmer Press.

Fullan, M. (2001). Leading in a culture of change. San Francisco: Jossey-Bass.

Hallinger, P., & Heck, R. H. (1996). Reassessing the principal’s role in school effectiveness:
A review of empirical research, 1980–1995. Educational Administration Quarterly,
32(1), 5–44.

Halverson, R. R. (2003). Systems of practice: How leaders use artifacts to create profes-
sional community in schools. Education Policy Analysis Archives, 11(37).

Heritage, M., & Chen, E. (2005). Why data skills matter in school improvement. Phi Delta
Kappan, May.

Heritage, M. , & Yeagley, R. (2005). Data use and school improvement: Challenges and
prospects. In Herman, J. L., & Haertel, E. H. (Eds.), The uses and misuses of data
for educational accountability and improvement—104th yearbook for the National
Society of the Study of Education, Part II (pp, 320–339). Malden MA: Blackwell
Publishing.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 47
 Hermon, J., & Gribbons, B. (2001). Lessons learned in using data to support school
inquiry and continuous improvement: Final report to the Stuart Foundation. Los
Angeles: Center for the Study of Evaluation.

Hess, R. (2003). Refining or retreating? High-stakes accountability in the states. In Peter-

son & M. West (Eds.), No child left behind? The politics and practice of school
accountability (pp. 55–79). Washington, D.C.: The Brookings Institution.

Holcomb, E. (1999). Getting excited about data: How to combine people, passion, and
proof. Thousand Oaks, CA: Corwin Press.

Holly, P. J. (2003). Data-driven school improvement series. Princeton, NJ: Educational

Testing Service.

Honig, M. (2006). Building policy from practice: Implementation as organizational learn-

ing. In Author (Ed.), New directions in educational policy implementation: Con-
fronting complexity. Albany, NY: SUNY Press.

Honig, M., & Coburn, C. E. (2005). When districts use evidence for instructional improve-
ment: What do we know and where do we go from here? Urban Voices in Educa-
tion 6, 22–26.

Huber, J. (1991). Organizational learning: The contributing processes and the literatures.
Organization Science, 2(1), 88–115.

Ingram, D., Louis, K. S., & Schroeder, R. G. (2004). Accountability policies and teacher
decision making: Barriers to the use of data to improve practice. Teachers College
Record, 106, 1,258–1,287.

Knapp, M. S., Copland, M., & Talbert, J. E. (2003). Leading for learning: Reflective tools
for school and district leaders. Seattle, WA: University of Washington Center for the
Study of Teaching and Policy.

Kerr, K. A., Marsh, J. A., Ikemoto, G. S., & Barney, H. (2006). Strategies to promote data
use for instructional improvement: Actions, outcomes, and lessons from three urban
districts. American Journal of Education 112(4), 496–520.

Killion, J., & Harrison, C. (2005). Data coach. Austin, TX: National Staff Development
Council.

Leavitt, P., McDaniel, R., & Skogstad, E. (2004). Disaggregating data in schools: Leverag-
ing the information you have. Houston TX: APQC.

Leithwood, K., Aitkin, R., & Jantzi, D. (2001). Making schools smarter. Thousand Oaks,
CA: Corwin Press.

Lemons, R., Luschei, T. F., & Siskin, L. S. (2003). Leadership and the demands of
standards-based accountability. In M. Carnoy, R. Elmore & L. S. Siskin (Eds.), The
new accountability: High schools and high-stakes testing (pp. 99–127). New York:
Routledge-Falmer.

Marzano, R. J., Waters, T., & McNulty, B. A. (2005). School leadership that works. Alex-
andria VA: Association for Supervision & Curriculum Development.

48 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
Mason, M. (2002). Turning data into knowledge: Lessons from six Milwaukee public
schools. Madison, WI: Wisconsin Center for Education Research.

Massell, D. (2001). The theory and practice of using data to build capacity: State and local
strategies and their effects. In S. H. Fuhrman (Ed.), From the capitol to the class-
room: Standards-based reform in the states (pp. 148–169). Chicago: University of
Chicago.

McLaughlin, M., & Talbert, J. (2002). Bay Area School Reform Collaborative: Phase one
(1996–2001) evaluation. Stanford, CA: Stanford University, Center for Research on
the Context of Teaching.

McNeil, L. M. (2000). Contradictions of school reform: Educational costs of high-stakes

testing. New York: Routledge.

McCaffrey, D., Lockwood, J. R., Koretz, D., & Hamilton, L. (2004). Evaluating value-
added models for teacher accountability. Washington D.C.: Rand.

Mieles, T., & Foley, E. (2005). From data to decisions: Lessons from school districts using
data warehousing. Providence, RI: Annenberg.

Murphy, J., & Hallinger, P. (1988). The characteristics of instructionally effective school
districts. Journal of Educational Research, 81(3), 175–181.

National Research Council (U.S.). Committee on Developments in the Science of Learning

(2000). How people learn: Brain, mind, experience, and school. Washington, D.C.:
National Academy Press.

Popham, J. (1987). The merits of measurement-driven instruction. Phi Delta Kappan, 68(9),
679–682.

Portin, B., Beck, L., Knapp, M. S., & Murphy, J. (2003). The school and self-reflective
renewal: Taking stock and moving on. In Authors (Eds.), Self-reflective renewal
in schools: Local lessons from a national school renewal initiative (pp. 179–199).
Westport, CT: Greenwood Publishing Group.

Project SOAR. Retrieved August 26, 2005, 2005, from http://www.battelleforkids.org/b4k/

rt/about/our_work/data/SOAR

Sarmiento, J. W. (2004). Technology tools for the analysis of achievement data: An intro-
ductory guide for educational leaders. Philadelphia: Mid-Atlantic Laboratory for
Student Success.

Schmoker, M. (1996). Results: The key to continuous school improvement. Alexandria,

VA: Association for Supervision and Curriculum Development.

Schwartz, D. L., Bransford, J., & Sears, D. (2005). Efficiency and innovation in transfer.
In J. Mestre (Ed.), Transfer of learning: Research and perspectives. Greenwich, CT:
Information Age Publishing.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 49
 Scribner, J. P., Cockrell, K. S., Cockrell, D. H., & Valentine, J. W. (1999). Creating pro-
fessional communities in school through organizational learning: An evaluation
of a school improvement process. Educational Administration Quarterly 35(1),
130–160.

Senge, P. (1999). The dance of change. New York: Doubleday.

Sergiovanni, T. (1992). Moral leadership: Getting to the heart of the matter. San Francisco:
Jossey-Bass.

Shepard, L. (2000). The role of assessment in a learning culture. Educational Researcher,

29, 4–14.

Snipes, J., Doolittle, F. & Herlihy, C. (2002). Foundations for success: Case studies of how
urban school systems improve student achievement. Washington D.C.: Council of
Great City Schools.

Snow-Renner, R., & Torrence, M. (2002). No child left behind policy brief: State informa-
tion systems. Denver, CO: ECS.

Spillane, J. P. (2006). Distributed leadership. San Francisco: Jossey-Bass.

Spillane, J. P., Diamond, J. B., Burch, P., Hallett, T., Jita, L., & Zoltners, J. (2002). Manag-
ing in the middle: School leaders and the enactment of accountability policy. Educa-
tional Policy, 16, 731–762.

Spillane, J. P., Reiser, B. J., & Reimer, T. (2002). Policy implementation and cognition:
reframing and refocusing implementation research. Review of Educational Research,
72, 387–431.

Stein, M. (2003). Making sense of the data: Overview of the K–12 data management and
analysis market. Boston: Eduventures, Inc.

Stoll, L., Fink, D., & Earl, L. (2003). It’s about learning [and it’s about time]: What’s in it
for schools? London & New York: Routledge Falmer.

Streifer, P. A. (2002). Using data to make better educational decisions. Lanham, MD: Scare-
crow Press.

Stringer, E. (2003). Action research in education. New Jersey: Prentice Hall.

Stringfield, S., Wayman, J. C., & Yakimowski, M. (2005). Scaling up data use in classrooms,
schools and districts. In C. Dede, J. P. Honan & L. C. Peters (Eds.), Scaling up suc-
cess: Lessons learned from technology-based educational innovation. San Francisco:
Jossey-Bass.

Supovitz, J., & Klein, V. (2003). Mapping a course for improved student learning: How in-
novative schools use student performance data to guide improvement. Philadelphia:
Consortium for Policy Research in Education.

Togneri, W., & Anderson, S. (2003). Beyond islands of excellence: What districts can do
to improve instruction and achievement in all schools. Washington, D.C.: Learning
First Alliance.

50 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
Wayman, J. C. (2005). Involving teachers in data-based decision-making: Using computer
data systems to support teacher inquiry and reflection. Journal of Education for
Students Placed At Risk, 10(3), 295–308.

Wayman, J. C., & Stringfield, S. (2006). Technology-supported involvement of entire facul-

ties in examination of student stat for instructional improvement. American Journal
of Education 112(4), 549–571.

Wayman, J. C., Stringfield, S., & Yakimowski, M. (2004). Software enabling school im-
provement through analysis of student data. Baltimore, MD: Center for Research
on the Education of Students Placed At Risk, Johns Hopkins University.

Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks, CA: SAGE Publications.

Weiss, C. (1995). Nothing as practical as a good theory: Exploring theory-based evalua-

tions for comprehensive community-based initiatives for children and families. In
Connell, J., Kubisch, A., Schorr, L., & Weiss, C. (Eds.), New approaches to evaluat-
ing community initiatives. Washington, D.C.: The Aspen Institute.

Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. Cambridge,

UK, & New York: Cambridge University Press.

Witherspoon, P. D. (1997). Communicating leadership: An organizational perspective.

Boston: Allyn & Bacon.

Young, V. (2006). Teachers’ use of data: Loose coupling, agenda setting, and team norms.
American Journal of Education 112(4), 521–548.

Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n 51
 Endnotes
1.
This vignette was derived from ongoing exploratory research currently undertaken by
the Center for the Study of Teaching and Policy in several urban and suburban school
districts in the Northwest.
2.
For specific guidelines, see http://www.ed.gov/programs/titleiparta/reportcardsguidance.doc.
3.
Though not all of the assertions made in this line of inquiry have yet been tested empiri-
cally, there is accumulating evidence that both directly and indirectly learning-focused
leaders at both school and district level can realize substantial improvements in the
performance of students. See summaries of this work in Hallinger, P., & Heck, R. H.
(1996). Reassessing the principal’s role in school effectiveness: A review of empirical
research, 1980–1995. Educational Administration Quarterly, 32(1), 5–44; Leithwood,
K., & Riehl, C. (2003). What do we already know about successful school leadership?
Chicago: American Educational Research Association; Leithwood, K., Louis, K. S., An-
derson, S., & Wahlstrom, K. (2004). How leadership influences student learning. New
York: The Wallace Foundation.
4.
Many of these examples now under way in state and local sites are supported by The
Wallace Foundation.
5.
Excerpt from an unpublished descriptive account of the state policy environment in one
state included within the Study of Policy Environments and Teaching Quality, under-
taken by the Center for the Study of Teaching and Policy, University of Washington. The
account here reflects practices in place in the year 1999.
6.
Just for the Kids can be found at http://www.just4kids.org and School Matters at http://
www.schoolmatters.com.
7.
See http://www.seattleschools.org/area/siso/test/smallsisoschoollist.xml.
8.
See http://www.dataqualitycampaign.org/activities/elements.cfm.
9.
Only 48 states responded to the survey. There are no data from New York and New
Hampshire.
10.
See max.kde.state.ky.us.
11.
See http://www.teradata.com/t/page/145997/.
12.
Sample Balanced Scorecard: http://www.fultonschools.org/media-bin/documents/BSC.pdf.
13.
See http://www.galeaders.org/site/publications/publicationitems/pub_8roles.htm for a
description of all eight roles.
14.
See http://boston.k12.ma.us/bps/bpsglance.asp#improve for all Six Essentials for Whole
School Improvement.

52 Da t a - I n f o r m e d L e a d e r s h i p i n E d u c a t i o n
CTP—a national research consortium home-based at the University of Washington
and including three other research partners (Stanford, University of Pennsylvania,
and University of Michigan)—studies the way policies, leadership, and conditions
in schools, districts, states, the federal government, and the education profession
shape the quality of teaching and learning in the nation’s schools. The Center pays
particular attention to the ways these forces and conditions interact with each other
to influence what teachers, learners, and educational leaders do in daily practice.
A major goal of the Center’s program of research is to discover and
document the means for improving practice so that the nation’s young people
experience a challenging and equitable education. To that end, the Center’s re-
search products are designed to inform policymakers, practicing leaders, and the
educational reform community, along with scholars.

The Wallace Foundation seeks to support and share effective ideas and practices
that expand learning and enrichment opportunities for all people.
Its three current objectives are:
• Strengthen education leadership to improve student achievement
• Enhance out-of-school learning opportunities
• Expand participation in arts and culture
For more information and research on these and other related topics, please visit
our Knowledge Center at www.wallacefoundation.org.

KSA-Plus Communications: design and editorial

 Commissioned by

Miller Hall 404, College of Education • University of Washington • Box 353600 • Seattle, WA 98195-3600
Phone: (206) 221-4114 • Fax: (206) 616-8158 • Email: [email protected]