Scaffolding Complex Learning: The
Mechanisms of Structuring and
Problematizing Student Work
Brian J. Reiser
School of Education and Social Policy
There has been much interest in using software tools to scaffold learners in complex
tasks, that is, to provide supports that enable students to deal with more complex con-
tent and skill demands than they could otherwise handle. Many different approaches
to scaffolding techniques have been presented in a broad range of software tools. I ar-
gue that two complementary mechanisms can explain how a diversity of scaffolding
approaches in software act to support learners. Software tools can help structure the
learning task, guiding learners through key components and supporting their plan-
ning and performance. In addition, tools can shape students’ performance and under-
standing of the task in terms of key disciplinary content and strategies and thus
problematize this important content. Although making the task more difficult in the
short term, by forcing learners to engage with this complexity, such scaffolded tools
make this work more productive opportunities for learning. I present arguments for
these mechanisms in terms of the obstacles learners face, and I present several brief
examples to illustrate their use in design guidelines. Finally, I examine how the
mechanisms of structuring and problematizing are sometimes complementary and
sometimes in tension in design, discuss design tradeoffs in developing scaffolded in-
vestigation tools for learners, and consider the reliance of scaffolding on a classroom
system of supports.
There is much interest in education reform in using technology to support learners.
One aspect of the argument for technology has been that software can be used to help
learners succeed in more complex tasks than they could otherwise master (Davis &
THE JOURNAL OF THE LEARNING SCIENCES, 13(3), 273–304
Copyright © 2004, Lawrence Erlbaum Associates, Inc.
Correspondence and requests for reprints should be sent to Brian J. Reiser, School of Education and
Social Policy, Northwestern University, 2120 Campus Drive, Evanston, IL 60208. E-mail:
Linn, 2000; Edelson, Gordin, & Pea, 1999; Guzdial, 1994; Quintana, Eng, Carra,
Wu, & Soloway, 1999; Reiser et al., 2001). Researchers have invoked the notion of
scaffolding, a construct originally crafted to characterize how more experienced
peers or adults can assist learners. As defined and used in early research, scaffolding
is said to occur when a more knowledgeable person helps a learner succeed in tasks
that would be otherwise beyond their reach (Wood, Bruner,& Ross, 1976). In the last
two decades of learning sciences research, scaffolding has become increasingly
prominent. Scaffolding is a key strategy in cognitive apprenticeship, in which stu-
dents can learn by taking increasing responsibility and ownership for their role in
complex problem solving with the structure and guidance of more knowledgeable
mentors or teachers (Collins, Brown, & Newman, 1989).
Many different approaches to scaffolding have emerged from the design re-
search on interactive learning environments, and a variety of design guidelines or
principles have been proposed (Edelson et al., 1999; Guzdial, 1994; Kolodner,
Owensby, & Guzdial, 2004; Linn, 2000; Reiser et al., 2001). To engage in princi-
pled development and empirical study of design guidelines requires greater clarity
concerning what is meant when one says that a tool has scaffolded learners, and
requires a model of how the tool has benefited learners. In particular, it is important
to characterize the mechanisms by which a software tool can provide scaffolding
for learners. Developing a common system of design guidelines for scaffolded
software requires such a model of mechanisms that explain why a tool reflecting
these guidelines would benefit learners.
In this article, I present an analysis of two general mechanisms to character-
ize how scaffolded tools can support learning. I describe how these dual mecha-
nisms can address the challenges learners face by structuring tasks to make them
more tractable and to shape tasks for learners in ways that makes their problem
solving more productive. I develop the argument for these mechanisms by first
considering how tools affect the experience of tasks for learners. Then I review
some of the critical challenges learners face in complex domains such as science
and mathematics learning. In describing each mechanism, I present brief exam-
ples of software environments to illustrate the mechanisms in practice. Finally, I
consider how the mechanisms can interact and discuss issues of the embedding
of tools in classroom contexts.
TRADITIONAL APPROACHES TO SCAFFOLDING
To consider how software tools can scaffold learners, I first review the source of
the scaffolding metaphor. The term scaffolding has traditionally been used to refer
to the process by which a teacher or more knowledgeable peer assists a learner, al-
tering the learning task so the learner can solve problems or accomplish tasks that
would otherwise be out of reach (Collins et al., 1989; Wood et al., 1976). The cen-
tral component of this definition is that another person intervenes at times appro-
priate for that learner in that context, and what the learner can accomplish in-
creases with these interventions. For example, a teacher may help a child in a board
game by reminding him or her of the rules or by suggesting strategic steps if the
child is stuck. The conception is associated with Vygotsky’s (1978) notion of the
zone of proximal development, which characterizes the region of tasks between
what the learner could accomplish alone and what he or she could accomplish (and
master) with assistance (Rogoff, 1990).
The idea of scaffolding is now in increasing use in educational design. In these
contexts, the intention is that the support not only assists learners in accomplishing
tasks but also enables them to learn from the experience. The use of the notion of
scaffolding has not always been explicitly limited to learning settings. For exam-
ple, one might consider an adult providing support to a child for some task (such as
observing an animal at the zoo) in which there is no intention that the child learns
to perform the task in the future more effectively. For educational settings, it is im-
portant to stress the dual aspects of both (a) accomplishing the task and (b) learning
from one’s efforts, that is, improving one’s performance on the future tasks in the
process. If learners are assisted in the task but are not able to understand or take ad-
vantage of the experience, the assistance will have been local to that instance of
scaffolding but will not have provided support for learning. Thus, scaffolding en-
tails a delicate negotiation between providing support and continuing to engage
learners actively in the process (Hogan, Nastasi, & Pressley, 1999; Merrill, Reiser,
Merrill, & Landes, 1995). Lepper, Woolverton, Mumme, and Gurtner (1993) de-
scribed this as maintaining an “optimum” level of challenge for learners. I return to
the need for balancing assistance with ensuring the work on the task is productive
in later discussions of the two scaffolding mechanisms.
Recent design research on interactive learning environments has adapted the
notion of scaffolding (Davis & Linn, 2000; Edelson et al., 1999; Guzdial, 1994;
Quintana et al., 1999; Reiser et al., 2001). This vision of scaffolding refers to
ways the software tool itself can support learners rather than only teachers or
peers. As applied to software, scaffolding refers to cases in which the tool
changes the task in some way so that learners can accomplish tasks that would
otherwise be out of their reach. Software scaffolding provides some aspect of
support that helps make the learning more tractable for learners. For example,
the software might provide prompts to encourage or remind students what steps
to take (Davis & Linn, 2000), graphical organizers or other notations to help stu-
dents plan and organize their problem solving (Quintana et al., 1999), or repre-
sentations that help learners track what steps they have taken (Collins & Brown,
1988; Koedinger & Anderson, 1993). In all these cases, the software provides
additional assistance beyond what a simpler, more basic tool would have pro-
vided to allow learners to accomplish more ambitious tasks. Sherin, Reiser, and
Edelson (this issue) argue that software scaffolding should be characterized in
STRUCTURING AND PROBLEMATIZING 275
terms of the differences the scaffolding creates in comparison to some presum-
ably more difficult reference version of the task.
This work on scaffolded software tools has been very encouraging, and scaffold-
ing promises to be an important benefit in integrating technological tools into class-
rooms. However, there has been a wide range of approaches to designing software
scaffolds and many kinds of design principles. Integrating different scaffolding ap-
proaches into a common framework requires an analysis of how scaffolding can oc-
cur in the interactions between learners and software tools. How can researchers
characterize the mechanisms by which software scaffolding assists learners?
Although there have been many different design principles proposed, I argue
that underlying these principles are some common assumptions about how to make
more productive learning experiences for students. The focus of this article is to
consider design arguments and principles that have been proposed for software
scaffolding and to characterize the common mechanisms by which these strategies
achieve benefits for learners. Such an analysis of mechanisms is needed to clarify
and evaluate what types of scaffolding are effective. In this work, I build on the
framework developed by Quintana et al. (this issue). This framework consists of a
set of design guidelines and specific design strategies that synthesize design ideas
across a range of software tools and grounds these strategies in the types of obsta-
cles learners need to overcome. The scaffolding mechanisms proposed here are
meant to explain why these design guidelines work to support learners in master-
ing complex tasks.
To construct this argument, I focus on scaffolding in the discipline of science.
Much of the work on scaffolding tools has taken place in this domain, and there is a
rich literature on the obstacles learners face. Furthermore, tools to access and inter-
pret data are a central part of the practices of scientific investigation, so this domain
is a productive context in which to explore the design of scaffolded tools.
NEEDS OF LEARNERS
A principled analysis of the manners in which tools can influence learning must
begin with an analysis of the needs of learners and the ways that shaping the tool
can affect the ability of learners to overcome these challenges. In this section, I
briefly consider the challenges of learners in the discipline of science to delineate
the opportunities for a software tool to help learners overcome these challenges.
Instructional approaches in science emphasize learning by engaging in knowl-
edge construction practices. In the case of science, this entails learning science
through investigation and argumentation (Olson & Loucks-Horsley, 2000). In pro-
ject-based science, students learn general principles in the context of investigating
particular problem scenarios such as learning introductory chemistry by analyzing
the quality of air in the local community (Blumenfeld et al., 1991; Edelson, 2001;
Hmelo, Holton, & Kolodner, 2000). In addition to constructing conceptual under-
standing, students need to acquire new disciplinary strategies to guide reasoning in
the domain (Schauble, Glaser, Raghavan, & Reiner, 1991; Tabak, this issue).
These approaches to learning through inquiry, although providing the potential
to connect knowledge more effectively to real-world contexts, also pose particular
challenges for learners. Quintana et al. (this issue) consider the challenges learners
face and organize them around three constituent processes involved in learning
through scientific investigation—sense making, process management, and articu-
lation and reflection. Each type of process is challenging for learners.
Sense making entails constructing and interpreting empirical tests of hypothe-
ses. Students need to coordinate their reasoning about experiments or data compar-
isons with the implications of the findings for an explanation of the scientific phe-
nomena. This coordination and mapping task is complex and requires rich subject
matter knowledge to design data comparisons and interpret findings in light of hy-
potheses (Klahr & Dunbar, 1988; Schauble et al., 1991).
Strategic guidance is critical to manage the complex investigation process. In-
vestigations require an iterative processes of designing an investigation, collecting
data, constructing and revising explanations based on data, evaluating explana-
tions, and communicating arguments (Olson & Loucks-Horsley, 2000). This re-
quires strategic knowledge to plan, conduct investigations, and make decisions
about next steps based on interim results. These require both discipline-specific
processes and content knowledge that may be new to learners.
Finally, investigations require the complementary processes of reflection and
articulation as students monitor and evaluate their progress, reconsider and refine
their plans, and articulate their understanding as they proceed. These communica-
tive and metacognitive skills pose additional challenges for learners.
Thus, learners face challenges at several levels. Students must master concep-
tual knowledge, domain process skills, domain-specific strategies, and more
general metacognitive processes. In addition to cognitive and metacognitive
challenges, these practices include a social dimension, as investigations involve
working together in teams, planning and negotiating within a group, communi-
cating, and debating with peers about scientific interpretations (Brown &
Campione, 1994). These social practices and the discourse practices they entail
are potentially unfamiliar, and pose additional social interaction challenges for
learners (Webb & Palincsar, 1996). Next, I consider specific obstacles that arise
as learners grapple with these challenges.
Sophisticated problem solving relies on strategies for planning and guiding rea-
soning. These heuristic strategies in science are needed to plan investigations, se-
lect data comparisons, and synthesize findings. These strategies involve general
STRUCTURING AND PROBLEMATIZING 277
strategies for scientific inquiry (Klahr, 2000; Krajcik et al., 1998; White &
Frederiksen, 1998) and discipline-specific explanatory frameworks (Passmore &
Stewart, 2002; Reiser et al., 2001; Sandoval & Reiser, in press; Tabak, Smith,
Sandoval, & Reiser, 1996). A key challenge is that this knowledge is typically tacit
for more experienced reasoners and may be taken for granted. Instruction often
fails to make these strategies explicit for learners. Learners experience challenges
in using general strategies for designing empirical tests of hypotheses (Klahr, Fay,
& Dunbar, 1993) and in using specific domain knowledge to plan and guide inves-
tigations (Schauble et al., 1991).
Learners tend to focus on products rather than on explanatory and learning goals
(Perkins, 1998; Schauble, Glaser, Duschl, Schulze, & John, 1995). For example,
they focus on achieving desired results rather than on understanding the principles
behind the results and become distracted by superficial aspects of the products they
need to construct (Krajcik et al., 1998). The difficulty in managing investigations
leads to insufficient attention devoted to reflection and reevaluation (Loh, 2003;
Loh et al., 2001). Lack of content knowledge further complicates the process of
evaluating the progress of an investigation. Another aspect to this challenge is that
learners may need assistance in generalizing appropriately from their work on spe-
cific problem scenarios. For learning through investigation to succeed, students
must not only construct solutions to the particular scenario but must connect the
explanations or arguments they construct to more general disciplinary frameworks
Fragile and Superficial Understanding
Learners tend to focus on superficial details and have difficulty seeing the under-
lying structure that is visible with more experience (Chi, Feltovich, & Glaser,
1981). They may have difficulty mapping between their intuitive understandings
and more precise scientific constructs and to formal representations that are the
medium for representing work in the domain (Reif & Larkin, 1991; Sherin,
2001). Furthermore, they may be too quick to decide interpretations are war-
ranted without sufficient evaluation of alternatives (Klahr, 2000; Kuhn, Amsel,
& O’Loughlin, 1988). Learners are not always effective in analyzing whether
they have understood and may be overconfident in their self-assessments (Chi,
Bassok, Lewis, Reimann, & Glaser, 1989; Davis, 2003).
Unfamiliar Social Interaction Practices
Scientific investigations require practices that include both cognitive and social in-
teraction components. These include constructing scientific arguments to persuade
peers, receiving questions and critiques, and improving explanations based on
feedback (Bell & Linn, 2000; de Vries, Lund, & Baker, 2002; Driver, Newton, &
Osborne, 2000; Kuhn, 1993; Sandoval & Reiser, in press). These processes are
typically conducted in classrooms in teams, requiring collaborative planning, ne-
gotiation, and self-assessment. This complex collaborative work presents social
interaction challenges such as weighing opinions and keeping track of alternatives
proposed by all group members, ensuring participation from all group members,
and learning to offer and receive critiques (Coleman, 1998; Webb & Palincsar,
1996). The success of group work can be compromised by difficulties in these so-
cial interactions (Barron, 2003; Kurth, Anderson, & Palincsar, 2002).
Unfamiliar Discourse Practices
The third component of practices, in addition to cognitive strategies and social in-
teractions, consists of characteristic discourse practices. Discourse plays a particu-
larly central role in the practice of scientific inquiry and places demands on learn-
ers (Lemke, 1990). Scientific practices are implemented in particular discourse
practices and uses of language—for example, for expressing hypotheses, arguing
from evidence, critiquing an idea, and so on—which may be unfamiliar (Rosebery,
Warren, & Conant, 1992). The linguistic and cognitive aspects of these practices
are interrelated, with important constructs signaled by language and precise lan-
guage used to communicate about scientific processes (e.g., “support,” “argue,”
“falsify,” etc.). There may be a disconnect between some learners’ understanding
of the practices underlying language use and the scientifically accepted practice
(Moje, Collazo, Carrillo, & Marx, 2001; Reif & Larkin, 1991). Learners may need
support in using language appropriately and in connecting the language with im-
plementation in scientific practices.
In summary, the task demands of engaging in scientific investigations reveal a
system of challenges for learners. Difficulties arise from the cognitive complex-
ity of the practices as well as from new social interaction and discourse chal-
lenges. The cognitive complexity arises from unfamiliar general and disci-
pline-specific strategies that are required in sense making and process
management. Learners may not articulate their ongoing understanding, focusing
instead on pragmatic goals of creating required products. They may need to be
prompted to be more reflective and focused on understanding rather than perfor-
mance and to go beyond superficial solutions to problems. Students may need to
become proficient in new practices for social interaction and discourse associ-
ated with knowledge building. The focus on public knowledge building and the
specialized uses of language it requires pose an ongoing challenge during the in-
vestigation. They may need assistance in scientific discourse to move beyond de-
scription and communicate a scientific argument.
STRUCTURING AND PROBLEMATIZING 279
In the next section, I consider how software tools influence the nature of tasks for
learners. With that as a foundation, I then propose specific mechanisms to explain
how software scaffolding can assist with the specific challenges identified here.
HOW CAN TOOLS HELP LEARNERS?
Traditional views of scaffolding have focused on interactions with teachers or
peers as the source of assistance, articulating how a more knowledgeable person
can provide assistance in the context of a task (Hogan & Pressley, 1997; Wood et
al., 1976). The focus of the last two decades of research on the learning sciences is-
sues in technology design has illuminated ways in which technological tools may
provide some types of scaffolding functions. Instructional designers have investi-
gated how to create tools that can help learners accomplish complex tasks. In con-
sidering how to design effective scaffolded tools, it is important to reconceptualize
the learning problem from that of an individual working on tasks, perhaps with as-
sistance of another more knowledgeable person, to a consideration of the context
in which the people are acting, the tools they use, and the knowledge embedded in
this context. Rather than considering what the individual can accomplish, this view
of distributed cognition focuses on what a person or group working with tools can
accomplish as a system (Hollan, Hutchins, & Kirsh, 2000). The structure of a tool
shapes how people interact with the task and affects what can be accomplished. I
consider how the nature of the tool can alter the task facing learners.
Tools Can Distribute Work and Reduce What Is Required of
One clear way that tools change the nature of tasks for learners is in automating as-
pects of tasks and thereby limiting the part of the task the learners need to perform,
potentially enabling them to focus on more productive parts of the tasks (Salomon,
Perkins, & Globerson, 1991). Salomon et al. considered how the partnership of per-
son and tool can accomplish tasks together that extend beyond what the person could
accomplish alone. This is the perhaps the most straightforward sense of scaffolding.
For example, calculators can offload simple computations, allowing people to focus
on other parts of the data manipulation tasks such as considering what calculations to
compose together to solve a problem. Word processors with spelling checkers can al-
low writers to focus more on the construction of their prose rather than devotingtime
to checking spelling in dictionaries. The result of offloading aspects of the task may
be to reduce the overall complexity. If offloading these aspects of the task allows
learners to focus more effectivelyon the conceptually important aspects and thereby
learn from their experience, the tool has scaffolded that learning.
Tools Can Transform Tasks
Tools can have even more dramatic effects in the way they transform the nature of
the tasks. In fact, the tools people use can be a critical factor in how people envision
and engage in the tasks they perform (Hutchins, 1995; Norman, 1991). This is par-
ticularly true when tasks involve accessing, manipulating, storing, or reasoning
about information. Norman (1991) described cognitive artifacts, or tools that are
used to represent and manipulate information in a task. Cognitive artifacts can
change the task in fundamental ways. Users translate their intentions into actions
to be taken in the tool and construct an understanding of the state of the external
world from the representation the tool provides. The tool provides an inscription or
encoding of information about the world. In this way, the tool provides the repre-
sentation of external states and the vehicle for operating on the environment. Be-
cause of the central role of tools in effecting actions in information domains, the
task cannot be defined independently of the tools that people use in the practice of
that task (Bannon & Bødker, 1987).
The framework of distributed cognition has been used to describe these
transformative tools (Collins, 1991; Hollan et al., 2000; Pea, 1992; Pea & Gomez,
1992). In this framework, tools affect and may extend what users can do. Tools may
make some types of computation unnecessary for users (such as symbolic calcula-
tors) or may create inscriptions that encode information in a more usable form. Be-
cause cognitive artifactsmediate between people and the world, such tools can trans-
form the task in the way in which they represent and allow people to manipulate
information. If the inscription provided by the tool enables useful inferences more
effectively, it can extend the range of what users can do. Visualization tools that pro-
vide conceptually meaningful representations are designed to help users form deep
models of an underlying system (Hollan, Bederson, & Helfman, 1997). For exam-
ple, studying atmospheric sciences phenomena using tools to access and construct
sophisticated visualizations of primary data enables scientists to construct and test
conjectures about the phenomena. These specialized inscriptions are designed to en-
able users to notice patterns in data and fundamentally alter the kinds and level of sci-
entific reasoning they can do about the material (Edelson et al., 1999). Similarly, di-
rect-manipulation interfaces allow users to control a process by appearing to act on it
directly through a visual metaphor (Hutchins, Hollan, & Norman, 1986).
The tools can also affect the nature of interactions between collaborators. For
example, the structure of the tool may influence the design of the task into constit-
uents, influencing how roles are defined. The structure of the tool may also influ-
ence the focus of conversation between collaborators. For example, through its or-
ganization of functionality, the tool may focus conversation on particular choices
or may highlight certain relations and influence the course of decision making by
providing a vehicle for students to articulate aspects of their understanding
(Scardamalia & Bereiter, 1994).
STRUCTURING AND PROBLEMATIZING 281
Thus, the design of the tool itself is an important constituent in defining tasks.
The nature of the task emerges from the interactions of people, subject matter, and
tools. The nature of the tool clearly affects how tractable the performance of a task
is for learners.
Norman (1991) described the challenges of designing effective tools for com-
plex tasks. Because tools mediate users’ interactions with the environment, users
need to map between understanding, a tool’s representation, and the world it repre-
sents. Difficulties in this mapping create challenges and can lead to extra steps or
levels of indirection in reasoning. The goal of human–computer interaction design
research is to make that mapping between a tool’s representation and what it repre-
sents “transparent.” That is, users should be able to “see” the core meaning in a rep-
resentation (such as noticing which regions are hotter by comparing red and blue
colors on a temperature map) rather than getting bogged down in reasoning to
translate from the representation to its underlying meaning. Simplifying this trans-
lation can reduce the complexity of the task, minimize errors, and extend what is
possible with the tool (Norman, 1993).
In the design of scaffolded tools, instructional designers can use this mapping to
an advantage. Rather than striving only for transparency between the representa-
tion and the world it represents, designers can bend that representation to instruc-
tional purposes. Cognitive artifacts provide a lever for designers to shape how
learners think about tasks. In the next section, I consider how cognitive artifacts
can be used to instructional advantage.
MECHANISMS OF SCAFFOLDING: STRUCTURE AND
I have discussed how changing the nature of tools people use can fundamentally
transform the task, determining the cognitive and social interaction demands of a
task. Now I consider the kinds of transformation involved in tools that are designed
to scaffold learners.
How can software tools provide scaffolding? First, I revisit the definition of
scaffolding as applied to science. There are two critical notions in scaffolding: (a)
learners receive assistance to succeed in more complex tasks that would otherwise
be too difficult, and (b) learners draw from that experience and improve in process
skills and/or content understanding. Focusing on the representational properties of
tools provides a dimension on which to consider how the specific design of the tool
can support learning tasks.
I propose two complementary mechanisms of scaffolding in software tools—it
can help structure the task of problem solving, and it can problematize subject mat-
ter, and thus provoke learners to devote resources to issues they might not other-
wise address. I describe each mechanism and the nature of the scaffolding influ-
ence on learners. To illustrate each mechanism in action, I present an example
learning environment and its associated design guidelines that rely on that manner
of support for learners. I then discuss some of the design tradeoffs and tensions that
exist in utilizing these two mechanisms.
The roots of these ideas of structuring and problematizing are in traditional no-
tions of scaffolding as developed with parents, tutors, and teachers. However, al-
though both of these ideas are implicit in the underlying designs of software tools
developed in design research, the function of problematizing subject matter has
not received much attention in the design arguments and theoretical accounts of
software scaffolding. My goal in this article is to help make explicit both of these
mechanisms in theoretical accounts of software scaffolding.
Structuring the Task
The first sense of scaffolding is the most common in design research on scaffold-
ing, and it is the most straightforward. If reasoning is difficult due to complexity or
the open-ended nature of the task, then one way to help learners is to use the tool to
reduce complexity and choice by providing additional structure to the task. For ex-
ample, this may be done by providing structured work spaces to help learners de-
compose a task and organize their work or prompts to help learners recognize im-
portant goals to pursue.
The notion of supporting problem solving by providing more structure has been
a central constituent of scaffolding theories beginning with the analyses of Wood
et al. (1976). Wood et al. characterized scaffolding as involving “reduction in de-
grees of freedom” and support to help learners “maintain direction.” The core idea
is that by providing structure or constraints, perhaps in the form of explicit direc-
tion or by narrowing choices, the complexity facing the learner is reduced and the
problem solving is more tractable.
Influence of Structuring Student Work
There are several types of learner challenges for which this approach to struc-
turing work can help learners including decomposing tasks, focusing effort, and
Decomposing complex tasks.
The tool’s interface can be organized to help
learners decompose open-ended problems. For example, Model-It™, a tool for
helping learners construct qualitative models (represented as causal maps), uses
functional modes to structure the task of modeling into plan, build, and test pro-
cesses and organizes relevant functions in each of those modes (Jackson, Krajcik,
& Soloway, 1998). This can help guide what actions to take, their order, or neces-
sary aspects of work products. In this way, the software tool helps overcome some
STRUCTURING AND PROBLEMATIZING 283
of the obstacles of unfamiliar strategies by helping the learners attend to important
goals. It may also help nonreflective work by helping learners keep track of what
goals have been addressed and what aspects of the task are pending. Several in-
quiry support systems use checklists or diagrams to help learners identify and im-
plement aspects of the process they may otherwise neglect to perform (Davis,
2003; Davis & Linn, 2000; Linn, Bell, & Davis, 2004; Quintana et al., 1999). This
assistance with decomposition can be a resource groups can use to organize work
Restricting the problem space, for example, by narrowing
options, preselecting data, or offloading more routine parts of the task, can help
learners focus resources on the aspects of the task more productive for learning.
For example, visualization software for learners may include specialized tools that
automatically access anchoring reference information (such as names of countries
or cities) not typically represented in scientific tools for experts (Edelson et al.,
1999). Functionality can be organized to restrict options to those relevant to the
learner’s current goals, again helping learners focus resources in productive ways
(Quintana et al., 1999). Like decomposition, these design strategies can help learn-
ers overcome obstacles of unfamiliar strategies by reducing the overload experi-
enced in handling strategic decisions concurrent with managing the implementa-
tion of a plan. The support in the tool that helps focus effort may provide a resource
to help learners work together more effectively. For example, guidance in structur-
ing the work may help reduce options and ambiguities that groups face when nego-
tiating possible directions (Barron, 2003; Krajcik et al., 1998).
Explicit structures such as prompts, agendas, or graphical orga-
nizers can help learners keep track of their plans and monitor their progress. This
type of structuring is characteristic of a number of scaffolding approaches. Guidance
embedded in software can remind learners of important goals to pursue or criteria to
apply to their work. In the Knowledge IntegrationEnvironment, prompts can remind
learners of important criteria they should apply to their work (Davis, 2003; Davis &
Linn, 2000). In these cases, the scaffolded tool addresses the tendency toward
nonreflective work by helping learners construct their plans, consider the possible
actions relevant to each stage of the process, monitor the plan, and tie in relevant dis-
ciplinary ideas as they make sense and communicate about their data.
Examples of Structuring Student Work
The following example illustrates how a software environment can support
learners through the scaffolding mechanism of structuring their work. I de-
scribe a learning environment and the design guidelines it represents and con-
sider how the guideline acts to structure student work. The goal is to illustrate
this mechanism rather than to present an extensive review of types of design
guidelines that utilize the mechanism (see Quintana et al., this issue, for a de-
sign guidelines framework that synthesizes a broad range of design approaches
from the field).
One approach for structuring student work is represented by the design guide-
line “facilitate ongoing articulation and reflection during the investigation”
(“Guideline 7” in Quintana et al., this issue). To see this guideline and its realiza-
tion of structuring, consider the Biology Guided Inquiry Learning Environments
(BGuILE) software tool ExplanationConstructor (Reiser et al., 2001; Sandoval,
1998, 2003; Sandoval & Reiser, in press). ExplanationConstructor is a com-
puter-based science journal in which students construct their scientific explana-
tions. While working on an investigation, students record their research questions
and new subquestions as they emerge, construct candidate explanations and asso-
ciate them with their research questions, and record evidence for each assertion
(see Figure 1). Students use ExplanationConstructor in concert with other applica-
tions that contain data or a simulation, and it provides a structured work space
where they can record the sense they are making of their findings.
STRUCTURING AND PROBLEMATIZING 285
FIGURE 1 The ExplanationConstructor used to articulate questions, explanations, and back-
ing support. The outline of students’ questions, subquestions, and explanations is shown in the
upper left Organizer panel; Explanation Guides specific to the explanatory framework selected
are shown in the upper right.
ExplanationConstructor embodies two kinds of structuring that embody scaf-
folding strategies for articulation and reflection: provide guidance for planning
and guidance for monitoring (Strategies 7a and 7b in Quintana et al., this issue).
The environment contains tools geared to help students articulate their research
questions and the links between candidate explanations and these questions. En-
couraging students to record pending questions and to be explicit about the rele-
vance of their findings for those questions is intended to provide needed structure,
encouraging students to record their overall plans and continuously monitor their
progress. These are elements of articulation and reflection that students frequently
omit in the rush toward producing final products.
Another aspect of the structuring is apparent in the explanation guides that
serve as prompts for critical constituents of an explanation. These reflect an addi-
tional scaffolding strategy of highlighting epistemic features of scientific products
(Strategy 7d in Quintana et al., this issue). These prompts are intended to help stu-
dents structure their interpretations in light of relevant disciplinary frameworks.
Sandoval and Reiser (in press) analyzed the use of ExplanationConstructor in
conjunction with the BGuILE investigation environment Galápagos Finches
(Reiser et al., 2001; Tabak et al., 1996) in a high school biology classroom. The
Galápagos Finches enables learners to investigate changes in populations of plants
and animals in an ecosystem and serves as a platform for learning principles of
ecosystems and natural selection. Students can analyze a population across time or
according to some dimension of interest such as comparing male to female or
young to adult. Students used the Galápagos Finches to analyze data to understand
what killed many of the finches in the a population of finches during a crisis period
and why some finches were able to survive. They used ExplanationConstructor to
help manage the investigation, record their questions, construct explanations, and
link in evidence for their claims.
Sandoval and Reiser (in press) described the use of several aspects of the system
as productive structure for planning and self-monitoring in the investigations. For
example, the list of explanation guides (“existing variation,” “change in environ-
ment,” etc.) and the students’ own generated subquestions both served as external
aides to help students track what issues their explanation needs to address. These
provided an anchor to which students returned to evaluate their in-progress expla-
nations. Students continued to refer to their list of research questions to remind
themselves which questions were accomplished and which were pending—for ex-
ample, “we are still answering that question [pointing to question in journal].”
They negotiated what tasks to work on next by revisiting their questions—“How
are we gonna answer that?; Where are we gonna go look?” They returned to ques-
tions and discovered “we don’t know that yet,” helping them determine where they
needed to focus next in the data. In general, the explanation facilities seemed to
provide a structure students could use to construct a representation of their plans,
and they used this representation to guide and monitor their own investigation.
Problematizing Aspects of Subject Matter
The second mechanism for scaffolding is to make some aspects of students’ work
more “problematic” (Hiebert et al., 1996) in a way that increases the utility of the
problem-solving experience for learning. That is, the software tools help students
see something as requiring attention and decision making that they might other-
wise overlook. Rather than simplifying the task, the software leads students to en-
counter and grapple with important ideas or processes. This may actually add diffi-
culty in the short term, but in a way that is productive for learning.
This notion of problematizing as supporting learners has its roots in the original
analysis of scaffolding presented by Wood et al. (1976). Wood et al. characterized
part of the functions of scaffolding as focusing attention by “marking critical fea-
tures” of a task, which may involve highlighting “discrepancies” between what a
child might produce and “correct production.” The scaffolding function of “main-
taining direction” mentioned by Wood et al. may also serve as problematizing, fo-
cusing learners on aspects of the task not yet performed. In a way, problematizing
is the flip side of structuring—although a structure may be helpful, it also may in-
troduce a cost in reworking one’s ideas in terms of the presented framework.
The importance of the mechanism of problematizing has been highlighted in re-
cent studies of teachers’ support of inquiry classrooms. In a study of a community
of learners classroom, Engle and Conant (2002) found that making subject matter
problematic and connecting students’ work to disciplinary frameworks are key as-
pects of a productive discussion-based classroom. Similarly, the teaching strategy
of inducing “cognitive conflict” involves drawing attention to problems in stu-
dents’ work in progress (Webb & Palincsar, 1996). In their analyses of expert tu-
tors, Lepper et al. (1993) pointed out that tutors modulate their support to target an
optimal level of difficulty. Tutors appear to seek a balance in eliciting the learner’s
active engagement with the problem and providing enough support to prevent frus-
tration and nonproductive floundering. They refrain from providing too much help
that would remove productive complexity or prevent errors that are learning oppor-
tunities (Lepper et al., 1993; Merrill et al., 1995). These studies suggest a role for
guiding learning not only by simplifying but also in encouraging learners to face
some of the complexity of the domain in productive ways.
Problematizing in a learning situation consists generally of several characteris-
tics. First, it involves focusing students’ attention on an aspect of a situation that
needs resolution. Second, it involves engaging students—eliciting students’ com-
mitment of attention and resources to reasoning about an aspect of a problem. This
may involve creating a sense of dissonance or curiosity. Finally, it may involve an
affective component—creating interest in some aspect of a problem or getting stu-
dents to care about understanding or resolving an issue.
How can software tools achieve some of these same ends? I discussed earlier
how tools can shape users’ conceptions of a task. A tool’s interface can shape the
STRUCTURING AND PROBLEMATIZING 287
concepts and language students use to identify actions. This may provide useful
structure, but it may also act to problematize important disciplinary constructs. For
example, in Belvedere, a software environment in which students create arguments
represented as a graph of claims and evidence, students must decide whether each
assertion is a claim or evidence and must indicate the relations between claims and
evidence in their representation (Cavalli-Sforza, Weiner, & Lesgold, 1994; Toth,
Suthers, & Lesgold, 2002). This can focus students on worrying about how a par-
ticular idea should be classified, something that simply expressing an argument in
the fuzziness of natural language may gloss over.
Despite some suggestions of the importance of engaging students with com-
plexity in early accounts of scaffolding with human teachers, accounts of scaffold-
ing in software have tended to emphasize the structuring aspect of support, making
tasks more tractable by reducing difficulty, reducing overwhelming options, or
embedding prompts or guides to help learners focus their efforts. There are some
discussions in these accounts of providing an optimal level of such support and
fading that support when appropriate (Collins et al., 1989; Soloway, Guzdial, &
Hay, 1994). However, the notion of problematizing goes beyond avoiding giving
too much help or fading the help as learners develop increasing expertise. This core
of this approach is to guide the learner into facing complexity in the domain that
will be productive for learning, for example, by connecting their work on a prob-
lem to disciplinary frameworks.
The general goal of increasing learners’ engagement with complex disciplinary
ideas is a key aspect of the last few decades of education reforms. Reform ideas
such as conceptual understanding versus rote learning, engaging in authentic prac-
tices, and learning by doing all create a need to focus learners on unfamiliar but
productive disciplinary ideas (American Association for the Advancement of Sci-
ence, 1990; Bruner, 1966; Cohen, 1988). Although software tools are often de-
signed with the goal of eliciting deeper engagement and reasoning, little attention
in theoretical definitions of scaffolding or design arguments has been focused on
the problematizing aspect. The mechanisms of structuring and problematizing are
an attempt to characterize how software tools can help learners grapple with more
ambitious learning goals.
This focus of resources inherent in problematizing can address the problems
raised earlier of nonreflective work and superficial analyses, encouraging and re-
quiring students to address critical questions and ideas in the discipline. In these
cases, requiring students to connect their thinking to disciplinary issues can pro-
voke deliberations, debate, and decisions that are productive for students as they
make sense of the findings of their investigation and manage its progress.
By leading students to encounter particular ways of thinking, scaffolding can
provoke students, “rocking the boat” when they are proceeding along without be-
ing mindful enough of the rich connections of their decisions to the domain con-
tent. It may challenge their self-assessments that they have really addressed as-
pects of the research question appropriately (Davis, 2003). This provocation may
occur as the tools force them into decisions or commitments required to use the vo-
cabulary and machinery of the interface. This type of scaffolded tool may create
short-term costs, preventing students from rushing through their work in a problem
without being mindful of the subject-matter issues that are the goal of the instruc-
tion. Although this may be a short-term challenge rather than directly assisting
with more expeditious solutions, such a tool may make the students’ efforts in the
problem a more productive learning opportunity.
The social context of collaborative problem solving is often integral to the
problematizing nature of the tool. Students must make their understanding public
when using tools that represent conceptual distinctions explicitly. Such tools re-
quire students to discuss disciplinary ideas to effect actions in the tool. In this way,
the artifact students use to examine data and the external records they create of
their interpretations become a vehicle for negotiation of understanding about the
disciplinary ideas and their application to the task at hand. The pressure to be ex-
plicit in a shared external representation can serve as a catalyst for negotiation of
ideas (Pea & Gomez, 1992; Roschelle, 1992; Teasley & Roschelle, 1993).
Influence of Problematizing Student Work
There are several ways that problematizing tools can help address learner
The tools can lead learners to be more explicit about their
reasoning by providing a restricted representation that makes important distinc-
tions explicit. This can help counter the tendency toward superficial and
nonreflective work. For example, Computer-Supported Intentional Learning Envi-
ronments (CSILE) requires students to indicate the rhetorical connection of their
comment to an ongoing discussion (Scardamalia & Bereiter, 1994). The explicit
representation in the artifact also provides concrete evidence of the work that can
enable later reflection and revisiting of the history of the investigation (Collins &
Brown, 1988; Loh et al., 2001). The explicit representation may also aid the group
dynamics. For example the inscription of the work so far can make explicit the con-
sensus that has been reached and thereby help the group avoid reinventing solu-
tions to decisions already constructed. Productive discussions in community of
learners classrooms require connecting the discussion to disciplinary norms
(Engle & Conant, 2002). Software tools may provide a context for more productive
discussions within a group if they can be designed to encourage these connections
and thereby focus discussion to deal with these critical aspects (de Vries et al.,
2002; Sandoval & Reiser, in press).
Software tools can employ explicit representations that are
more precise than natural language or students’ paper-and-pencil work. Requiring
STRUCTURING AND PROBLEMATIZING 289
students to select from limited options can encourage them to grapple with deci-
sions they might otherwise overlook such as classifying the way evidence connects
to positions in an argument (Bell & Linn, 2000; Toth et al., 2002). This may help
students face the challenges of unfamiliar strategies as they engage in and docu-
ment the decision making elicited in interactions with the tool.
Surface gaps and disagreements.
The explicitness required in con-
strained artifacts can lead students to discover and address disagreements that may
be beneath the surface in their individual interpretations of the group’s work. Ar-
ticulating interpretations in a joint product, particularly within a constrained repre-
sentation, helps make clear where there is disagreement and need for resolution
(Kyza, Golan, Reiser, & Edelson, 2002; Teasley & Roschelle, 1993). Barron
(2003) argued that collaborative work in investigations requires a convergence of
social and discourse practices with the cognitive practices of constructing and ex-
ploring a joint problem space. Productive classroom discussions require not just
participation but active engagement in particular learning processes including un-
covering divergent positions and resolving them through fair and moderated argu-
ment (Barron, 2003; Jimenenez-Aleixandre, Rodriguez, & Duschl, 2000; Phelps
& Damon, 1989; van Zee & Minstrell, 1997). Software tools have the potential to
aid teams of learners by surfacing disagreements critical to disciplinary goals.
Examples of Problematizing Student Work
To illustrate scaffolding through problematizing, I consider another scaffolding
strategy associated with the guideline of facilitating articulation and reflec-
tion—“highlight epistemic features of scientific practices and products” (Strategy
7d in Quintana et al., this issue). As students interact with the tools and create work
products, they need to use the important epistemic features of the discipline and
may need to reframe their own ways of thinking in terms of these features. For ex-
ample, as mentioned earlier, Belvedere requires students to clearly indicate the re-
lation (supporting or disconfirming) between evidence and claims (Toth et al.,
2002). Similarly, CSILE requires students to indicate the rhetorical relation be-
tween their contribution and other entries in an online discussion such as elabora-
tion, new question, disagreement, and so on (Scardamalia & Bereiter, 1994).
Transparency in cognitive artifacts that people use can affect the ease with
which they can achieve their desired ends through actions in the interface
(Hutchins et al., 1986; Norman, 1988, 1991). Here, the goal, rather than solely be-
ing only the ease of mapping between learners’ intentions and the representations
and language of the tool, is the fit of the interactions with the tool with the type of
thinking needed. Tools can provide structure and focus learners on important con-
stituents of tasks such as argument structure. However, the need to use the struc-
tures in the tool may uncover complexity in the domain. This can help learners
avoid solutions that are too superficial and lead them to focus resources on produc-
tive issues. The thinking required to make decisions in authoring an argument in
Belvedere or participating in a CSILE discussion goes beyond what may be re-
quired to produce more traditional artifacts such as verbal essays in which the re-
search behind the essay and the deeper argument structure embedded in the essay
may be very difficult to discern. Interacting with a more general tool like a word
processor may not create the same pressures to face the difficult decisions entailed
in these scaffolded systems.
The following example, drawn from Sandoval and Reiser (in press), demonstrates the
problematizing nature of the software tools. In this example, three high school biology
students are working in a group using Galápagos Finches and ExplanationConstructor to
investigate changes in the finches population. In this excerpt, the students are deliberat-
ing which explanatory framework of explanation guides to select. One student’s sugges-
tion to use the natural selection framework provokes debate on one of the key ideas in the
domain, the nature of traits. In considering whether this framework fits their interpreta-
tion of the problem, the group disagrees about whether food choice qualifies as a trait. In
the course of this debate, the group brings in key ideas about physical traits and the rela-
tion between structure and function.
Evan: (Reading prompt on framework) “Environment causes …”
Evan: Yeah, “to be selected for …”
Janie: Yeah, but that means like …
Evan: // what food they eat //
Janie: … organism with these trait
Evan: // the trait being the food
Franny: Yeah, that’s right.
Janie: No, because like, if my trait is to eat steak, and there’s no steak, I’m im-
mediately gonna go to something else.
Evan: If you’re only a vegetarian and you only eat … you don’t eat meat,
you’re not gonna eat meat. Well, that depends …//
Janie: Are you insane!?
Franny: OK, OK. Don’t think of people. Think of these guys (the finches). If
they only eat one type of seed with their beaks and that seed is gone then
they can’t live anymore.
This example demonstrates aspects of discourse that the tool is designed to cat-
alyze and support and suggests a way that tools can help problematize subject mat-
ter. Having to structure the analysis of their findings in terms of the theoretical
framework embedded in the tool required students to frame their understanding in
terms of principles of the domain. Rather than just writing down their explanation,
the tool forces them to consider how to express their hypothesis and its support
STRUCTURING AND PROBLEMATIZING 291
within a disciplinary framework such as natural selection. Decisions about the use
of the artifact became the context for negotiation between the students of these im-
portant disciplinary ideas such as the nature of a trait and the difference between a
specie’s characteristic traits and learned behaviors. The tool was a context for use-
ful conversations that helped students overcome the limitations of unfamiliar strat-
egies and helped them avoid superficial solutions not connected to disciplinary
ideas. In this case, both sense making (interpreting the observed differences in in-
dividuals as candidates for traits supporting differential survival) and articulation
were supported by the problematizing nature of the tool.
It is important to note that the tool was a support for these practices, but its ef-
fective use relies on other factors such as the inclination of the students to engage
sufficiently in handling the complexity uncovered in their interactions with the tool
as well as expectations and classroom norms fostered by the teacher (Tabak, this
issue). I return to the role software scaffolds can play in the larger classroom sys-
tem in the Discussion section.
Another example of problematizing reflects the scaffolding strategy “make dis-
ciplinary strategies explicit in the artifacts learners create,” associated with the
guideline to “organize tools and artifacts around the semantics of the discipline”
(Strategy 2b and “Guideline 2” in Quintana et al., this issue, respectively). The
BGuILE environment Animal Landlord is designed to provide tools to analyze an-
imal behavior that help students understand the strategies and constructs in the dis-
cipline (Smith & Reiser, 1998). In the Animal Landlord, students study examples
of animal behavior to isolate and analyze the key components of complex animal
behavior. Students analyze digital video clip examples of behavior such as hunting
or eating and deconstruct the complex sequence into what they see as the important
causal events. Students identify significant events by selecting frames from a clip,
categorizing them from a behavioral taxonomy, and annotating them with their in-
terpretations (see Figure 2). In this way, students build an annotated storyboard
that describes the progression of critical events in a behavior. Students use the tool
to compare and contrast examples in the corpus such as different patterns in lion
hunting or how different animals obtain food (Golan, Kyza, Reiser, & Edelson,
2001; Smith & Reiser, 1998).
The type of artifact students create is designed to provide useful structure for the
students’investigation and to focus attention and make problematic disciplinary dis-
tinctions. The structure of the artifact, a sequence of labeled and annotated frames,
clearly represents the structure of the process they use—decomposing complex be-
havior into its critical constituents, classifying and interpreting each constituent. In
this way, the inscriptions students create to record their analyses are a relatively
transparent representation of the processes students need to use—decomposition
and categorization. Furthermore, the analyses are organized in these inscriptions
into two distinct categories—observations refer to what one can see directly in the
data, and interpretations represent what one infers from these observations. This is a
key epistemic distinction in the scientific practice students are learning; thus, this as-
pect of the tool also reflects the scaffolding strategy “highlight epistemic features”
mentioned earlier in the ExplanationConstructor example.
Interacting with the tool to create these artifacts can also help problematize key
aspects of the practice. Requiring students to categorize their observations in terms
of disciplinary frameworks pushes students to articulate their understanding and
represent it in the artifact. The explicit distinction between observation and inter-
pretation is intended to elicit and support discussions geared at understanding the
relation between the two.
The following episode illustrates how the structure of the artifacts can lead stu-
dents to grapple with disciplinary content (Golan et al., 2001). This debate was re-
corded from a group of three seventh-grade students who were watching a clip fea-
turing two Golden Lion Tamarin monkeys eating a grape. One of the monkeys (the
male) had the grape first. The female took part of the grape away from him, and then
STRUCTURING AND PROBLEMATIZING 293
FIGURE 2 Artifacts constructed in the Animal Landlord. Students decompose complex be-
havior into its constituents, categorize each constituent, and record their observations and inter-
jumped away to another branch. Students had selected frames from the clip to anno-
tate that included the point at which the female jumped. The students were arguing
about this event, and disagreed on their interpretation. Two of the students believed
that the event was an instance of “mount” behavior, whereas the third student did not
agree. (The term mount was one of 15 built-in labels available;others included sniff,
rest, move, take food, follow, etc.) One student (Chandan, who was currently con-
trolling the keyboard), had labeled the clip as “mount” and was bringing up what to
type into the observation annotations associated with that event. In essence, this ar-
gument was how to define the behavior mount—as merely contact between two
animals or as a specialized kind of contact.
Chandan: What did we observe as mount? (reading the prompt from the
Danny: No, that one is yours because I totally disagree with you guys!
Chandan: Good for you! Come on man, you see in the clip it just looks mounting,
they got on top of each other.
Danny: No, he jumped over her.
Dennis: She, she jumped over him …
Danny: Whatever, she jumped over him.
Chandan: I know, but still, the contact …
Danny: She jumped over him, doesn’t matter, contact is not what you guys are
talking about. Shoot, you are talking about like getting on top of each
other and staying on top for a couple of minutes.
Dennis: No, no, no. We are not talking about that!
Chandan: No, no, not that, no, no, that’s not. See, look, watch this contact.
(replaying the clip beginning at the selected frame) Boom! Look at
Danny: Jumping over!Over!
This argument was finally settled by one of the researchers clarifying what the
behavior mount means in the domain of animal behavior. This discussion surfaced
discrepancies in students’ implicit definitions of the behavior. Converging on clear
and explicit definitions is an important step in developing and applying a categori-
zation scheme. Making these decisions as part of their analysis and clearly repre-
senting these distinctions in the artifacts they create provoked these and similar ar-
guments about the meaning or indicators of behavior, surfacing disagreements and
eventually refining students’ definitions of these behaviors.
It is interesting that this disagreement had been brewing for some time in the
group, but it had not yet been addressed. Here we see the potential for software to
problematize aspects of the subject matter. Had the group merely been asked to re-
port a summary of their observations or turn in a simple text report, it is possible
this discussion would not have occurred. Requiring students to articulate their un-
derstanding and represent it explicitly using menu item labels and the observation
and interpretation structure finally surfaced the differences in interpretation and
provoked these productive discussions. For several reasons, students may be in-
clined to downplay the type of confusion or disagreement that finally sur-
faced—for example, a tendency to accept solutions that appear on the surface suf-
ficient and the inclination to downplay the significance of difference of opinion to
avoid dissension. Yet when required to commit their analysis in the precise repre-
sentation (identifying each important event as an example of a small number of be-
havioral categories) as required by the software tool, this disagreement came to the
surface and became the focus of conversation in the group.
I have argued that software scaffolding can help learners by providing needed
structure and by problematizing important subject matter. I presented brief exam-
ples of learners’ encounters with software environments to help illustrate these
mechanisms. The aim of scaffolding design is to find ways to use the nature of
learners’ interactions with the tool to help shape their thinking in productive ways.
I discussed how the explicitness of representations and ability to represent impor-
tant conceptual aspects of a discipline in tools can play a role in structuring and
problematizing learners’ engagement with the subject matter.
In this section, I consider a number of general issues about the scaffolding
mechanisms of structuring and problematizing. First, I consider how the two
mechanisms of structuring and problematizing are related to one another. I discuss
design tensions that arise in trying to design learning environments that provide
structure and lead students to problematize subject matter. Finally, I consider other
aspects of the context that are important for scaffolding in software tools to influ-
ence learning practices.
Complementarity and Tensions
The two mechanisms of structuring and problematizing are often complemen-
tary. I described earlier how representations in tools can be designed to provide
needed structure that students can use to help guide and organize their work. The
structure can help students decompose tasks, monitor the performance of com-
ponent tasks, and encode the results of their work in useful forms. I suggested
that the features of a software tool that provide useful structure may also elicit
attention to critical issues in subject matter, for example, calling attention to im-
portant distinctions such as observation and interpretation. Thus, as shown in the
examples with ExplanationConstructor and Animal Landlord, the same charac-
teristic of a tool designed to help structure students’ engagement in a task, for
example, by providing guiding prompts or making explicit a set of subtasks, may
STRUCTURING AND PROBLEMATIZING 295
also problematize the disciplinary ideas by requiring students to make sense of
the options and connect their own work to these disciplinary ideas.
Yet the goals of supporting these two mechanisms may also be in tension—the
goals of simplifying the problem space may be somewhat in tension with encour-
aging learners to grapple with distinctions or disciplinary frameworks that are
likely to require more deliberation. For structure to be useful, learners must be able
to recognize and use the distinctions presented. Calling students’ attention to and
requiring use of unfamiliar strategies may work against the system’s usefulness for
guiding students’ investigations. It may require additional reasoning steps that
work counter to the structures intended to be useful. Or, if the strategies are unfa-
miliar enough and students cannot make the connections to their owns ways of
thinking, they may use the systems’structuring improperly or superficially. For ex-
ample, despite careful crafting of prompts to guide students’ work, students may
treat the software environment as “just another worksheet” and ignore the fine dis-
tinctions in the system’s attempt to structure the reporting of their work, or may
enter minimal answers rather than carefully considering what is needed.
The proper balance is tricky to achieve. Vygotsky’s (1978) notion of zone of
proximal development is important in characterizing the sense of balance needed
in scaffolded software tools. The technical tools of a scientist make extensive use
of disciplinary ideas, yet these tools are often inappropriate for learners. They are
not supportive enough to engage learners in grappling with the complex ideas in
the domain that they represent. Introducing overwhelming complexity without
needed structure does not lead to problematizing subject matter. On the other hand,
a tool that strips away the complexity, guiding students in lockstep, may not en-
gage them in grappling with the complexity and reasoning through the solutions
they are led to construct. Thus, achieving both functions requires a careful balance.
Design Tensions in Crafting Scaffolded Tools
Considering the tensions between attempting to problematize and provide struc-
ture suggests several important design tensions in designing scaffolded tools.
Support intuitive strategies versus problematize disciplinary
Designs have to strive for an optimal balance between connecting with
students’ intuitive strategies on one hand and requiring students to work within dis-
ciplinary frameworks on the other. Attempts to provide structure may focus atten-
tion and highlight critical features, but the problematizing is only effective if the
students can make the connections in bridging from their own intuitive strategies to
the structures enforced by the tool. Quintana et al. (this issue) review a number of
scaffolding strategies that take on this challenge—using authentic complex repre-
sentations, but tailored to better connect with learners’ intuitive understandings.
For example, Model-It™enables learners to construct dynamic systems models but
provides an intuitively compelling representation of a concept map and enables
learners to simply link two factors with qualitative types of influence rather than
getting bogged down in cumbersome mathematical relations between two vari-
ables (Jackson et al., 1998).
Generality versus specificity.
General guidance may be very helpful for
structuring work. For example, a general structure for posing questions, designing
investigations, and interpreting data may provide useful structures to enable learn-
ers to organize their investigation. It may have the benefit of introducing learners to
a small set of very powerful disciplinary ideas such as controlling variables or
graphically representing data. These ideas would potentially be applicable to
broad range of new problems. Yet to problematize subject matter, more specific
guidance may be needed such as the disciplinary frameworks embedded in
ExplanationConstructor. Students may have difficulty acting on such general guid-
ance, for example, knowing counterevidence is important in an argument but hav-
ing difficulty figuring out what types of evidence to collect or evaluate (Sandoval,
2003; Sandoval & Reiser, in press). Designers have to find the right balance for the
particular target learners in the tension between broadly applicable guidance and
guidance more tailored to subject matter.
Student responsibility and control versus more constrained
Ultimately the goal is for students to be able to direct their own inves-
tigations and to be involved in defining a problem, planning a solution, and con-
ducting empirical investigations. In scaffolded inquiry approaches, the learner
may be engaged directly in only some subset or in narrowed versions of these com-
ponents. For example, often in project-based science, the general problem is pro-
vided to students, although great care is taken to contextualize that problem in stu-
dents’ experiences and to create a sense of ownership as students explore their own
solutions to that problem. However, the risk of this narrowing to provide helpful
structure is that it may lead to learners just “going through the motions” rather than
being reflective about what is being required and why. In contrast, problematizing
by requiring important steps or distinctions may take some of the control for their
own investigations away from students.
Studies of teaching approaches exhibited by tutors and classroom teachers have
revealed a careful balance between providing help of various sorts to assist prob-
lem solving and help avoid difficulties on one hand while not providing too much
support that curtails learners’ active engagement or circumvents learners’ involve-
ment in the target reasoning practices (Hogan & Pressley, 1997; Lepper et al.,
1993; Merrill et al., 1995). Similarly, design of effective software scaffolding must
negotiate such a balance between these tensions. For example, tools may attempt
to link more intuitive representations with scientific language, provide prompts
that continually act to operationalize important strategies, or help students docu-
STRUCTURING AND PROBLEMATIZING 297
ment the connections between strategic decisions and the results in the investiga-
tion (Quintana et al., this issue), thereby attempting to bridge learners to expert
practice using representations that are elaborated and connected to learners’ start-
An important limitation in finding an optimal balance between these tensions
between structuring and problematizing is the limited ability of most scaffolded
tools to individualize their support. The notion of tailoring support to the needs of
particular learners and learning situations has been an important aspect of scaffold-
ing in its application to teaching and tutoring (Webb & Palincsar, 1996; Wood et
al., 1976). Whereas reasoning about the states of individual learners has been ex-
plored in intelligent tutoring systems for more procedural domains (Anderson,
Corbett, Koedinger, & Pelletier, 1995), the approach in the scaffolded cognitive
tools has been to embed support within the system as prompts or as representations
in the structure of the tool itself. Such tools are adaptable, to a limited extent, under
the control of the learners who can explore additional prompts or assistance avail-
able, attempt to follow or work around the system’s advice, or perhaps under con-
trol of teachers who may tailor messages or functionality provided. However, the
sense in which tools can scaffold learners under such conditions are clearly quite
different than expert teachers who can tailor their advice to an assessment of the in-
dividual learner state. Of course, there are other pressures on classroom teachers
who must deal with the needs of many learners simultaneously, so their ability to
scaffold dozens of learners simultaneously faces other challenges.
Scaffolding Requires a System
A final caution to be discussed in exploring models of scaffolding in software tools
is that learners, tools, and teachers work together as a system, and it is an oversim-
plification to consider how tools can scaffold learners without considering the
other aspects of this system. Learners come with attitudes and expectations toward
the subject matter and toward learning, and these expectations are shaped in part
by the classroom culture created by teachers. Tools cannot force learners to reason
about an idea or to use kinds of language. Rather, tools can provide support that in
the right context may influence the directions and practices of learners and teach-
ers. Thus, scaffolded tools can create opportunities, but whether learners capitalize
on these opportunities depends on the expectations and practices established in the
classroom. The nature of the scaffolding is that it may act to provoke or catalyze,
but of course the software tools cannot require that learners mindfully engage with
A first critical factor is that the ways of thinking the tool is designed to sup-
port must be threaded through all aspects of the classroom system—in the cur-
ricular activities that surround the tool, in teachers’ support working with indi-
vidual groups, and in the teachers’ structuring and guidance of whole class
discussions. Although software may provoke attention to an idea, such as the
distinction between observation and interpretation, the classrooms practices need
to take up this idea in classroom discourse for these ideas to take on real mean-
ing (Hogan et al., 1999; Hogan & Pressley, 1997; Lemke, 1990; Tabak & Reiser,
1997). Tabak (2002, this issue; Tabak & Reiser, 1997) has characterized this as a
synergistic process in which teachers capitalize on and reinforce helpful struc-
ture in tools, and the influence of tools relies on how teachers cultivate their use.
Software tools may influence the focus of attention, and teachers can then capi-
talize on and reinforce in their questioning and guidance of students (Tabak,
2002). For these distinctions in the tool’s representations to be taken seriously
and treated mindfully by learners, teachers need to structure their interactions
with students around the same framework. The software tool provides a way to
reinforce a way of talking and thinking about the data, but it is only part of the
classroom system in which teachers frame the way learners think about material.
Teachers and software tools may work in concert, with the software tools provid-
ing a concrete representation of distinctions that teachers have brought into
classroom discussions (Kemp, Tzou, Reiser, & Spillane, 2002; Reiser et al.,
2001; Tabak & Reiser, 1997; Tzou, Reiser, Spillane, & Kemp, 2002). Thus, soft-
ware scaffolds can influence learning by supporting more productive prob-
lem-solving conversations among learners and among learners and teachers
(Teasley & Roschelle, 1993).
Another important role of the classroom system is the availability of re-
sources to support the investigation. Problematizing is a process of focusing at-
tention along productive dimensions, but naturally, it does not guarantee that this
focus of attention will lead to productive results. This is particularly critical
when considering the problematizing nature of software tools. Provoking pro-
ductive conflict within a group requires that the group have access to resources
needed to resolve the conflict such as other information resources or teacher
In summary, I have argued that scaffolding occurs through two mechanisms,
structuring and problematizing. Most current accounts of scaffolding define sup-
port that helps students proceed through tasks by providing structure. However,
given the importance of connecting students’problem-solving work to disciplinary
content, skills, and strategies, it is important to provoke issues in students, veering
them off the course of nonreflective work and forcing them to confront key disci-
plinary ideas in their work. Tools shape users’ engagement with tasks. As such,
tools can be designed to provide a context that can influence users’perceptions, the
discourse between learners and between learners and teachers, and the ways they
represent their thinking in artifacts of their work. The artifacts students use and
create can be designed to map onto important disciplinary ideas and strategies,
thereby problematizing these ideas as students use the tool to work through the
task and represent the products of their work.
STRUCTURING AND PROBLEMATIZING 299
An earlier version of this article was presented at the Computer Support for Col-
laborative Learning 2002 Conference (January 2002, Boulder, CO). This research
was funded by the National Science Foundation under Grants REC–9980055 to
the KDI/ASSESS project and REC–9720383 to the Center for Learning Technol-
ogies in Urban Schools and by Grant 97–57 from the James S. McDonnell Founda-
tion, Cognitive Studies for Educational Practice. The opinions expressed herein
are those of the author and not necessarily those of these foundations. For addi-
tional information about BGuILE software and curricula, see
http://www.letus.org/bguile/ and http://www.hi-ce.org/iqwst/
This article builds on the design research of the BGuILE group at Northwestern
University, and I am grateful for conversations with my colleagues Iris Tabak,
Brian Smith, William Sandoval, Ravit Golan Duncan, and Elena Kyza about prin-
ciples of scaffolding in learning environments. I am also grateful to my colleagues
in the KDI/ASSESS group for discussion of these ideas—Joseph Krajcik, Danny
Edelson, Elliot Soloway, Chris Quintana, and Elizabeth A. Davis—and to Bruce
Sherin for many fruitful discussions. This article was greatly improved by com-
ments from reviewers Mimi Recker and Phil Bell; and editors Elizabeth A. Davis
and Naomi Miyake.
American Association for the Advancement of Science. (1990). Science for all Americans: Project
2061. New York: Oxford University Press.
Anderson, J. R., Corbett, A. T., Koedinger, K. R., & Pelletier, R. (1995). Cognitive tutors: Lessons
learned. The Journal of the Learning Sciences, 4, 167–207.
Bannon, L. J., & Bødker, S. (1987). Beyond the interface: Encountering artifacts in use. In J. M. Carroll
(Ed.), Interfacing thought: Cognitive aspects of human–computer interaction (pp. 227–253). Cam-
bridge, MA: MIT Press.
Barron, B. (2003). When smart groups fail. The Journal of the Learning Sciences, 12, 307–359.
Bell, P., & Linn, M. C. (2000). Scientific arguments as learning artifacts: Designing for learning from
the web with KIE. International Journal of Science Education, 22, 797–817.
Blumenfeld, P. C., Soloway, E., Marx, R. W., Krajcik, J. S., Guzdial, M., & Palincsar, A. (1991). Moti-
vating project-based learning: Sustaining the doing, supporting the learning. Educational Psycholo-
gist, 26, 369–398.
Brown, A. L., & Campione, J. C. (1994). Guided discovery in a community of learners. In K. McGilly
(Ed.), Classroom lessons: Integrating cognitive theory and classroom practice (pp. 229–270). Cam-
bridge, MA: MIT Press.
Bruner, J. S. (1966). Toward a theory of instruction. Cambridge, MA: Belknap Press/Harvard Univer-
Cavalli-Sforza, V., Weiner, A., & Lesgold, A. M. (1994). Software support for students engaging in sci-
entific activity and scientific controversy. Science Education, 78, 577–599.
Chi, M. T. H., Bassok, M., Lewis, M. W.,Reimann, P., & Glaser, R. (1989). Self-explanations: How stu-
dents study and use examples in learning to solve problems. Cognitive Science, 13, 145–182.
Chi, M. T. H., Feltovich,P., & Glaser, R. (1981). Categorization and representation of physics problems
by experts and novices. Cognitive Science, 5, 121–152.
Cohen, D. K. (1988). Teaching practice: Plus que ca change. In P. W. Jackson (Ed.), Contributing to ed-
ucational change: Perspectives on research and practice (pp. 27–84). Berkeley: McCutchan.
Coleman, E. B. (1998). Using explanatory knowledge during collaborative problem solving in science.
Journal of the Learning Sciences, 7, 387–427.
Collins, A. (1991). Cognitive apprenticeship and instructional technology. In L. Idol & B. F. Jones
(Eds.), Educational values and cognitive instruction: Implications for reform (pp. 121–138).
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Collins, A., & Brown, J. S. (1988). The computer as a tool for learning through reflection. In H. Mandl
& A. Lesgold (Eds.), Learning issues for intelligent tutoring systems (pp. 1–18). New York:
Collins, A., Brown , J. S., & Newman, S. E . (1989). Cogn itive apprenticeship: Teaching the crafts of
reading, writin g, and mathematics. In L. B. Resni ck (Ed.), Knowing, le arning, and instruc tion:
Essays in honor of Ro bert Glaser (pp. 453–494). Hillsdale, NJ: Lawrence Erlbaum Associates,
Davis, E. A. (2003). Prompting middle school science students for productive reflection: Generic and
directed prompts. The Journal of the Learning Sciences, 12, 91–142.
Davis, E. A., & Linn, M. C. (2000). Scaffolding students’knowledge integration: Prompts for reflection
in KIE. International Journal of Science Education, 22, 819–837.
de Vries, E., Lund, K., & Baker, M. (2002). Computer-mediated epistemic dialogue: Explanation and
argumentation as vehicles for understanding scientific notions. The Journal of the Learning Sci-
ences, 11, 63–103.
Driver, R., Newton, P., & Osborne, J. (2000). Establishing the norms of scientific argumentation in
classrooms. Science Education, 84, 287–312.
Edelson, D. C. (2001). Learning-for-use: A framework for integrating content and process learning in
the design of inquiry activities. Journal of Research in Science Teaching, 38, 355–385.
Edelson, D. C., Gor din, D. N., & Pea, R. D. (1999). Addre ssing the challenges of inquir y-based
learning throug h technology and curriculum de sign. The Journal of the Learning Sciences, 8,
Engle, R. A., & Conant, F. R. (2002). Guiding principles for fostering productive disciplinary engage-
ment: Explaining an emergent argument in a community of learners classroom. Cognition and In-
struction, 20, 399–483.
Golan, R., Kyza, E. A., Reiser, B. J., & Edelson, D. C. (2001, March). Structuring the task of behavioral
analysis with software scaffolds. Paper presented at the annual meeting of the National Association
of Research in Science Teaching, St. Louis, MO.
Guzdial, M. (1994). Software-realized scaffolding to facilitate programming for science learning. In-
teractive Learning Environments, 4, 1–44.
Hiebert, J., Carpenter, T. P., Fennema, E., Fuson, K., Human, P., Murray, H., et al. (1996). Problem solv-
ing as a basis for reform in curriculum and instruction: The case of mathematics. Educational Re-
searcher, 25(4), 12–21.
Hmelo, C. E., Holton, D. L., & Kolodner, J. L. (2000). Designing to learn about complex systems. The
Journal of the Learning Sciences, 9, 247–298.
Hogan, K., Nastasi, B. K., & Pressley, M. (1999). Discourse patterns and collaborative scientific rea-
soning in peer and teacher-guided discussions. Cognition and Instruction, 17, 379–432.
Hogan, K., & Pressley, M. (1997). Scaffolding scientific competencies within classroom communities
of inquiry. In K. Hogan & M. Pressley(Eds.), Scaffolding student learning: Instructional approaches
and issues (pp. 74–107). Cambridge, MA: Brookline.
STRUCTURING AND PROBLEMATIZING 301
Hollan, J. D., Bederson, B. B., & Helfman, J. (1997). Information visualization. In M. G. Helander, T.
K. Landauer, & V. Prabhu (Eds.), Handbook of human–computer interaction (pp. 33–48). Essex,
England: Elsevier Science.
Hollan, J. D., Hutchins, E., & Kirsh, D. (2000). Distributed cognition: Toward a new foundation for hu-
man–computer interaction research. ACM Transactions on Computer–Human Interaction, 7,
Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press.
Hutchins, E., Hollan, J. D., & Norman, D. A. (1986). Direct manipulation interfaces. In D. A. Norman
& S. Draper (Eds.), User centered system design: New perspectives in human–computer interaction.
Hillsdale, NJ: Lawrence Erlbaum Association, Inc.
Jackson, S. L., Krajcik, J., & Soloway, E. (1998). The design of guided learning-adaptable scaffolding
in interactive learning environments. In C.-M. Karat, A. Lund, J. Coutaz, & J. Karat (Eds.), Proceed-
ings of CHI 98: Human factors in computing systems (pp. 187–194). Reading, MA: Addison-Wesley.
Jimenenez-Aleixandre, M. P., Rodriguez, A. B., & Duschl, R. A. (2000). “Doing the lesson” or “doing
science”: Argument in high school genetics. Science Education, 84, 757 –792.
Kemp, E. K., Tzou, C. T., Reiser, B. J., & Spillane, J. P. (2002). Managing dilemmas in inquiry science
teaching. In P. Bell, R. Stevens, & T. Satwicz (Eds.), Keeping learning complex: The Proceedings of
the Fifth International Conference of the Learning Sciences (ICLS) (pp. 206–213). Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Klahr, D. (2000). Exploring science: The cognition and development of discovery processes. Cam-
bridge, MA: MIT Press.
Klahr, D., & Dunbar, K. (1988). Dual space search during scientific reasoning. Cognitive Science, 12,
Klahr, D., Fay, A. L., & Dunbar, K. (1993). Heuristics for scientific experimentation: A developmental
study. Cognitive Psychology, 25, 111–146.
Koedinger, K. R., & Anderson, J. R. (1993). Reifying implicit planning in geometry: Guidelines for
model-based intelligent tutoring system design. In S. P. Lajoie & S. J. Derry (Eds.), Computers as
cognitive tools (pp. 15–45). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Kolodner, J. L., Owensby, J. N., & Guzdial, M. (2004). Case-based learning aids. In D. H. Jonassen
(Ed.), Handbook of research for educational communications and technology (2nd ed., pp.
829–862). Mahwah, NJ: Lawrence Erlbaum Associates Inc.
Krajcik, J., Blumenfeld, P. C., Marx, R. W., Bass, K. M., Fredricks, J., & Soloway, E. (1998). Inquiry in
project-based science classrooms: Initial attempts by middle school students. Journal of the
Learning Sciences, 7, 313–350.
Kuhn, D. (1993). Science as argument: Implications for teaching and learning scientific thinking. Sci-
ence Education, 77, 319–337.
Kuhn, D., Amsel, E., & O’Loughlin, M. (1988). The development of scientific thinking skills. San
Diego, CA: Academic.
Kurth, L. A., Anderson, C. W., & Palincsar, A. S. (2002). The case of Carla: Dilemmas of helping all
students to understand science. Science Education, 86, 287–313.
Kyza, E. A., Golan, R., Reiser, B. J., & Edelson, D. C. (2002). Reflective inquiry: Enabling group
self-regulation in inquiry-based science using the progress portfolio tool. In G. Stahl (Ed.), Com-
puter support for collaborative Learning: Proceedings of CSCL 2002 (pp. 227–236). Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Lemke, J. L. (1990). Talking science: Language, learning, and values. Norwood, NJ: Ablex.
Lepper, M. R., Woolverton, M., Mumme, D. L., & Gurtner, J. (1993). Motivational techniques of expert
human tutors: Lessons for the design of computer-based tutors. In S. P. Lajoie & S. J. Derry (Eds.),
Computers as cognitive tools (pp. 75–105). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Linn, M. C. (2000). Designing the knowledge integration environment. International Journal of Sci-
ence Education, 22, 781–796.
Linn, M. C., Bell, B., & Davis, E. A. (2004). Internet environments for science education. Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Loh, B. (2003). Using articulation and inscription as catalysts for reflection: Design principles for re-
flective inquiry. Unpublished doctoral dissertation, Northwestern University, Evanston, IL.
Loh, B., Reiser, B. J., Radinsky, J., Edelson, D. C., Gomez, L. M., & Marshall, S. (2001). Developing
reflective inquiry practices: A case study of software, the teacher, and students. In K. Crowley, C. D.
Schunn, & T. Okada (Eds.), Designing for science: Implications from everyday, classroom, and pro-
fessional settings (pp. 279–323). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Merrill, D. C., Reiser, B. J., Merrill, S. K., & Landes, S. (1995). Tutoring: Guided learning by doing.
Cognition and Instruction, 13, 315–372.
Moje, E. B., Collazo, T., Carrillo, R., & Marx, R. W. (2001). “Maestro, what is ‘quality’?”: Language, liter-
acy, and discourse in project-based science. Journal of Research in Science Teaching, 38, 469–498.
Norman, D. A. (1988). Psychology of everyday things. New York: Basic Books.
Norman, D. A. (1991). Cognitive artifacts. In J. M. Carroll (Ed.), Designing interaction: Psychology at
the human–computer interface (pp. 17–38). New York: Cambridge University Press.
Norman, D. A. (1993). Things that make us smart: Defending human attributes in the age of the ma-
chine. Reading, MA: Addison-Wesley.
Olson, S., & Loucks-Horsley, S. (Eds.). (2000). Inquiry and the national science education standards:
A guide for teaching and learning. Washington, DC: National Academy Press.
Passmore, C., & Stewart, J. (2002). A modeling approach to teaching evolutionary biology in high
schools. Journal of Research in Science Teaching, 39, 185–204.
Pea, R. D. (1992). Practices of distributed intelligence and designs for education. In G. Salomon (Ed.), Distributed
cognitions: Psychological and educational considerations (pp. 47–87). New York: CambridgeUniversity Press.
Pea, R. D., & Gomez, L. M. (1992). Distributed multimedia learning environments: Why and how? In-
teractive Learning Environments, 2, 73–109.
Perkins, D. (1998). What is understanding? In M. S. Wiske (Ed.), Teaching for understanding: Linking
research with practice (pp. 39–58). San Francisco: Jossey-Bass.
Phelps, E., & Damon, W. (1989). Problem solving with equals: Peer collaboration as a context for
learning mathematics and spatial concepts. Journal of Educational Psychology, 81, 639–646.
Quintana, C., Eng, J., Carra, A., Wu, H.-K., & Soloway, E. (1999). Symphony: A case study in extend-
ing learner-centered design through process space analysis. In M. G. Williams, M. W. Altom, K.
Ehrlich, & W. Newman (Eds.), Proceedings of CHI 99 Conference on Human Factors in Computing
Systems (pp. 473–480). Reading, MA: Addison-Wesley.
Reif, F., & Larkin, J. H. (1991). Cognition in scientific and everyday domains: Comparison and learn-
ing implications. Journal of Research in Science Teaching, 28, 733–760.
Reiser, B. J., Tabak, I., Sandoval, W. A., Smith, B. K., Steinmuller, F., & Leone, A. J. (2001). BGuILE:
Strategic and conceptual scaffolds for scientific inquiry in biology classrooms. In S. M. Carver & D.
Klahr (Eds.), Cognition and instruction: Twenty-five years of progress (pp. 263–305). Mahwah, NJ:
Lawrence Erlbaum Associates, Inc.
Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. New York: Ox-
ford University Press.
Roschelle, J. (1992). Learning by collaboration: Convergent conceptual change. The Journal of the
Learning Sciences, 2, 235–276.
Rosebery, A. S., Warren, B., & Conant, F. R. (1992). Appropriating scientific discourse: Findings from
language minority classrooms. Journal of the Learning Sciences, 2, 61–94.
Salomon, G., Perkins, D. N., & Globerson, T. (1991). Partners in cognition: Extending human intelli-
gence with intelligent technologies. Educational Researcher, 20(3), 2–9.
Sandoval, W. A. (1998). Inquire to explain: Structuring inquiry around explanation construction in a
technology-supported biology curriculum. Unpublished doctoral dissertation, Northwestern Univer-
sity, Evanston, IL.
STRUCTURING AND PROBLEMATIZING 303
Sandoval, W. A. (2003). Students’ understanding of causal explanation and natural selection in a tech-
nology-supported inquiry curriculum. Journal of the Learning Sciences, 12, 5–51.
Sandoval, W. A., & Reiser, B. J. (2004). Explanation-driven inquiry: Integrating conceptual and
epistemic scaffolds for scientific inquiry. Science Education, 88, 345–372.
Scardamalia, M., & Bereiter, C. (1994). Computer support for knowledge-building communities. The
Journal of the Learning Sciences, 3, 265–283.
Schauble, L., Glaser, R., Duschl, R. A., Schulze, S., & John, J. (1995). Students’ understanding of the
objectives and procedures of experimentation in the science classroom. The Journal of the Learning
Sciences, 4, 131–166.
Schauble, L., Glaser, R., Raghavan, K., & Reiner, M. (1991). Causal models and experimentation strat-
egies in scientific reasoning. The Journal of the Learning Sciences, 1, 201–238.
Sherin, B. L. (2001). How students understand physics equations. Cognition and Instruction, 19,
Smith, B. K., & Reiser, B. J. (1998). National Geographic unplugged: Designing interactive nature
films for classrooms. In C.-M. Karat, A. Lund, J. Coutaz, & J. Karat (Eds.), Proceedings of CHI 98
(pp. 424–431). New York: ACM Press.
Soloway, E., Guzdial, M., & Hay, K. E. (1994). Learner-centered design: The challenge for HCI in the
21st century. Interactions, 1(2), 36–48.
Tabak, I. (2002). Teacher as monitor, mentor or partner: Uncovering participant structures involved in
supporting student-directed inquiry. In P. Bell, R. Stevens, & T. Satwicz (Eds.), Keeping learning
complex: The Proceedings of the fifth international conference of the learning sciences (ICLS) (pp.
466–472). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Tabak, I., & Reiser, B. J. (1997). Complementary roles of software-based scaffolding and teacher–stu-
dent interactions in inquiry learning. In R. Hall, N. Miyake, & N. Enyedy (Eds.), Proceedings of
Computer Support for Collaborative Learning ‘97 (pp. 289–298). Toronto, Ontario, Canada: Law-
rence Erlbaum Associates, Inc.
Tabak, I., Smith, B. K., Sandoval, W. A., & Reiser, B. J. (1996). Combining general and domain-spe-
cific strategic support for biological inquiry. In C. Frasson, G. Gauthier, & A. Lesgold (Eds.), Intelli-
gent tutoring systems: Third international conference, ITS ‘96 (pp. 288–296). Montreal, Quebec,
Teasley, S. D., & Roschelle, J. (1993). Constructing a joint problem space: The computer as a tool for
sharing knowledge. In S. P. Lajoie & S. J. Derry (Eds.), Computers as cognitive tools (pp. 229–258).
Hillsdale, NJ: Lawrence Erlbaum Associates Inc.
Toth, E. E., Suthers, D. D., & Lesgold, A. M. (2002). “Mapping to know”: The effects of representa-
tional guidance and reflective assessment on scientific inquiry. Science Education, 86, 264–286.
Tzou, C. T., Reiser, B. J., Spillane, J. P., & Kemp, E. K. (2002). Characterizing the multiple dimensions
of teachers’ inquiry practices. Paper presented at the annual meeting of the American Educational
Research Association, New Orleans, LA.
van Zee, E. H., & Minstrell, J. (1997). Reflective discourse: Developing shared understandings in a
physics classroom. International Journal of Science Education, 19, 209–228.
Vygotsky, L. S. (1978). Mind in society: The development of the higher psychological processes (A.
Kozulin, Trans.). Cambridge, MA: Harvard University Press.
Webb, N. M., & Palincsar, A. S. (1996). Group processes in the classroom. In D. C. Berliner & R. C.
Calfee (Eds.), Handbook of educational psychology (pp. 841–873). New York: Macmillan.
White, B. Y., & Frederiksen, J. R. (1998). Inquiry, modeling, and metacognition: Making science ac-
cessible to all students. Cognition and Instruction, 16, 3–118.
Williams, S. M. (1992). Putting case-based instruction into context: Examples from legal and medical
education. The Journal of the Learning Sciences, 2, 367–427.
Wood, D., Bruner, J. S., & Ross, G. (1976). The role of tutoring in problem solving. Journal of Child
Psychology and Psychiatry and Allied Disciplines, 17, 89–100.