Features

2009 Wallace Foundation

Distinguished Lecture
Nested Learning Systems for the
Thinking Curriculum
Lauren B. Resnick
The 21st century will require knowledge and skill well beyond the
basic levels of reading and arithmetic that American schools know
how to produce more or less reliably. Delivering a thinking curriculum to all American students requires major reform in the
ways schools and districts organize their work. The transformation
of the institution of schooling that will be needed to make this
aspirational goal a real achievement is daunting. This article examines cognitive science, systems engineering, and social science concepts that are pointing toward a new foundation for policies and
practices that may radically improve the proportion of students
who can achieve true 21st-century skills.
Keywords: education organization; high-demand curriculum;
human capital; instructional leadership; learning
systems; professional development; routines; social
capital; systems engineering; 21st-century skills

decade into the 21st century, we still find ourselves

nearly as unprepared for what our students and society
need from education as we were when A Nation at Risk
(National Commission on Excellence in Education, 1983) was
first published. We are calling for career and college readiness
for all American high school students and for a focus on learning
21st-century skills, yet we find these terms hard to define,
much less translate into meaningful action. The overused phrase
21st-century learning can take on real meaning only if we compare
the challenges educators face now with those faced at the turn of
the past century. The idea that virtually all students can, and
should, learn a high-demand curriculum, focused on thinking
and reasoning and grounded in mastery of complex bodies of
knowledge, would have seemed quixotic to thinkers a century
ago. In the last part of the 20th century, we began to imagine
such possibilities and even establish them as national goals
(National Education Goals Panel, 1991). But it has only been in
Educational Researcher,Vol. 39, No. 3, pp. 183197
DOI: 10.3102/0013189X10364671
2010 AERA. http://er.aera.net

the present century that the concept of college readiness for all
has taken root as a serious education policy target.
Figure 1 provides a conceptual graph of the radically changed
education aspirations that characterize our present efforts to
teach 21st-century skills to virtually all American students
(cf. D. P. Resnick & Resnick, 1977). The very concept of schooling for all is only a few centuries old in Western countries. It is a
17th-century invention, born during the spread of Protestant
Christianity in Northern Europe and then taken up in Southern
Europe as part of the Catholic Reformation. In the 19th century,
basic schooling for all became a national aspiration in Europe and
North America, aimed at creating citizens and competent participants in national defense efforts. These initial mass schooling
efforts (the top left points in the graph) aimed to make high
proportions of the population literate but set a low criterion of
what counted as literacy.1 Catechism, in which individuals were
asked a set series of questions culled from specific religious texts
and were expected to provide standardized answers, was a basic
form of instruction. Participants were judged literate if they
could recite familiar texts and answer simple questions on which
they had been drilled.
As schooling became more widespread, the catechism form of
instruction moved into the lay classroom. The content changed
to include basic arithmetic, geography, history, and some science,
along with a broader range of texts for reading and writing. But
schoolroom discourse remained remarkably unchanged. Students
were assigned a text to read or a problem to work, and they were
then quizzed by the teachers with a set of questions that basically
checked on whether the students had done the assignment
(Mehan, 1979; Resnick, Wiliam, Apodaca, & Rangel, in press).
Across Europe and America, schooling became part of most
young peoples experience. But mass schooling did not even try
to engage most pupils in the kinds of knowledge-based reasoning
and problem solving that characterized elite schooling from
ancient times. This elite type of schooling became institutionalized in academies and technical institutions in the 19th and
20th centuries (see the bottom right point in the Figure 1 graph).
A sharp distinction in expectations for mass, or basic, education and what was taught to an elite minority still held in the

april 2010

183

new (some students have been successfully learning them in some

schools from the beginning of civilization). But the aspiration to
successfully teach knowledge-grounded reasoning competencies
to everyone is still just thatan aspiration. Is it a sensible one? Is
there any reasonable prospect of meeting it? What would it take?
Those are the questions I aim to answer in this essay. To anticipate,
I will argue that basic human capacity for learning and thinking
makes the aspiration humanly possible, if we think in terms of
the learning capabilities of most individuals. But the transformation of the institution of schooling that will be needed to come
close to making the aspirational goal a real achievement is huge.
I will suggest some steps we might take in the near future. To do
this, I will take the reader on a quick journey through a slice of
national achievement data from the past two decades and a summary of a consensual cognitive sciencebased theory of instruction that most scholars of learning and teaching now agree on.
I will then examine systems engineering and social science concepts that point toward a new foundation for policies and practices that may radically improve the proportion of students who
can achieve true 21st-century skills.
Reaching for the Star: Caught in the Basics Trap
FIGURE 1. Criteria and expectations of literacy from the 17th to
the 21st centuries. Adapted from D. P. Resnick & L. B. Resnick
(1977). The nature of literacy: An historical exploration. Harvard
Educational Review, 47(3), 370385.
United States at the turn of the 20th century. In 1892, the
Committee of Ten, a working group of educators from colleges
and universities across the country, attempted to create for the
first time a curriculum plan for all students who attended
American high schools (Hertzberg, 1988). The program that the
committee agreed on included the recommendation to teach
English, math, and history or civics to every student in every
academic year, along with recommendations for specific course
sequences. With expectable debate and quarrels about which
schools really counted as high schools and what all students
should be expected to learn, the committees program gradually
became the core curriculum for American secondary students.
Yet by 1910 only about 10% of students attended school beyond
the eighth grade. So the Committee of Ten curriculum is perhaps
best thought of as Americas response to Europes elite technical
schools.
The proportion of students attending secondary school in the
United States rose only very slowly over the succeeding decades.
At the centurys midpoint, roughly two thirds of the school-age
population was in high school. Thus the common core program for high school a century ago reached many fewer students
than today participate in our college preparatory high school programs. A rough estimate of literacy levels in the United States in
1900 (see central point in Figure 1) suggests that only about half
of young people were completing eighth grade. Those who did so
were literate at higher levels than the mass of 17th- and 18thcentury pupils but fell well short of the elite standard that the
Committee of Ten was attempting to codify.
Today we are aiming for something new in the world: An elite
standard for everyone (star at top right of the Figure 1 graph). That
is what the term 21st-century skills really means. The skills are not
184

educational Researcher

From the 1990s on, the public agenda of raising educational levels
for all has been promoted under the banner of the standards movement, often accompanied by the phrase All children can learn.
But neither term clarifies just what we have expected all children
to learn and thus what the standards ought to be. The evidence is
now pretty clear. We seem to have figured out how to teach the
basics to just about everyonewith special success in mathematics. But we are deeply unsuccessful in the rest of our 21st-century
agenda of moving beyond basic competencies to proficiency.
Figure 2 shows the National Assessment of Educational
Progress (NAEP) fourth-grade math scores over an 18-year
period spanning the end of the 20th and the beginning of the
21st centuries. The graph plots scores separately for Whites,
Blacks, and Hispanics. There has been a nearly continuous rise in
achievement scores over this period among all population groups.
The achievement gap has not closed, but it has shrunk somewhat.
In fact, and very much worth noting, Blacks and Hispanics were
doing as well in 2007 as Whites had been in 1990. The fourthgrade math gap would have closed completely if White students
had not continued to improve across the 18-year period! In
eighth-grade math, the pattern is similar (Figure 3), with Blacks
showing an especially steep rise, but a lower percentage of
students is meeting basic eighth-grade goals.
The story is less dramatic for reading, but there is evidence
that this achievement gap is shrinking somewhat as well. Over
the past 20 years, guided by a growing body of scientific research,
there has been substantially more teaching of the components of
basic literacy (phonemic awareness, phonics, vocabulary). And
this has shown up in higher first-grade scores on basic word
decoding skills (National Institute for Literacy, 2008). But by
fourth grade, when NAEP first measures reading, the focus is on
reading comprehensionunderstanding what you read. There,
the gains have been very small.
Overall, then, national achievement results suggest that as a
nation we are en route to eliminating basic illiteracy and innumeracy. The really troubling performances of the early 1990s, in

FIGURE 2. Disaggregated NAEP scores for 1990 through 2007, fourth-grade mathematics. Source: U.S. Department of Education,
Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), NAEP Data
Explorer, available online at http://nces.ed.gov/nationsreportcard.

FIGURE 3. Disaggregated NAEP Scores for 1990 through 2007, eighth-grade mathematics. Source: U.S. Department of Education,
Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), NAEP Data
Explorer, available online at http://nces.ed.gov/nationsreportcard.
which large numbers of our minority students, along with some
White children of poverty, seemed to be fundamentally illiterate
or innumerate, have changed. We are on the way to meeting our
basic education goalsand we have achieved this even as we
have absorbed growing numbers of students with limited English
proficiency into the nations schools. It appears that the standards
effort, including requirements for disaggregated test score reporting, is having the hoped-for equity effects. We are teaching basic
literacy and math to more and more of our elementary school
children, and fewer and fewer are being left way behind.
We are, however, very far from reaching the star. Proficiency
levels on the NAEP remain low, and there are very few students of
any subgroup reaching Advanced levels. Furthermore, it now seems
likely that the accountability regime that appears to be creating
much of the improvement in Basic skills may actually be limiting
progress toward the kinds of more challenging competencies we
seek. The effects of high-stakes, low-cognitive-demand tests on

instructional practice have been quite widely documented by

now (Koretz & Hamilton, 2006; McNeill, 2002). Most studies
show that state tests have led to a noticeable increase in the amount
of instructional time devoted to the tested subjects and a corresponding drop in nontested subjects (Center on Education Policy,
2008). Most districts that increased time for English language
arts or mathematics also reported substantial cuts in time for
other subjects, including social studies, science, art and music, and
physical education (Center on Education Policy, 2008).
Even within the tested subjects, it appears that test-based
accountability may be narrowing what is taught. In many urban
school districts, teachers are emphasizing test preparation over
other aspects of their districts official curricula (Shepard, 2002
2003). As end-of-year testing dates approach, teaching time is
spent on test practice. In one district that we have studied intensively, elementary students stop reading and discussing gradelevel-appropriate books in February and instead spend time
april 2010

185

FIGURE 4. Definitions of knowledge, competent performance,

instruction and learning, and aptitude and intelligence to which
learning scientists subscribed in the 1920s and today.

digesting brief passages, accompanied by multiple-choice test

items that mimic the ones that appear on the state tests. District
leadership reluctantly supports this practice because the tests
carry heavy consequences.
Although no one intended such an outcome, the test-based
accountability movement seems to have taken the nation back to
something like the minimum competency movement of the
1970s (Jaeger & Tittle, 1980), which was an effort to ensure that
poor and minority students would at least be taught the basics
but with no grounded approach to high-cognitive-demand learning for the great majority of students.
Reaching for the Star: The Thinking Curriculum
Despite the rhetoric of 21st-century skills, we have by and large
built our accountability system so that it actually suppresses the
kind of learning that the 21st century calls for. Since the middle
of the 20th century, the science of learning, and thus the underpinnings for trying to reach the gold star of knowledge-based
reasoning for all Americans, has expanded substantially. The recommendations now coming from an expanded, multidisciplinary
learning science community are substantially different from
those of the first half of the 20th century (Resnick, 1987b, 1999).
The transformation of learning theory over a century of its
attempted application to schooling is remarkable. Scientific
research on learning has produced changed concepts of knowledge
itself, new criteria for what counts as competent performance and
as intelligence, new principles for instruction, and even new theories of how educational organizations work.
Figure 4 compares definitions of knowledge to which learning
scientists subscribed in the 1920s and today. Instead of defining
knowledge in terms of a bounded list of facts (bonds as E. L.
186

educational Researcher

Thorndike, 1932, called them) coming from a small number of

controlled sources, we now define knowledge in terms of schemas
and conceptual structures. We recognize that knowledge comes
from multiple sources and that it is often public, rather than
controlled by academicians. And we know now that knowledge
is exploding every day, that it is emergent from the complex interactions in which people engage.
Our changed understanding of knowledge leads to changed
views of what counts as competent performance. Automated skill
in performance of routines still matters, but 21st-century skills
mostly focus on a persons ability to participate in argumentation
and discussion. Question-and-answer performances are replaced
by discursive processes that include productively challenging colleagues, paraphrasing, and interpreting presentations by others.
And although individual performances still matter, much
knowledge work is distributed, involving collaboration with
others (Resnick, 1987a).
These new concepts of knowledge and competence entail new
understandings of how instruction and learning can best proceed. Although practice and repetition still play a role in acquiring a relatively narrow set of skills and information (e.g., solving
algebra problems speedily and accurately), we now recognize that
reliable learning of complex material will proceed through a process of interpretation and explanation. Instead of just stamping
in correct answers and stamping out the incorrect (as
Thorndike taught us in the 1920s), we now try to teach students
the metacognitive capabilities of self-monitoring and selfmanagement of learning. And we recognize that there are important social aspects of learning, even when each individual is
responsible for mastering some body of skill or knowledge.
There have also been important changes in how we think
about aptitude and intelligence. Instead of intelligence being
viewed as an entity, something that people have a fixed amount
of and thatfor manylimits learning possibilities, we now
understand intelligence to be learnable (Dweck & Molden, 2005;
Resnick & Nelson-LeGall, 1997) through social processes that
include participation in certain forms of high-demand learning.
These changes in our understanding of learning point toward a
form of instruction that I have come to call the Thinking
Curriculum (Resnick, 1987b; Resnick & Klopfer, 1989). The
Thinking Curriculum calls for instruction that is high in cognitive
demand (conceptual learning, reasoning, explaining, and problem
solving are engaged daily) and that is embedded in specific, challenging subject matter. Evidence has accumulated that teaching
cognitive skills in the absence of specific content rarely works. It
appears that thinking abilities have to develop in the course of
reasoning about specific information and knowledge. At the same
time, there is plenty of evidence that drilling on the facts without
demands for explanation and reasoning produces fragile knowledge, which is likely to disappear once the test is over and is unlikely
to transfer (Chi, 2000). A form of the Thinking Curriculum that
uses guided classroom discussion of core disciplinary ideas (we call
this accountable talk) apparently yields both long-term retention
and transfer to other disciplines (Resnick, Michaels, & OConnor,
in press). The strongest examples come from controlled experiments in which an individual teaches elementary mathematics in
the Thinking Curriculum style and in the traditional recitation
style (e.g., Bill, Leer, Reams, & Resnick, 1992; OConnor, 2001).

Other studies, in which multiple teachers are trained to teach science or math discursively, produce less dramatic differences in
scores but still show significant transfer effects on measures of general cognitive functioning across disciplines (e.g., Adey & Shayer,
2001; Mercer, Dawes, Wegerif, & Sams, 2004).

Scaling the Thinking Curriculum: An Organizational

Design Problem
We know that the Thinking Curriculum can actually work for a
broad range of individual students. The question now is whether
we can figure out how to scale the kinds of teaching that are
needed to reach the 21st-century star. Doing so will mean preparing educators to adopt a significantly different way of teaching
than most of them experienced in the course of their own schooling. Doing so will call not just for pedagogical shifts but also for
deeper knowledge of core subject matter than many current
teachers haveperhaps especially in mathematics and science.
Consider mathematics, which is the field that has made the
greatest advances in codifying methods of teaching that ensure
both mastery of basic skills and conceptual understanding and
problem solving. To successfully lead a Thinking Curriculum
mathematics class, a teacher must be able to recognize the mathematical content embedded in initially ill-formed articulations of
concepts and explanations. Then, the skillful teacher orchestrates
classroom interactionsincluding challenges, revoicings, and
targeted discussions of student explanationsthat bring the
important concepts to light in a form that all students can share
(Michaels, OConnor, & Resnick, 2007).
Compared to traditional mathematics teaching (see Stigler &
Hiebert, 1999) in which the teacher leads the class through a
relatively simple script of choosing and naming the correct steps
in a procedural task, Thinking Curriculum instruction calls for
guiding an only partially scripted line of talk, one in which childrens initial formulations of ideas are halting and filled with nontechnical language.
This kind of teaching is not discovery learning in which
children are free to explore a problem space, with the teacher riding along for the adventure. Instead every class session has a clear
intellectual goal, a kind of macroscript for a directed conversation. But the class traverses the script in a series of byways,
some introduced by students who are struggling to articulate
their understanding of a concept, others introduced by students
already confident in their knowledge. Both kinds of students are
likely to be present in most classrooms.
Thinking Curriculum lessons are also not simply collaborative learning (Johnson & Johnson, 1986; Slavin, 1996), with
which they share important features (most obviously the importance of student talk and explanation and the possibility of
groups of students working on their own for portions of the
instructional time). In effective Thinking Curriculum classrooms the teacher does not simply step back and let students
discover knowledge or problem solutions for themselves. Instead,
the teacher guides a knowledge-driven discussion focused on
explaining concepts in the context of specific texts, tasks, and
interpretive questions. Every class sessionor small-group meetinghas a clear intellectual goal, with students working out specific conceptual understandings under teacher guidance. The
teacher is guided by a macroscript that specifies the goal (student

understanding of specific concepts) and some likely landmarks

along the way, with the route to the goal marked by explanatory
byways. These byways are crucial for building understanding,
but it is also crucial to return to the planned path (Ball &
Lampert, 1998; Lampert, 2001; Lampert & Ball, 1998).
All this requires confident subject-matter knowledge on the
part of teachers. What works in the Thinking Curriculum is not
generic but deeply subject-specific teaching. Yet considerable
research has documented the weak mastery of core subject matter
by many teachers educated in our current system (Clotfelter,
Ladd, & Vigdor, 2007; D. K. Cohen & Hill, 2000). So there is a
substantial challenge ahead for teacher preparation programs and
teacher credentialing in terms of creating incentives for high performance and on-the-job training. All these proposed solutions
focus on populating schools with better educated teachers (i.e.,
improving the human capital of the education sectorsee below
for further discussion). That is important. But by itself it will not
create enough good teaching to go aroundthat is, to reach all
of our students. The challenge to individual teachers is matched
perhaps exceededby the challenge to educational organizations
and the policy structures within which they act.
Current policy discussions often aim to solve the problem of
disappointing levels of learning by investing heavily in theories of
performance management. The prescription for better performing schools, according to this theory, is more frequent measures of
student performance and greater attention to this output data
(in economists terms, productivity). This has led to a virtual
industry of student measures that can be administered early and
often, in the form of interim, or benchmark, tests. As noted earlier,
these tests have come to control the de facto curriculum, as school
districts and school principalsworried about poor performance
on state accountability instrumentsprescribe more and more
test preparation, mostly in the form of practicing items that are
very like the ones that will appear on the state tests. These items,
for reasons of cost, familiarity, and certain psychometric considerations, are mostly simple multiple-choice questions, with little
opportunity for the kind of interpretive knowledge work that the
Thinking Curriculum calls for. This growing practice, encouraged by the offerings of test providers, inflates attention to the
end-of-year test items and exaggerates the basic skills character
of the standards movement.
Even if the accountability tests were to be changed substantially (along lines being discussed today in many venues), performance management based on student test scores alone would be
a far cry from what is needed to build a new educational and
organizational management system that can support, enhance,
and sustain the Thinking Curriculum. We need a method closer
to systems engineering (Resnick, Besterfield-Sacre, Mehalik,
Sherer, & Halverson, 2007), one that examines processes along a
chain of linked policies and actions.
Process management was widely embraced in the 1980s in the
business world under labels such as total quality management
(TQM) and ISO 9000. A catalyst in this movement was then
Secretary of Commerce Malcolm Baldridge, for whom the
Malcolm Baldridge National Quality Award was established by
an act of Congress in 1987. TQM is not specific to any one type
of organization and was considered applicable to education and
government agencies as well as the private sector. During the
april 2010

187

FIGURE 5. Feedback control system for a manufacturing process. Source: Modified from W. C. Turner, J. H. Mize, K. E. Case, & J. W.
Nazemetz. (1992). Introduction to industrial and systems engineering (3rd ed.). Upper Saddle River, NJ: Prentice Hall.
1990s there were attemptslargely encouraged and funded by
the American business communityto apply TQM to education
(e.g., a Malcolm Baldridge award for school systems was initiated). Key to the TQM management philosophy was that all
employeesfrom floor workers to CEOsbe engaged in seeking quality improvements and that processes as well as outputs be
measured and improved as necessary. It is odd that, just as we
have truly engaged the agenda of focusing on results, we seem to
have left behind the attention to organizational processes that is
a crucial aspect of quality management.
Engineering a Nested Learning System for
the Thinking Curriculum
The systems engineering concept of process control (Turner,
Mize, Case, & Nazemetz, 1992) provides a foundation for
organizational design that goes beyond just measuring outputs.
Originally introduced for manufacturing organizations, systems
engineering approaches have also been heavily applied to the service industry, including financial, medical, and educational organizations. A notable example of how process engineering has
focused on the values and needs of people is the redesign of hospital systemsincluding improved surgery room functionality,
reduction of errors in medical procedures and medicine distribution, improved diagnosis systems, improved scheduling to reduce
patient waiting times, and effective distribution of information
and resources to minimize hospital costs (Sahney, 1993).
Figure 5 provides a schematic of how a process control system
would work in a manufacturing setting. The production process
(circle in the bottom line of the figure) is where the fundamental
work on the product is done, using a variety of input resources
materials, people, and so forth (shown to the left of the production process circle). The quality of the end product and the
processes used to produce it are both continuously measured.
Results are compared to plans (diamond at far right), and a leadership and management team (central rectangle in the figure)
uses these measurement results (on outputs and processes) to
188

educational Researcher

decide whether the desired objectives and outputs have been

satisfactorily met (Davis, 1982). When results do not meet
expectations, the leadership and management team (central rectangle) takes corrective actions. Because processes as well as products are monitored, there are opportunities to determine where
the process has broken down. Initiatives and plans as well as operating procedures may be modified (far left rectangle) by the leadership and management team, and resources may be reallocated
to support the changes.
Figure 6 adapts the basic process control model to educational
organizationsspecifically large urban districts. The figure
shows a nested, or layered, system. The production process of
the preceding figure is now the classroom level. Leadership and
management processes are shared between the school and the district level. Outputs (student learning) flow out of the classroom
and produce data that allow results to be compared to the plan
(the diamond to the right). If they are acceptable, the process is
allowed to continue. If not, corrective action is determined. The
pipes symbolize the ways in which influence can flow between
and among levels of the system, sometimes enabling, sometimes
constraining, action at the next level.
As a step toward working with urban school districts to build
a process-engineering management system, the Processing
Engineering for Education Results (PEER) group at the
University of Pittsburgh, headed by engineer Mary BesterfieldSacre, developed a hypothetical flow model of processes from the
district to the school level. This is shown in Figure 7. The output
of the system (at the right) is expected to be student learning,
using multiple measures. Five types of classroom enablers are
specified. These enablers draw on a substantial body of research
much of it beginning with John B. Carrolls Model of School
Learning initially put forward in 1963 and modified by Carroll
himself (in 1989) and many other scholars (e.g., Berliner, 2006)
in the intervening years. Measures for many of these classroomlevel enablers exist and can be used in the process-engineering
effort of school districts.

FIGURE 6. Three-level conceptual model of K12 district system with process feedback. Source: L. B. Resnick, M. Besterfield-Sacre, M. M.
Mehalik, J. Z. Sherer, and E. R. Halverson. (2007). A framework for effective management of school system performance. In P. A. Moss (Ed.),
Evidence and decision making: The 106th yearbook of the National Society for the Study of Education (NSSE) (Part I, pp. 155185).
Malden, MA: Blackwell.

To the left in Figure 7 are hypothesized enablers (sometimes

constraints) that the school or district introduces (some of the
processes fall between district and school or between school and
classroom). The research basis for the elements further to the left
in the diagram is much thinner than for classroom processes, but
there is widespread agreement in the policy making and policy
research communities that each of the elements named is potentially important. In fact, many proposed policy initiatives are
based on assumptions involving these elements (e.g., hire and
reward teachers with more knowledge and skill, provide continuing professional development, modify principal hiring and school
assignment policies).
To check our hypothesized process model against the implicit
models of leaders in urban school districts, we invited key decision makers in several urban school districts to participate in a
series of mapping exercises. More than 100 urban district officials (including superintendents, deputy superintendents, chief
academic officers, instructional supervisors, and principals) participated. Participants were given a set of tiles, each containing
one of the elements in Figure 7. They were permitted to discard
any elements they did not deem centrally important and to add
new ones, if necessary, to reflect their views. Participants were
asked to create influence maps of their policies aimed at
improving student learning. Twenty-eight groups of district officials created 28 different maps using our organizational elements and enablers plus a small number of additional ones that
they added.

Although there were variations among the maps, certain

characteristics were largely shared. First, our hypothesized system elements did, according to our participants, constitute the
fundamentals of the K12 system. Even though district leaders were instructed to add additional elements as needed, few
were added, and there was no consistency among the additions. In addition, the pattern of influence revealed in the 28
separate practitioner maps was quite similar to what is shown
in Figure 7.
We combined the qualitative knowledge embodied in our district experts graphs using an algorithmic approach involving a
recursive path-counting routine written in VB.NET (Clark,
Sherer, Besterfield-Sacre, & Resnick, 2007). The results of the
VB.NET analyses identify frequently occurring paths among the
28 maps developed by our school district participants. These
paths represent prominent shared theories of action among our
experts for how to influence classroom processes to produce
improved student learning. Four high-frequency two-element
paths were identified:
Instructional leadership teacher beliefs
Quality of professional community teacher beliefs
Professional development teacher knowledge and skill
School calendar instructional schedule
Further, four individual elements (instructional leadership,
quality of professional community, teacher beliefs, teacher
april 2010

189

FIGURE 7. Fundamental elements of the K12 system according to participant practitioners.

knowledge and skill) also appeared in frequently identified paths

of three, four, or five elements. All 28 groups placed teacher
beliefs in their diagrams. All but one included teacher knowledge
and skill. Instructional leadership, quality of professional community, and professional development were used with high frequency. Thus, according to our participants, there is a solid core
of processes essential to enabling the classroom practices that we
know produce student learning.
Policy Planning: The New Reform Triangle
Although the terms human capital and social capital had deliberately not been used in our mapping exercises (to avoid having
technical terms block our experts ability to articulate their own
theories of action), the education leaders we worked with shared
the underlying ideas expressed by those terms. Our participants
affirmed the importance of organizational features of schools that
social scientists and policy experts have been addressing for some
time: human capital (expressed as teacher knowledge and skill,
teacher beliefs, and instructional leadership) and social capital
(expressed as quality of professional community and effortbased instructional culture). To these two terms, we add a third:
instructional tools and routines (expressed as appropriate assessments available, curriculum and materials, professional
development). These three organizational features comprise a
190

educational Researcher

new policy triangle (see Figure 8) that is beginning to guide policy designs for improved achievementalthough it is rare for
advocates or scholars to consider the three in combination.

Human Capital
Economists tend to be especially interested in human capital:
what people in the organization know and know how to do
(Harbison & Hanushek, 1992). Human capital is typically measured by credentials, performance observations, and individual
outputs (in education, student learning). Economists have related

FIGURE 8. Policy triangle to guide policy design in educational

settings.

student performance on academic measures to the number and

type of courses teachers have taken in college or graduate school
(Boyd, Grossman, Lankford, Loeb, & Wyckoff, 2006, 2008).
Some recent work using more refined measures of teacher knowledge also shows a significant positive relationship between teacher
knowledge and student achievement (Hill, Rowan, & Ball, 2005).
Our expert educators agreed that teacher knowledge is an
important enabler of the kinds of classroom activity that enhance
student learning, but they included teacher skill in their definition
of teacher competencereferring to teachers abilities to communicate that knowledge, engage students interest, and in general create classroom environments for successful learning. They
thus embraced the broad view of teacher knowledge introduced by
Lee Shulman (1987), who distinguished between content knowledge
and pedagogical content knowledge, arguing that both were essential to good teaching (see also Darling-Hammond & Bransford,
2005). Although popular among educators and education
researchers, it has not proved easy to measure pedagogical content
knowledge in a way that provides empirical evidence of making a
difference in student learning. Some progress is now being made
using structured classroom observations (Grossman & McDonald,
2008) and teacher logs (Rowan & Correnti, 2009).
The educators in our study also enriched the concept of
human capital by pointing to teacher beliefs (the extent to which
teachers believe their students can learn at high levels) as a
hypothesized producer of greater student learning. And they
identified instructional leadership as an important element in
increasing productivity, thus making the quality of school principals an important part of the human capital equation for their
schools.
Much current policy discussion is aimed at remedying the
human capital gap by processes of selection (including alternative
pathways into education careers), retention, dismissal, and differential pay (National Academy of Education, 2009). Across the
United States, several experiments are now under wayfor
example, the Denver school systems ProComp pay for performance systemlinking such incentives to specific training and
instructional programs.
Even if many of these policies were widely implemented, however, there would be a long period before most of the teaching and
leadership force in schools would count as highly qualified under
new and more stringent definitions of quality. In the interim, old
organizational practices might actually suppress the development
of the new, knowledge-based Thinking Curriculum that is our
goal. Thus we will need to consider how to create human capital
within education organizations. The two most promising routes
appear to be development of social capital within schools and
systematic introduction of instructional tools and routines that
have the power to directly change classroom practice and thereby
increase learning.

Social Capital
Social capital is a term introduced by sociologists (Becker, 1964;
Coleman, 1988) referring to resources for action that inhere in
the relations or interactions among peoplethe opportunities
that some people have, and that organizations can create, for
acquiring knowledge and other resources through interactions
with others. Social capital is used to refer to social ties and trustful

relationships (Adler & Kwon, 2002; Nahapiet & Ghoshal, 1998).

A number of sociologists studying processes of education reform
have begun to document links between social capital (e.g., groups
of teachers professionally engaged with one another within a
school) and the forms of knowledge-based thinking that cognitive and sociocognitive instructional theory recommends (e.g.,
Bryk & Schneider, 2002; Frank, Zho, & Borman, 2004;
Gamoran et al., 2003; McLaughlin & Talbert, 2001; Newman,
1996).
Social capital refers to the ways in which people in an organization share what they know. With whom do they talk? How
openly or widely do they share informationboth positive and
negativeabout their work? Do they know or care who has
expertise? How broad or narrow are their networks? Our expert
educator participants referred to social capital mainly as the quality of professional community within a school and viewed it as a
primary means of building human capital. Social science research
supports their practice-honed view. For example, in a large study
of social capital in New York City schools, high social capital (as
measured by structured surveys) apparently led many competent
teachers to stay in schools serving the poor, even if the teachers
had opportunities for better paying jobs nearer to their own
homes in the suburbs (Leana & Pil, 2009).

Instructional Tools and Routines

Human and social capital are powerful concepts, but they do not
tell the whole story. As do all organizations, schools function
through a set of more or less interconnecting routines
repetitive, recognizable patterns of interdependent actions,
involving multiple actors (Feldman & Pentland, 2003, p. 113).
These routines are critical for any organization to function effectively because they provide stability and continuity over time
(Feldman, 2000; Feldman & Pentland, 2003; March, 1981;
March & Simon, 1958, 1993) and structure action in organizations (Allison, 1971; Gersick & Hackman, 1990). Groups and
individuals in the organization develop routines that constitute
the normal ways in which work gets done. These routines are not
always in the official manuals, but they allow members to perform satisfactorily in the judgment of clients and supervisors and
for their own self-satisfaction. Such routines often involve adaptation to internal and external institutional constraints and may
also recruit the power of informal below the radar work groups,
as documented by sociocognitive research (Brown & Duguid,
2000; Orr, 1996; Resnick, Saljo, Pontecorvo, & Burge, 1997;
Suchman, 1996). Research has documented the ways in which
organizational routines, both formal and informal, frame and
enable interactions, provide stability across time, and assist in
socializing new organizational members (M. D. Cohen &
Bacdayan, 1994; Feldman & Pentland, 2003; Sherer & Spillane,
in press; Spillane, Mesler, Sherer, & Croegaert, 2010).
What kinds of routines might be introduced into schools
and school systems that would build the human and social
capital needed? There are several possibilitiesranging from
instructionally based supervision systems to tools and routines
for instruction. Stated most directly, it probably would help to
put curriculum of known effectiveness, along with materials
and procedures for classroom implementation, in the hands of
teachers.
april 2010

191

Throughout the first half of the 20th centurywhen school

attendance was expanding, people were moving from farm to
city, and American cities were absorbing then-unprecedented
numbers of internal (South to North) and external immigrants
large school districts laid out well-defined curriculum and
instruction plans and expected teachers to follow them. Starting
in the 1960s, educational tastes changed and large pressures
against industrial models of education developed. Over the
three and a half decades since David Tyack (1974) described the
functioning of centralized school systems as Tayloristic (referring to efficiency methods in which teachers were expected to
implement detailed programs of instruction on a strict schedule),
there has been a growing rejection of the idea of centrally imposed
instructional programs. This has been accompanied by a rhetoric
of professionalization of teachers, with the implication that
professionals should develop their own instructional plans and
programs.
Today, the language of professional independence for teachers
is so widespread that even when school districts attempt to implement systems of managed instruction as a way of improving educational provision for underserved populations of students, they
mostly cast their curriculum offerings as guidance for teachers
rather than as required programs. Textbooks are adopted but often
used only sporadically. In the face of high-stakes accountability for
student performance on state tests, most districts offer interim
assessments intended to provide guidance on how to meet student
needs. In many cases, however, the actual use of such assessments
is voluntary. In a recent PEER study of a large Eastern urban school
district, for example, teachers reported not regularly using the
district-supplied teaching materials in math (which they judged
too fast paced for their students), and most did not administer the
voluntary end-of-unit assessments that the district supplied.
The resistance to curriculum-based solutions is beginning to
decline. There is growing evidence that structured instructional
tools and routines for using them can be a powerful route to better teacher performance and increased student learning. In the
next section, I report evidence that the three elements of the new
policy triangle can be used together to meet 21st-century education reform goals.
Using the Policy Triangle to
Improve Educational Results

School-Based Instructional Tools and Routines

Although school districts have, by and large, been shying away
fromor not fully implementingcurriculum-based reform,
the instruction-based reform strategy has been kept alive in
American schools by some of the intervention models that
emerged as part of the 1990s reform strategy known as comprehensive school reform (CSR). In the CSR approach, an individual school uses state, federal, or philanthropic dollars to contract
with an organization (usually a nonprofit, sometimes a for-profit
or a university group) that provides a defined service program,
usually one that includes professional development and other
support for teachers.
The existence of these CSR schools, using a variety of reform
models, provided the opportunity for a research group at the
192

educational Researcher

University of Michigan (Correnti & Rowan, 2007) to compare

different reform strategies effectiveness in raising student
achievement in elementary schools. The groups Study of
Instructional Improvement research program compared schools
using three different models of school improvement: Success for
All (SFA), Americas Choice (AC), and the Accelerated Schools
Project (ASP). The first two (SFA and AC) are curriculum-based
models in which the external provider supplied texts and other
teaching materials, specified student grouping and instructional
processes in detail, and provided training in specific pedagogical
strategies. ASP, by contrast, provided structured support to school
staff members to develop their own instructional plans and
implementations around a broadly shared philosophy of learning
and teaching. ASP, in other words, focused on building social
capital, whereas AC and SFA used instructional tools and routines as a route to improved student learning.
The findings were striking. Using student learning as the criterion, the curriculum-based approaches (SFA and AC) performed
better than the ASP approach that worked mainly on developing
social capital. What is more, the learning gains were tightly tied
to the specific instruction that was the focus of the program. Each
program showed significant effects only for the core curriculum
component it focused onbasic reading skills in the case of SFA,
writing skills in the case of AC. In other words, well-implemented
curriculum worked, but the effects were specific to the tools and
routines introduced by the provider.

District-Based Instructional Tools and Routines

Can the instructional tools and routines strategy be used to
increase student learning across a broad swath of schools in a
district? This is the question we addressed in an experimental
study conducted in several dozen elementary schools in a large
urban district in the Southwest, a district struggling to raise
achievement among a large and growing population of Spanishspeaking immigrant children (Matsumura, Garnier, & Resnick,
in press). The curriculum approach we introduced was based on
the reading comprehension program Questioning the Author
(QtA) developed by Isabel Beck and Margaret McKeown (2002).
The tools and routines of QtA were introduced by placing dedicated literacy coaches into half of the schools under study and
training them using a program called Content-Focused Coaching
(CFC).
Coaching is a popular intervention in school districts that are
trying to raise achievement. It is, in theory, a form of professional
development that is school embedded and therefore close to
instructional practice. It alsoagain in theoryuses the best
teachers in the system to help build skill among the larger teaching force. Using the PEER process of specifying expected influences, coaching can be seen as a means of upgrading teacher
content and pedagogical knowledge, along with teacher beliefs
about student ability to learn. As shown in Figure 9, coaching
(the central diamond) is expected to enhance key enablers of
classroom practice such as content coverage and quality of classroom interaction, thus leading to gains in student learning.
Coaching is rarely enacted according to theory, however.
District practices of hiring and assigning coaches, sometimes governed by union contracts, along with job postings and salary

FIGURE 9. Coaching enhances key enablers and is constrained by district policies and practices.
policies, often mean that coaches are selected from the teaching
ranks by seniority or preference of principals rather than by demonstrated capacity to increase student learning. Coaches job
descriptions are often only vaguely specified. And, reporting usually to individual principals who do not have a developed understanding of what to expect from coaches, they are assigned to a
myriad of tasks (ranging from supervising testing, serving as substitute teachers, or providing personal teaching to underperforming students) and do not have the opportunity to develop a
systematic coaching relationship with teachers. Figure 9 also
illustrates how the quality of coaching in a school district is
dependent on a large set of policies and practices (those to the left
of the coaching diamond in Figure 9) that are heavily influenced
by the district central office.
In our study of CFC, we enacted a carefully focused program
of coaching in upper elementary classrooms teaching reading
comprehension (Matsumura, Sartoris, Bickel, & Garnier, 2009).
The program used to train coaches had been developed over
several years at the University of Pittsburghs Institute for
Learning (Bickel & Artz, 2001). We worked with district administrators to select demonstrably successful reading teachers to be
trained as coaches. We also worked with school principals to
develop agreements that would make it probable that coaches
assigned to their schools would be scheduled for regular meetings
with subgroups of teachers, would be allowed to make classroom
visits to individual teachers, and would not have competing work
assignments (Matsumura et al., 2009).
Twenty-nine of the lowest performing elementary schools, all
with high proportions of English language learners, were randomly assigned to either the CFC or a comparison condition.
Teachers and principals in both sets of schools responded to periodic surveys and interviews. There were systematic observations
of classroom text discussions and recording of the complexity of
the texts being used in instruction. Students reading test scores
on the state accountability tests were tracked over several years.
Teachers in the CFC intervention schools significantly
increased participation in coaching focused directly on classroom practice. The quality of the text discussions in their classrooms improved: The classes read more difficult texts, they

actively referred to the texts as they discussed them, and the

teachers used accountable talk (Resnick, Michaels, et al., in press)
classroom discourse strategies. Students in the CFC schools
showed significantly higher reading test scores (effect size = .25
after 2 years). The effect was strongest for English language learners (Matsumura, Garnier, & Resnick, in press; see Figure 10).

Combining Social Capital and Instructional Tools and

Routines in High Schools
Many scholars of education practice and reform suggest that the
quality of professional community in schools is tightly associated
with student achievement. Considerable work has been done to
develop ways of managing schools that are likely to build and
sustain professional communities (Loeb, Darling-Hammond, &
Luczak, 2005; McLaughlin & Talbert, 2001). Indeed, it is possible that the successes of curriculum-based whole-school models
are in part due to the professional community commitments that
are evoked by the forms of training and support that are part of
the implementation packages that the sponsoring organizations
build into their programs.
In 2001, the Institute for Learning began designing and piloting an intervention program that explicitly combined instructional tools and routines with professional development strategies

FIGURE 10. Content-Focused Coaching (CFC) schools showed

improved reading test scores.
april 2010

193

FIGURE 11. Pedagogy and content routine.

aimed at building professional learning communities (McConachie

& Petrosky, 2010). The institute worked in two urban districts: six
high schools in Austin, Texas, in which all teachers in the four core
high school disciplines (science, math, history/geography, and
English) participated, and schools in the Los Angeles Unified
School District in which the departments of mathematics and/or
English, along with their principals, agreed to participate.
In both districts, intensive professional development led by
subject matter experts was organized around a spine of curriculum units and lessons that were explicitly linked to the districts
official curriculum guidance documents. The units and lessons
were designed to educate teachers in high-cognitive-demand
forms of classroom instruction. The form of training used was
intended to induce new forms of professional engagement in
what we have termed a kerneling process (Resnick & Spillane,
2006; Resnick, Spillane, Goldman, & Rangel, in press) in which
formal routines embedded in an institution give rise to a next
generation of practices that are kin to the externally introduced
routine but not identical to it. Kernel routines are designed to
deliberately displace standard routines of practice. Participating
teachers are expected at first to follow the new routines quite
faithfully. However, the routines have been developed with the
goal of encouraging new forms of professional interaction that
are consonant with the curriculum plan but crafted by the participants to fit their interactive and learning preferences.
The pedagogy and content routine (PCR) begins by engaging
teachers in a tightly crafted routine consisting of a specific set of
194

educational Researcher

professional development practices. Training and practice of

PCR occurs separately for each teaching discipline, but when the
routine is used for several disciplines within a school (as it was in
our Austin implementation), substantial cross-seeding and
development of a larger institutional change within a school is
expected to occur.
Figure 11 depicts how PCR works. Teachers, coaches, and
lead teachers first experience the sequence of activities described
in column 2. They begin as learners of model lessons taught by a
trainer. They then engage in a trainer-led process of deconstructing or interpreting what their learning process has been and the
role of the trainer in evoking their learning processes. Next (still
in column 2), participants are asked to teach the model lesson or
a modification of their own design. They are observed by the
trainer and other teachers, and they then participate with the
observers in an analysis of the pedagogy and content of their
teaching. Teachers have the experience of both observing and
being observed during this phase of implementationalways
with careful attention to the content taught and the cognitive
processes evoked among students. They next modify or adapt the
lesson for future teaching and continue the teaching-observinganalyzing sequence. Alternatively, they can ask for one or more
additional model lessons from the spine.
Even as the teacher group cycles through the PCR, the kerneling process is expected to engender new forms of school practice.
As shown in column 3, we expect changes in leadership activity,
in norms of trust and collaboration, and in specific collaborative

routines and structures of interaction. All of these forms of social

capital should affect aspects of classroom practice (column 4) and
thus of student learning (column 5).
A research team headed by Joan Talbert of Stanford University
evaluated the use of PCR in the six Austin, Texas, high schools.
The evaluation report (Talbert & David, 2008) suggested that
PCR provides an effective vehicle for developing teacher collaboration centered on instruction, as well as for increasing the academic rigor of teaching and learning. However, the authors noted
that it would need further central administration support to realize its full effects. Another study in Los Angeles (David & Greene,
2008) yielded similar results in mathematics. Implementation in
English language arts, which received less intensive administrative
support, was not as effective.
In Conclusion
We stand at the cusp of some potentially important shifts in how we
think about education reform. Resistance to external specification
of routines and curriculum seems to be ebbing. But increased
policy interest in curriculum-specific instructional practices will
bear fruit only if we can learn how to embed detailed curriculum
guidance in organizational designs that support the complex
sociocognitive practices of participants and the diversity of students in our schools. Systems that aim to develop extended
knowledge and complex forms of argument and reasoning among
students will fail if teachers are restricted to scripted lessons that
close off discussion. Instructional tools and routines that seek
widespread use by overspecifying behaviors and conversations in
the classroom may help in meeting basic education goals, but
they will not take us far toward the 21st-century star.
Many scholars and practitioners today recognize that to reap
the benefits of more than half a century of cognitive research on
thinking and mental capacity building will require serious attention to how education organizations function as well as to how
individuals learn. Social science research has a long history of
studying organizations. But systematic applications to education
are more recent. And attempts to design education organizations
and test those designs empirically in a continuous cycle of
improvement are still rare. I hope the analysis and examples
offered here will become part of a growing program of education
research that brings the resources of cognitive science, sociocultural research, and organization theory and practice to bear on
our efforts to reach for the star.
Note

This work was supported in part by the National Science Foundation

(Scaling Up Mathematics: The Interface of Curricula With Human and
Social Capital, DRL-0228343) and the Pittsburgh Science of Learning
Center, also funded by the National Science Foundation (SBE0836012).
I wish to thank Elizabeth Rangel for her assistance, both substantive
and editorial, in preparing this article.
1The term literate here refers not just to reading and writing but also
to mathematics and basic knowledge in science, geography, and history.
References

Adey, P., & Shayer, M. (2001). Thinking science. London: Nelson

Thormes.

Adler, P. S., & Kwon, S. (2002). Social capital: Prospects for a new concept. Academy of Management Review, 27, 1740.
Allison, G. T. (1971). Essence of decision. New York: Little, Brown.
Ball, D. L. & Lampert, M. (1998). Multiples of evidence, time, and
perspective: Revising the study of teaching and learning. In E. C.
Lagemann & L. S. Shulman (Eds.), Issues in education research:
Problems and possibilities (pp. 371398). San Francisco: Jossey Bass.
Beck, I. L., & McKeown, M. G. (2002). Questioning the author:
Making sense of social studies. Educational Leadership, 60(3), 4447.
Becker, H. (1964). Against the code of ethics. American Sociological
Review, 29, 409410.
Berliner, D. C. (2006). Educational psychology: Searching for essence
throughout a century of influence. In P. A. Alexander & P. H. Winne
(Eds.), Handbook of educational psychology (2nd ed., pp. 327).
Mahwah, NJ: Lawrence Erlbaum.
Bickel, D. D., & Artz, N. J. (2001, August). Implementing content-focused
coaching as part of a district-wide reform effort. Paper presented at the
biannual meeting of the European Association for Research on
Learning and Instruction (EARLI), Fribourg, Switzerland.
Bill, V. L., Leer, M. N., Reams, L. E., & Resnick, L. B. (1992). From
cupcakes to equations: The structure of discourse in a primary mathematics education classroom. Verbum, 15(1), 6385.
Boyd, D., Grossman, P., Lankford, H., Loeb, S., & Wyckoff, J. (2006).
How changes in entry requirements alter the teacher workforce and affect
student achievement. Education Finance and Policy, 1(2), 176216.
Boyd, D., Grossman, P., Lankford, H., Loeb, S., & Wyckoff, J. (2008,
September). Teacher preparation and student achievement (NBER
Working Paper No. W14314). Cambridge, MA: National Bureau of
Economic Research. Available at http://ssrn.com/abstract=1264576.
Brown, J. S., & Duguid, P. (2000). The social life of information.
Cambridge, MA: Harvard Business School Press.
Bryk, A. S., & Schneider, B. L. (2003). Trust in schools: A core resource for
improvement. New York: Russell Sage Foundation.
Carroll, J. B. (1963). A model of school learning. Teachers College Record,
64, 723733.
Carroll, J. B. (1989) The Carroll model: A 25-year retrospective and
prospective view. Educational Researcher, 18(1), 2631.
Center on Education Policy. (2008). Instructional time in elementary schools:
A closer look at changes for specific subjects. Washington, DC: Author.
Chi, M. T. (2000). Self-explaining expository texts: The dual processes
of generating inferences and repairing mental models. In R. Glaser
(Ed.), Advances in instructional psychology (pp. 161238). Mahwah,
NJ: Lawrence Erlbaum Associates.
Clark, R. M., Sherer, J. Z., Besterfield-Sacre, M., & Resnick, L. (2007,
April). Combining expert directed graphs of the K12 educational system.
Paper presented at the American Educational Research Association
Annual Meeting, Chicago, Illinois.
Clotfelter, C. T., Ladd, H. F., & Vigdor, J. L. (2007). How and why do
teacher credentials matter for student achievement? (NBER Working
Paper No. 12828). Cambridge, MA: National Bureau of Economic
Research.
Cohen, M. D., & Bacdayan, P. (1994). Organizational routines are
stored as procedural memory: Evidence from a laboratory study.
Organizational Science, 5, 554568.
Cohen, D. K., & Hill, H. (2000). Instructional policy and classroom
performance: The mathematics reform in California. Teachers College
Record, 102, 294343.
Coleman, J. S. (1988). Social capital in the creation of human capital.
American Journal of Sociology, 94, S95S120.
Correnti, R., & Rowan, B. (2007). Opening up the black box:
Literacy instruction in schools participating in three comprehensive school reform programs. American Educational Research
Journal, 44, 298338.
april 2010

195

Darling-Hammond, L., & Bransford, J. (Eds.). (2005). Preparing teachers for a changing world: What teachers should learn and be able to do.
San Francisco: Jossey-Bass.
David, J. L., & Greene, D. (2008). Improving mathematics instruction in
Los Angeles high schools: Follow up to the evaluation of the PRISMA Pilot
Program. Palo Alto, CA: Bay Area Research Group.
Davis, L. E. (1982). Organizational design. In G. Salvendy (Ed.),
Handbook of industrial engineering (pp. 2.1.12.1.29). New York:
John Wiley.
Dweck, C. S., & Molden, D. C. (2005). Self-theories: Their impact on
competence motivation and acquisition. In A. Elliot & C. Dweck
(Eds.) Handbook of competence and motivation (pp. 122140). New
York: Guilford.
Feldman, M. S. (2000). Organizational routines as a source of continuous change. Organization Science, 11, 611629.
Feldman, M. S., & Pentland, B. T. (2003). Reconceptualizing organizational routines as a source of flexibility and change. Administrative
Science Quarterly, 48, 94118.
Frank, K. A., Zhao, Y., & Borman, K. (2004). Social capital and the
diffusion of innovations within organizations: The case of computer
technology in schools. Sociology of Education, 77, 148171.
Gamoran, A., Anderson, C. W., Quiroz, P. A., Secada, W. G., Williams,
T., & Ashmann, S. (2003). Transforming teaching in math and science:
How schools and districts can support change. New York: Teachers
College Press.
Gersick, G. J., & Hackman, J. R. (1990). Habitual routines in taskperforming groups. Organizational Behavior and Human Decision
Process, 47, 6597.
Grossman, P., & McDonald, M. (2008). Back to the future: Directions
for research in teaching and teacher education. American Educational
Research Journal, 45, 184205.
Harbison, R., & Hanushek, E. (1992). Educational performance for the poor:
Lesson from rural northeast Brazil. Oxford, UK: Oxford University Press.
Hertzberg, H. (1988). Foundations. The 1982 Committee of Ten. Social
Education, 52(2), 4445.
Hill, H., Rowan, B., & Ball, D. (2005). Effects of teachers mathematic
knowledge for teaching on student achievement. American Educational
Research Journal, 42, 371406.
Jaeger, R. M., & Tittle, C. K. (1980). Minimum competency testing:
Motives, models, measures and consequences. Berkeley, CA: McCutchan.
Johnson, R. T., & Johnson, D. W. (1986). Action research: Cooperative
learning in the science classroom. Science and Children, 24, 3132.
Koretz, D., & Hamilton, L. (2006). Testing for accountability in K12.
In R. L. Brennan (Ed.), Educational measurement (5th ed., pp. 531
578). Westport, CT: American Council on Education/Praeger.
Lampert, M. (2001). Teaching problems and the problems of teaching.
New Haven, CT: Yale University Press.
Lampert, M., & Ball, D. (1998). Teaching, multimedia, and mathematics:
Investigations of real practice. New York: Teachers College Press.
Leana, C., & Pil, F. (2009). Applying organizational research to public
school reform: The effects of teacher human and social capital on
student performance. Academy of Management Journal.
Loeb, S., Darling-Hammond, L., & Luczak, J. (2005). How teaching
conditions predict teacher turnover in California schools. Peabody
Journal of Education, 80(3), 4470.
March, J. G. (1981). Footnotes to organizational change. Administrative
Science Quarterly, 26, 563577.
March, J. G., & Simon, H. A. (1958). Organizations. New York: John Wiley.
March, J. G., & Simon, H. A. (with H. Guetzkow). (1993). Organizations
(2nd ed.). Cambridge, MA: Blackwell.
Matsumura, L. C., Garnier, H., & Resnick, L. B. (in press). Implementing
literacy coaching: The role of school social resources. Educational
Evaluation and Policy Analysis.
196

educational Researcher

Matsumura, L. C., Sartoris, M., Bickel, D., & Garnier, H. (2009).

Leadership for literacy coaching: The principals role in launching a new
coaching program. Educational Administration Quarterly, 45, 655694.
McConachie, S.M., & Petrosky, A.R. (2010). Content matters: A disciplinary literacy approach to improving student learning. San Francisco:
Jossey-Bass.
McLaughlin, M. W., & Talbert, J. E. (2001). Professional communities
and the work of high school teaching. Chicago: University of Chicago
Press.
McNeill, L. (2002). Contradictions of school reform: Educational costs of
educational testing. New York: Routledge.
Mehan, H. (1979). Learning lessons. Cambridge, MA: Harvard
University Press.
Mercer, N., Dawes, L., Wegerif, R., & Sams, C. (2004). Reasoning as a
scientist: Ways of helping children to use language to learn science.
British Educational Research Journal, 30, 359377.
Michaels, S., OConnor, C., & Resnick, L. B. (2007). Deliberative discourse idealized and realized: Accountable talk in the classroom and
in civic life. Studies in Philosophy and Education, 27, 283297.
Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital
and the organizational advantage. Academy of Management Review,
23, 242266.
National Academy of Education. (2009). Education policy white paper:
Teacher quality. Washington, DC: Author.
National Commission on Excellence in Education. (1983). A nation at
risk: The imperative for educational reform. Washington, DC: Author.
National Education Goals Panel. (1991). The National Education Goals
report 1991: Building a nation of learners. Washington, DC: Author.
National Institute for Literacy. (2008). Developing early literacy: Report
of the National Early Literacy Panel. Jessup, MD: Author.
Newman, F. M. (1996). Authentic achievement: Restructuring schools for
intellectual quality. San Francisco, CA: Jossey-Boss.
OConnor, M. C. (2001). Can any fraction be turned into a decimal?
A case study of a mathematical group discussion. Educational Studies
in Mathematics, 46, 143185.
Orr, J. (1996). Talking about machines. Ithaca, NY: Cornell University Press.
Resnick, D. P., & Resnick, L. B. (1977). The nature of literacy: An historical exploration. Harvard Educational Review, 47, 370385.
Resnick, L. B. (1987a). Learning in school and out. Educational
Researcher, 16(9), 1320.
Resnick, L. B. (1987b). The Thinking Curriculum. Washington, DC:
National Academy Press.
Resnick, L. B. (1999). From aptitude to effort: A new foundation for our
schools. American Educator, 23, 1417.
Resnick, L. B., Besterfield-Sacre, M., Mehalik, M. M., Sherer, J. Z.,
& Halverson, E. R. (2007). A framework for effective management
of school system performance. In P. A. Moss (Ed.). Evidence and
decision making: The 106th yearbook of the National Society for the
Study of Education (NSSE) (Part I, pp. 155185). Malden, MA:
Blackwell.
Resnick, L. B., & Klopfer, L. E. (Eds.). (1989). Toward the Thinking
Curriculum: Current cognitive research (ASCD Yearbook). Alexandria,
VA: Association for Supervision and Curriculum Development.
Resnick, L. B., Michaels, S., & OConnor, C. (in press). How (wellstructured) talk builds the mind. In R. Sternberg & D. Preiss (Eds.),
From genes to context: New discoveries about learning from educational
research and their applications. New York: Springer.
Resnick, L. B., & Nelson-Le Gall, S. (1997). Socializing intelligence. In
L. Smith, J. Dockrell, & P. Tomlinson (Eds.), Piaget, Vygotsky and
beyond (pp. 145158). New York: Routledge.
Resnick, L. B., Saljo, R., Pontecorvo, C., & Burge, B. (Eds.). (1997).
Discourse, tools, and reasoning: Essays on situated cognition. Berlin:
Springer-Verlag.

Resnick, L. B., & Spillane, J. P. (2006). From individual learning to

organizational designs for learning. In L. Verschaffel, F. Dochy,
M. Boekaerts, & S. Vosniadou (Eds.), Instructional psychology: Past,
present and future trends. Sixteen essays in honor of Erik De Corte
(pp. 259276). Oxford, UK: Pergamon.
Resnick, L. B., Spillane, J., Goldman, P., & Rangel, E. (in press).
Implementing innovation: From visionary models to everyday practice. In D. Istance, F. Benavides, & H. Dumont (Eds.), Innovative
learning environments. Paris: Organisation for Economic Co-operation
and Development.
Resnick, L. B., Wiliam, D., Apodaca, R., & Rangel, E. (in press). The
relationship between assessment and the organization and practice of
teaching. In B. McGraw, P. Peterson, & E. Baker (Eds.), International
encyclopedia of education (3rd ed.). London: Elsevier.
Rowan, B., & Correnti, R. (2009). Studying reading instruction with
teacher logs: Lessons from a study of instructional improvement.
Educational Researcher, 38, 120131.
Sahney, V. K. (1993). Evolution of hospital industrial engineering: from
scientific management to total quality management. Journal of the
Society of Health Systems, 4(1), 317.
Shepard, L. (20022003, Winter). The hazards of high stakes testing.
Issues in Science and Technology, pp. 5358.
Sherer, J., & Spillane, J. (in press.) Constancy and change in work
practice in schools: The role of organizational routines. Teachers
College Record.
Shulman, L. S. (1987). Knowledge and teaching: Foundations of the
new reform. Harvard Educational Review, 57(1), 122.
Slavin, R. E. (1996). Research on cooperative learning and achievement:
What we know, what we need to know. Contemporary Educational
Psychology, 21, 4369.
Spillane, J. P., Mesler, L., Sherer, J. Z., & Croegaert, C. (2010). Organi
zational routines as recoupling mechanisms: Policy, school administration,
and the technical core. Manuscript submitted for publication.

Stigler, J., & Hiebert, J. (1999). The teaching gap: Best ideas from the
worlds teachers for improving education in the classroom. New York:
Free Press.
Suchman, L. (1996). Constituting shared workspaces. In. Y. Engestrm
& D. Middleton (Eds.), Cognition and communication at work
(pp. 3560). Cambridge, UK: Cambridge University Press.
Talbert, J. E., David, J. L., with Lin, W. (2008, September). Evaluation
of the Disciplinary Literacy-Professional Learning Community (DL-PLC)
Initiative in Austin Independent School District. Final Report. Palo Alto,
CA: Center for Research on the Content of Teaching, Stanford
University.
Thorndike, E. L. (1932). Fundamentals of learning. New York: Teachers
College, Columbia University.
Turner, W. C., Mize, J. H., Case, K. E., & Nazemetz, J. W. (1992).
Introduction to industrial and systems engineering (3rd ed.). Upper
Saddle River, NJ: Prentice Hall.
Tyack, D. B. (1974). The one best system: A history of American urban
education. Cambridge, MA: Harvard University Press.
AUTHOR

LAUREN B. RESNICK is Distinguished University Professor of

Psychology and Cognitive Science and of Learning Sciences and
Education Policy, former director of the Learning Research and
Development Center, and founder and director of the Institute for
Learning at the University of Pittsburgh, Learning Research and
Development Center, 3939 OHara Street, Pittsburgh, PA 15260;
[email protected]. Her current research focuses on school reform, assessment, effort-based education, the nature and development of thinking
abilities, and the role of talk and discourse in learning.
Manuscript received January 10, 2010
Revision received January 15, 2010
Accepted January 25, 2010

april 2010

197