ArticlePDF Available

Medium-based design: Extending a medium to create an exploratory learning environment


Abstract and Figures

This article introduces medium-based design—an approach to creating exploratory learning environments using the method of extending a medium. First, the character-istics of exploratory learning environments and medium-based design are described and grounded in related work. Particular attention is given to extending a medium— medium-based design's core method. Then, the product and process of two envi-ronments are detailed to show how medium-based design enables the creation of ex-ploratory learning environments. Finally, a study compares medium-based design to a conventional design process to test the hypothesis that medium-based design is partic-ularly effective in creating exploratory learning environments.
Content may be subject to copyright.
Medium-Based Design: Extending a Medium to Create
an Exploratory Learning Environment
Jochen Rick
College of Computing
Georgia Institute of Technology
phone 404.385.1105
K.K. Lamberty
Division of Science and Mathematics
University of Minnesota, Morris
phone 320.589.6351
This article introduces medium-based design—an approach to creating exploratory
learning environments using the method of extending a medium. First, the character-
istics of exploratory learning environments and medium-based design are described
and grounded in related work. Particular attention is given to extending a medium
medium-based design’s core method. Then, the product and process of two envi-
ronments are detailed to show how medium-based design enables the creation of ex-
ploratory learning environments. Finally, a study compares medium-based design to a
conventional design process to test the hypothesis that medium-based design is partic-
ularly effective in creating exploratory learning environments.
1 Exploratory Learning
In an elementary school classroom, a fourth grader uses a computer program to design a
quilt with the challenge to cover
of the quilt block with one color and
with another
color. He struggles for some time using a 4-by-4 grid. Eventually, he raises his hand and
remarks to the researcher with an emphatic and confused look, “I think it’s really hard to
make one third. Indeed, with the pieces provided, it’s impossible. The researcher agrees
with him and suggests that he try showing
using a 3-by-3 grid.
At a university, two physics-of-music students examine the spectral output of a musical
keyboard using a computer program. Surprisingly, the same note played with the organ
looks different than played with the piano—there’s an extra harmonic above the fundamen-
tal. After some experimenting and reflecting on their own experience, it clicks for one of
the students. “Oh! It’s because you’re getting this pedal tone [points to a harmonic below
This article was published in the journal Interactive Learning Environments
(2005), 13(3):179–212. Interactive Learning Environments is available online at:
Rick & Lamberty Draft: Do Not Distribute
the fundamental] and that [extra harmonic above the fundamental] is a harmonic of that.
Using the program, she explains it to the other student.
Both of these are examples of exploratory learning—learning that occurs through learner-
driven reflective inquiry. Instead of being passive consumers of information, learners are
active explorers of their own understanding. In education, learning by inquiry has a long
history of being championed by theorists, such as Dewey (1938), Fröbel (the originator of
Kindergarten), and Montessori:
Both Fröbel and Montessori emphasize the importance of assisting this “little
explorer” in his researches. With the former this is to be accomplished by the
direct help of the adult; with the latter more indirectly by means of a “prepared
environment” so simplified and set in order that the objects in it easily and
systematically reveal their qualities to the inquiring mind of the little scientist.
(Standing, 1957, p. 327)
Like in Montessori’s system, the learners in our two examples are supported in their inquiry
by the environment. They are environments where learners can act as explorers and discover
the principles of a certain domain. In both cases, the exploration in the environment yields
expectation failure—learners expect something to occur, but what actually occurs fails to
meet those expectations. Expectation failure is particularly important to learning as it allows
learners to reflect on their own understanding (Kolodner, 1997). In the quilt example, the
learner asks the researcher for assistance to resolve his problem; in the organ example,
the learners resolve their problem using previous knowledge and the affordances of the
environment. In both cases, the learning environment enables learners to reflect on their
exploration and thereby build their understanding.
In this article, we concern ourselves with these exploratory learning environments
environments that enable learner-driven reflective inquiry. We are interested in both the
product—what makes for an effective exploratory learning environment (ELE)—and a pro-
cess to support the creation of such environments. Based on previous research, we char-
acterize six defining properties of effective ELEs. While we are able to characterize ELEs
from that literature, a formalized grounded-in-theory approach for creating ELEs is miss-
ing. We articulate and ground such an approach. Drawing on media theory (McLuhan,
1964; Bolter & Grusin, 1999; Kay & Goldberg, 1977), medium-based design (MBD) uses
the technique of extending a medium to create an ELE.
This article is organized into five body sections, concentrating on our goal (ELEs), our
approach for achieving that goal (MBD), and the method by which we do it (extending a
medium). First, we synthesize previous research to characterize ELEs. Second, we de-
tail the important phases of MBD, grounding them in theory. Third, we demonstrate the
power of MBD’s core method of extending a medium. Fourth, we demonstrate how MBD
can be used to create effective ELEs by examining two environments, DigiQuilt and Au-
dioExplorer, designed with MBD. Fifth, we test our hypothesis that MBD is a particularly
effective design method for creating ELEs. After these body sections, we conclude by
discussing the affordances and limits of our approach.
Rick & Lamberty Draft: Do Not Distribute
2 The Goal: Exploratory Learning Environments (ELEs)
In this section, we more closely define what we mean by an exploratory learning environ-
ment. An ELE is an environment that supports learners in constructing their understanding
about a specific subject through learner-driven reflective inquiry. We synthesize related
threads in the literature to arrive at six characteristics of an effective ELE. The first four
(empty, open-ended, structurally simplistic, and concrete) are sufficient for an environment
to be an ELE. The last two (personal connections and epistemological connections) are nec-
essary for the ELE to effectively engage both the learner and important subjects. Naturally,
we are interested in producing effective ELEs—those that meet all six characteristics.
We do not argue that all learning environments should be ELEs nor that all effective
learning environments are ELEs. There are many examples of effective learning environ-
ments that fail to meet some or all of these characteristics. Instead, we argue that these
characteristics define a useful class of environments to support exploratory learning. We
argue that it is the combination of the characteristics that has specific power, more than the
sum of its parts; therefore, it is useful to define ELEs as a separate class.
We do not argue that all learning should be learner-driven and can be solely supported
by the learning environment. Social interaction, such as teacher or expert guidance, is ex-
tremely important to learning. Even in an effective ELE, there is still a need for social sup-
port (Bruckman, 1998); although, the nature of that social support changes. In exploratory
learning, the focus is on the learner learning, rather than the teacher teaching. This does not
mean that the teacher is unimportant. The teacher is still important, but the role changes
from being a “sage on the stage” to being a “guide on the side. Instead of being the trans-
mitter of information, the teacher acts as a guide to assist the learners’ inquiry.
In education, ELEs have a substantial history. They have been championed by educa-
tional theorists (Bruner, 1966; Zucchermaglio, 1993) and educational technologists (Papert,
1993a; diSessa, 2000), under classifications such as microworlds (Papert, 1987), construc-
tion kits (Resnick, Bruckman, & Martin, 1996), media creation tools (Kay & Goldberg,
1977), and inquiry tools (Soloway, Guzdial, & Hay, 1994). Part of what makes ELEs useful
as a class of learning environments is that it is broad enough to encompass a broad range of
previous work, yet narrow enough to have definable characteristics. We start by detailing
the four sufficient characteristics (empty, open-ended, structurally simplistic, and concrete)
for an environment to be considered an ELE:
Empty An ELE is empty of content; the learner “fills” it to establish its meaning.
Zucchermaglio (1993) broadly categorizes learning environments
into two groups: empty
and full. Full environments are full of content for the learner to absorb. In contrast, empty
environments do not contain explicit content. Instead, the message of the empty environ-
ment is only realized when it is engaged by the learner.
Zucchermaglio actually uses “technology, instead of “environment. In this research, we do not believe
that technology (or implicitly new technology) deserves special treatment from previous technologies. After
all, it is only in relationship to humans that old and new technologies achieve their meaning.
Rick & Lamberty Draft: Do Not Distribute
Full environments, such as the majority of Computer-Assisted Instruction and Intelli-
gent Tutoring Systems, are designed under the premise that learning is a process of trans-
ferring content from the environment to the learner (Zucchermaglio, 1993). The full envi-
ronment acts as a repository for content; the content is transferred to the learner. In con-
trast, empty environments are based on a constructivist model of learning, where learners
construct their own understanding (Zucchermaglio, 1993). The learner engages the envi-
ronment to construct meaning.
For example, a textbook is a full environment. It is full of content for the learner to
absorb. The message of the textbook is its content. In contrast, a coloring book is an empty
environment. The learner engages it by filling it with color. The message of the coloring
book is the skill the reader achieves by engaging it. At the end of the learning process, the
content of the textbook is unchanged from reading, while the content of the coloring book
is largely constructed by the learner.
Another example of the difference between empty and full is that of piano versus stereo.
The piano is empty; a player must actively engage it to produce music. The stereo is full; a
listener passively engages it to consume music. As a learning environment to foster inquiry,
the piano is more promising than the stereo (Resnick et al., 1996).
Open-Ended The interaction facilitated by ELEs is open-ended; the learner chooses how
to explore.
This characteristic addresses learner control. Is the learner in control of his or her learning
process? ELEs are open-ended—the learner is significantly in control of their own learning
process. Learners can use them at different rates and in different ways. As such, they
can support different ways of approaching subjects, which is important as people learn in
different ways (Turkle & Papert, 1991).
In many common learning environments, learner control is non-existent or superficial;
it is the environment that is in control (Papert, 1987). Consider an educational television
show or a typing tutor. In an educational television show, the learner has no real choices
about the information being transmitted or the style of interaction. Learner control is non-
existent. In a typing tutor, the learner has significant interaction (typing in letters) with the
tutor, but the tutor is in control of the content (letters to type, difficulty, level, assessment,
etc.). Any sense of learner control is superficial.
In contrast, either a coloring book or a piano allows for open-ended interaction. One
person can concentrate on drawing within the lines. Another person can concentrate on
using matching colors. One person can concentrate on learning classical piano technique.
Another person can concentrate on using fake books
to play popular hits for singing along
with. For both a coloring book and a piano, the learner has a real choice of how they engage
the environment.
While ELEs can be characterized as open-ended, their use need not be totally unre-
stricted. For instance, it would be foolish to sit a novice in front of a piano and expect them
A fake book contains condensed information about songs (lyrics, melodies, and chord progressions). It is
used by musicians as an abridged substitute for standard sheet music.
Rick & Lamberty Draft: Do Not Distribute
to learn to play beautiful music by themselves. Appropriate challenges, such as playing
a particular piece of music, and guidance can aid the learning process tremendously. Yet,
the piano allows the type of challenge and guidance to be influenced by the desires of the
Structurally Simplistic An ELE is structurally simplistic; in the environment, the learner
engages a few powerful structures and strictures.
Simplicity has been touted as an important concept in design. In user-interface design,
simplicity is often seen as a fundamentally good property (Norman, 1988). Here, we hope to
go beyond simplicity in interface. ELEs are simple in a specific way—they are structurally
simplistic. This is a term we take from diSessa’s (1985) descriptions on how he designed
the Boxer computation system. To assure that users can accomplish many things while only
mastering a few structures, diSessa made structural simplicity an explicit goal of his design.
ELEs are characterized by a few fundamental structures (constructs) and strictures
(constraints). In an ELE, the learner masters the environment. Having few structures and
strictures enables that mastery to be accomplished. Mastering the fundamental structures
and strictures of the environment allows learners to grasp the embodied concepts—to learn.
A good exploratory learning environment has only a few structures and strictures that offer
powerful affordances for learning, while simultaneously being simple enough to master. In
other words, the structures and strictures should offer a good payoff for effort. One way they
often offer a good payoff for effort is that they afford multiple representations of the fun-
damental structures. Different representations can have different affordances for learning,
so providing multiple representations (often simultaneously) is useful for learning (Suthers
& Hundhausen, 2003; Kaput, 1989). Because there are a small number of fundamental
structures, it is possible to represent them in different ways.
Consider the piano. It is structurally simplistic. It has a limited amount of structures.
Even on full-size pianos, there are only 88 keys to strike. Furthermore, those keys are orga-
nized into octaves, with the black-and-white-key pattern repeating every 12 keys (i.e. one
octave). It has a limited amount of strictures. Each key is restricted to a single fundamental
frequency. There are plenty of musical tones between C and C#, but the piano does not
allow you to play them; other instruments, such as the violin, would. Of course, the success
of the piano is not just because it has structures and strictures, but because those structures
and strictures map to qualities we look for in western music. The notes played by the piano
keys are the ones used in western music. Even the black-and-white-key pattern is represen-
tative of western music—a piece of music restricted to the key of C major or A minor only
uses the white keys. So, when learning about music, it is useful to start with a piano, rather
than an entire orchestra (Rick, 2002b). The orchestra is much more complicated than the
piano and that complexity would obscure those structural connections to western music. In
contrast, the piano is simplistic in such a way to have obvious structural connections to the
nature of western music.
Concrete An ELE offers concrete connections to important concepts; the learner is able to
reflectively engage the concepts in the environment.
Rick & Lamberty Draft: Do Not Distribute
A domain of study can often be represented in an abstract hierarchy, with concrete instanti-
ations at the bottom of that hierarchy. Justified by a striving towards incremental learning,
some draw the conclusion that the abstractions should be learned first. Intuitively, it seems
easier to first teach the abstractions, so that this knowledge can be transferred to the instan-
tiations. In this top-down approach to learning, learners first address the top classifications
(or abstractions) and then move down the hierarchy to the instantiations (concrete exam-
ples). But, it is often the objects at the bottom of that hierarchy with which we are most
familiar and have the most connections. As such, we can draw upon those previous con-
nections to understand the more abstract concepts (Wilensky, 1991). Learning research
has recently revalued concrete ways of learning and shown that starting from the concrete
with a bottom-up process can be useful (Kolodner, 1997; Turkle & Papert, 1991). ELEs, in
essence, enable the learner to use these concrete connections as a bridge to the more abstract
concepts. As part of the inquiry process, learners engage the domain of study concretely in
the environment.
Consider learning about heat conductivity from a laboratory experiment versus learning
it from a book. While they might cover the same material, the laboratory experiment repre-
sents the material in a concrete manner, while the book represents the material in an abstract
manner. Besides having hands-on experience with relevant materials, the laboratory exper-
iment has the advantage of allowing for reflection and expectation failure. Learners can try
something to test their understanding. So, ELEs are concrete to facilitate reflective inquiry
So far, we have characterized ELEs as empty, open-ended, structurally simplistic, and
concrete. While some of these characteristics could be viewed independently as beneficial,
we feel that there is a strong synergy between these characteristics. Just because an environ-
ment is empty does not mean that inquiry learning will occur. Just because an environment
is open-ended does not mean that learners will use it in a way that is useful to their learning
style. The structures and strictures the learner engages have to embody concrete connec-
tions to the subject in order for the exploration in the empty open-ended environment to
be fruitful. Because exploratory learning environments are meant to be mastered (due to
their structural simplicity) and are empty, they allow for open-ended use that can be quite
motivating for the learner. The learner can chose what to explore, construct, or analyze.
They can do that in a way that is personally interesting to them.
While these four characteristics are sufficient for an environment to be considered an
ELE, this does not guarantee that it is an effective learning environment. Consider the piano.
It meets the four characteristics. Yet, it is doubtful that simply sitting a novice in front of
a piano will lead to useful learning. This is all the more true if the learner is unfamiliar
with western music. To address effectiveness for a given learning situation (environment,
learners, social interaction, setting, etc.), two other characteristics are added, based on work
by Resnick et al. (1996). While they champion these characteristics for construction kits,
we believe they apply to all exploratory learning environments.
Personal Connections An effective ELE is able to leverage previous interests and experi-
ences to connect new concepts to pre-existing concepts (knowledge, intuitions, etc.).
Rick & Lamberty Draft: Do Not Distribute
Epistemological Connections An effective ELE is able to address important concepts (do-
mains of knowledge, ways of thinking, etc.) through natural exploration (inquiry,
construction, etc.).
In the case of the complete novice, the obstacle to the piano being an effective ELE is that
the novice needs more support than is provided in the environment to usefully leverage
previous interests and experiences. In contrast, an experienced composer may find that a
bare piano is all that is necessary to explore and create new music. To be effective, an
ELE needs to be both appropriate to the learners (personal connections) and to the subject
(epistemological connections).
2.1 ELEs as Scaffolding Sense-Making
ELEs may be a new concept (in this classification anyway), but the support they provide
for learning has already been established. To further that argument, we show how the
characteristics of ELEs address the three sense-making guidelines in Quintana et al.s (2004)
framework for scaffolding science inquiry learning.
1. Use representations and language that bridge learners’ understanding. Because of
their structural simplicity, ELEs can link representations to increase learner under-
standing. The concrete nature of ELEs allows learners to manipulate things they are
familiar with to engage more abstract domain concepts. Because of their personal
connections, ELEs leverage previous experiences to connect the new concepts to pre-
existing ones, thereby bridging learners’ understanding.
2. Organize tools and artifacts around the semantics of the discipline. Because of their
concrete connections to important concepts, ELEs allow learners to engage disci-
plinary concepts while engaging in the kind of open-ended exploration common to
science disciplines.
3. Use representations that learners can inspect in different ways to reveal important
properties of underlying data. Much of what an ELE tries to accomplish is to allow
the learner to explore new aspects of the environment through a variety of lenses or
tools. Because of their empty and open-ended nature, ELEs can be used in different
ways. Because the lenses focus the learner’s attention in different ways, they reveal
important properties of the underlying environment.
While the correspondence is not one-to-one, the characteristics of ELEs match well to the
guidelines that Quintana et al. posit for supporting science inquiry. Not all exploratory
learning is inquiry science learning, but the two seem to be congenial. By showing that
ELEs address important goals in supporting inquiry science learning, we aim to show that
ELEs may be of interest to that established research community.
Rick & Lamberty Draft: Do Not Distribute
2.2 Demarcation: ELE’s Boundaries
So far, we have woven several intellectual threads together to form our ELE classification.
Consequently, it is no surprise that the category is rather broad, covering under its umbrella
what others have termed microworlds, construction kits, inquiry tools, and media creation
tools. What are the boundaries of the ELE category?
Like other important dimensions in education (Reeves, 1994), the boundaries of the
individual characteristics are fluid. Environments span the range from being empty to being
full, from being open-ended to being constrained, from being structurally simplistic to being
complex, and from being concrete to being abstract. Consequently, the boundaries of the
ELE category are fluid as well—there is no definitive boundary. Instead, we define a loose
boundary: effective ELEs are the group of environments that meet all six characteristics to
a fair extent. As the characteristics interact, it is more important that an environment meets
all characteristics than that it excels in a subset of them.
Many of the six defining characteristics have been touted by their originators to apply
to a broad class of learning environments. Zucchermaglio (1993) prefers empty environ-
ments. Kolodner (1997) advocates a concrete approach. As their arguments have merit, it
is only natural that many effective learning environments meet their characteristics. Yet, an
effective environment can meet some of our characteristics and fail to meet others.
For instance, goal-based scenarios (Schank, Fano, Bell, & Jona, 1994) explicitly feature
concrete connections; however, they are not empty. Much of the knowledge gained from
using a goal-based scenario is a consequence of the scenario being full of content. So, goal-
based scenarios are not ELEs. Similarly, many collaborative learning environments, such
as CoWeb (Guzdial, Rick, & Kehoe, 2001) and CSILE (Scardamalia & Bereiter, 1991), are
empty, open-ended, and structurally simplistic; however, as domain-independent collabo-
ration tools, they do not offer concrete connections to specific learning goals. They are not
By classifying these environments outside the ELE umbrella, we do not mean to ques-
tion their effectiveness or marginalize them. To the contrary, we offer them as examples of
effective learning environments that are not ELEs. It is not necessary for an environment
to be an ELE to be an effective learning environment. Yet, we do believe there is value to
being an ELE.
2.3 From Goal to Approach
As designers, we are interested in creating effective ELEs—that is our goal. In this section,
we detailed the characteristics of that goal. While knowing these characteristics allows us
to identify ELEs and reflect on whether our design meets these characteristics, there is more
to effective design than that. If designing an ELE is similar to creating a movie, knowing
the characteristics of ELEs is similar to being able to identify the properties of a good film.
It may be necessary to make a good movie, but it is not sufficient. The approach towards
that goal is anything but clear. In the next session, we describe an approach (MBD) that we
have found useful for achieving our goal (creating ELEs).
Rick & Lamberty Draft: Do Not Distribute
3 The Approach: Medium-Based Design (MBD)
Making the simple complicated is commonplace; making the complicated sim-
ple, awesomely simple, that’s creativity. —Charles Mingus, Jazz composer and
By their nature, exploratory learning environments are simple. Unfortunately, simplicity is
hard to design for; it takes creativity. Many learning environments suffer from too many
features. One reason for this is that a conventional problem-based design approach tends
to add more structures and strictures over time. As more are added, those structures and
strictures lose their simplicity and thereby their conceptual coherence. ELEs need to be
structurally simplistic; unfortunately, a conventional hard-style design makes simplicity a
hard goal to achieve. So, we are offering an alternative approach that, in our experience, is
useful for designing ELEs.
Turkle and Papert (1991) note how learners approach tasks with two distinct styles: a
hard and a soft one. Conventional design methods have valued a hard-style approach, ne-
glecting the soft-style designer. The soft style is different from but not worse than the hard
style (Turkle & Papert, 1991). The hard style is a structural and top-down approach. The
large problem is divided into smaller more-manageable problems; solving the smaller prob-
lems solves the larger problem. In contrast, the soft style is bottom-up and negotional. It
involves a closeness to objects. While hierarchy and abstraction are valued by the structured
problem solvers, these bricoleurs prefer negotiation and rearrangement of their materials
(Turkle & Papert, 1991). Traditionally, the hard style has been given a higher status, while
the soft style has been maligned. Yet, the soft style too has its advantages; even in Western
science, where a distance from the object of study has been strongly encouraged, there are
examples of breakthrough success using a soft style (Turkle & Papert, 1991).
Our design method, medium-based design, offers designers an alternative—a soft-style
approach to designing learning environments. MBD extends a medium to create an ELE.
We draw on media theory (McLuhan, 1964; Bolter & Grusin, 1999) to realize the power of
a medium (a way of creating, consuming, and transmitting information) to have a message
to those that engage that medium. We draw on the power of the computer to create new
media (Kay & Goldberg, 1977) to realize the potential of a new medium as a learning
We do not believe that MBD is a completely new approach to designing learning en-
vironments. Rather, we believe that designers have intuitively used similar methods, and
that, by describing details of MBD and providing guidelines that follow this approach, we
are showing support and value for this approach. This section describes the four important
phases of MBD, grounding each in media theory.
Phase 1 Begin by choosing a medium you know well, care about, and feel has learning
Man’s use of mind is dependent upon his ability to develop and use “tools” or
Rick & Lamberty Draft: Do Not Distribute
“instruments” or “technologies” that make it possible for him to express and
amplify his powers. (Bruner, 1966, p. 24)
A medium is a way (as in the Bruner quote above) to amplify the powers of man (McLuhan,
1964). We base our design on that medium, in an effort to harvest the powers of the medium
for learning. From there, we seek to realize an exploratory learning environment.
MBD is a soft style of design. We aim not to hide this, but rather to embrace it and
use it to our advantage. So, the MBD designer should start with a medium that they are
close to—they understand it and care about it. This closeness is an asset as the designer is
seeking to create an environment that has personal connections to the learners who use the
environment. To better illustrate our notion of closeness, consider Papert’s (1993b, preface)
fond recollection of the differential gear. As a child, he (to use his words) fell in love
with the differential gear and the mathematical concepts it embodied. He connected new
concepts to the differential gear; the differential gear became a powerful medium for him
to understand his world. For Papert, the differential gear is a medium he is close to—he
knows it well, cares about it, and feels it has a learning potential. It is that kind of closeness
to the medium we are looking for.
Phase 2 Explore the medium to understand its affordances for promoting learning.
The medium is the message. (McLuhan, 1964, p. 7)
What McLuhan means by his infamous proclamation is that “societies have always been
shaped more by the nature of the media by which men communicate than by the content of
the communication” (McLuhan & Fiore, 1967, p. 8). The nature of the media by which we
communicate not only carry content, but also carry a message independent of the content we
try to communicate. That message is important to the individual engaging the medium; it
works him or her over completely (McLuhan & Fiore, 1967). The medium has affordances
and the message to the individual is strongly affected by those affordances.
The goal of MBD is not to use various media to convey some message independent of
those media, but rather to design one medium so that it is the message. Yet, the affordances
of a medium are not inherently obvious. In particular, the importance of a new medium is
often not discovered until it has been used for a while (McLuhan, 1964; Bolter, 2001). Peo-
ple tend to fit the affordances of established seemingly-related media into the new medium;
through this process they discover the unique affordances of the new medium (Bolter &
Grusin, 1999). For instance, early cinema was often hardly more than filmed stage theatre;
over time, film evolved into its own genre. Understanding a new medium is a journey of
discovery that requires serious exploration.
Consider diSessa’s (1987) enhancement to Logo. Traditionally, Logo allows users to
program a digital turtle to draw shapes on the monitor screen, a 2-dimensional plane.
diSessa wondered if anything could be gained by mapping the 2-dimensional plane onto
a 3-dimensional cube. Mathematically, it was doable; however, it was not clear that the new
environment would afford any new uses. For several weeks, he experimented with the idea
Rick & Lamberty Draft: Do Not Distribute
on paper, finding no useful differences; turtles behaved the same in every scenario he imag-
ined. Consequently, he became discouraged with the idea. As a last resort, he implemented
a working prototype. Almost immediately upon completing the prototype, he discovered
several interesting differences. For instance, if the turtle was positioned properly near the
corner of the cube, it was possible to draw a triangle using three 90-degree angles. diSessa
was only able to realize the affordances of his new medium after he built it. Only after he
built it could he properly explore it.
Fundamental to MBD is that the designer must explore what the message is of the
medium they are working with. As with any design process, this must be a reflective prac-
tice (Schön, 1987). The designer must explore what the medium is good for. In our expe-
rience, this exploration cannot be done theoretically. The designer must actually play with
the (new) medium and try to figure out what the affordances are. In practice, this involves
building prototypes (implementations that work to a certain extent). These prototypes must
be sufficiently developed and useful to allow the designer to actively explore (and even
extend) the medium.
Phase 3 Extend the medium to change those affordances.
We shape our tools and they in turn shape us. (McLuhan, 1964, p. xxi)
It is our experience that differences in a medium can create large differences in its affor-
dances for learning. In essence, extending a medium creates a new medium with different
affordances (McLuhan, 1964). So, it is the MBD designer’s task to extend the medium
(shape the tools) so that important learning goals (how it shapes us) are realized. Extend-
ing a medium is the primary method of MBD on both a large and a small scale.
Large extensions extend the medium along a fundamental (physical, temporal, or quan-
titative) dimension. Typically, they dramatically alter the learning affordances of a medium.
Consider, for instance, extending a video camera along a temporal dimension. A standard
video camera captures 30 frames per second; this is useful for recording real-time events.
In contrast, a time-lapse camera may only capture one frame every five minutes; this is
useful for recording large-scale events, such as the lunar cycle (Terry, Brostow, Ou, Tyman,
& Gromala, 2004). As a learning tool, the time-lapse camera focuses the user on a different
subject than the video camera.
Small extensions change the medium by adding new structures (constructs) and stric-
tures (constraints). Typically, they highlight specific learning goals. Different modes can
focus a learner on different aspects of the subject. A new lens can offer a new representation
for the learner to connect with (Kaput, 1989).
Because of its importance to MBD, extending a medium will be further detailed in this
article. Section 4 details two examples of how extending a medium works on a large scale.
Section 5 details how this method functions on a small scale as we detail the design of two
ELEs created with MBD.
Creating a medium and realizing its affordances in ways that others will be able to en-
gage them is a difficult task. Fortunately, adding computation gives us additional flexibility
Rick & Lamberty Draft: Do Not Distribute
in designing media (Resnick, 1998). The computer as the first meta-medium, a medium
to reinterpret previous mediums and create entirely new ones, provides a great vehicle for
designing new media (Kay & Goldberg, 1977).
Phase 4 Transition from the solution (the medium) to the problem (a clear message).
MBD proceeds by investigating the environmental needs and social context necessary for
making those affordances recognizable and graspable. In a hard-style design process, it
is important to first clarify and investigate an important problem. In contrast, the MBD
designer may find that the medium actually supports learning goals that are substantially
different than that first intuition. As such, the MBD designer cannot be solely ruled by the
constraints of the problem (the message), but must also take into considerations the con-
straints imposed by the solution (the medium). Therefore, in MBD, it is important to first
clarify and investigate the solution—the medium. Taking the constraints of the solution
space seriously and, at times, redefining the problem space to better match that solution
space is something accomplished designers do often (Goel & Pirolli, 1992). Solving an
important learning problem is still essential to the goals of MBD, but that does not neces-
sitate that the method have its initial focus on the problem. In both hard-style design and
MBD, solution and problem evolve together; the difference is that the initial focus is on the
problem in the hard-style design and the solution in MBD.
The closer a design comes to fruition, the less it can be dependent on MBD, or any de-
sign method, whether it be hard or soft. The situation matters (Lave & Wenger, 1991). As a
learning situation can be realized into concrete scenarios (Carroll, 2000), the specific needs
of that learning situation (students, teacher, school, etc.) become primary. At that point,
designers have to do what is right for their learners’ needs (Soloway et al., 1994) and the
design process can best be characterized as learner-centered iterative development. Some-
times, the medium is not enough and other learning supports will need to be added (Joseph
& Nacu, 2003). Social and process support may be needed to support the environmental
affordances (Quintana et al., 2004). For example, in MOOSE Crossing (an on-line commu-
nity for kids to share their creative writing through programming), the on-line community
provides both motivation and learning support for its members (Bruckman, 1998).
4 The Method: Extending Media
One of the foundations of MBD is that extending a medium is a powerful way of creating
a learning environment, since the extended medium has different affordances for learning
than the original medium. In this section, we aim to demonstrate this with two cases. In the
first, we examine two physical manipulatives: fraction sticks and algebra rectangles. In the
second, we examine two computer-based learning environments: Logo and StarLogo. In
each case, one medium extends the other leading to different learning goals being addressed
by the learning environment. We detail both a physical manipulative and a computer envi-
ronment to show that the principle of extending a medium applies to all media, from simple
blocks to sophisticated programming languages.
Rick & Lamberty Draft: Do Not Distribute
Figure 1: Fraction Sticks—different lengths represent different fractions
+ + +++
+ +
Figure 2: Examples of how 1 can be split up into fractions using Fraction Sticks
4.1 Example 1: Sticks and Rectangles
Fraction sticks (see Figure 1) are sticks of various lengths that represent different fractions
with their length (the height and width are the same for all sticks). So, a fraction stick
is twice the length of a stick representing
. Fraction sticks are easy to align
end-to-end. Combining fraction sticks in this manner essentially creates a new fraction stick
whose value is the sum of its component sticks. So, the sticks afford easy (visual) addition.
Also, they afford comparing, as it is easy to see that one stick is longer, shorter, or equal to
the length of another stick by laying it next to the other stick. Visually, a
stick is larger
than a
stick, like
is a larger fraction than
. As learners play with these sticks, they
can see how various fractions compare to each other. Combining these two affordances,
learners can see how the sum of various fractions compare to each other (see Figure 2). So,
it is easy to see that three
sticks and two
sticks have the same length, or, mathematically,
Mapping fractions to the lengths of these sticks creates a useful medium that affords
learners to encounter important learning goals for fraction learning—how fractions compare
to each other and how the sum of these fractions create new fractions. Other themes can be
usefully mapped to these sticks. Cuisenaire rods, for example, are very similar (rods differ
Rick & Lamberty Draft: Do Not Distribute
Figure 3: Algebra Rectangles—different areas represent different quantities (a x by 1 rect-
angle has an area of x)
x 2
x +
is equal to
x(x+4)+4 (x+2)
x +4x+4= =
Figure 4: Using Algebra Rectangles to show that different equations are the same
in length), but they are not necessarily mapped to fractions. Yet, fractions work well as
there is a clear and simple mapping between the learning goals (addition and comparison
of fraction) and what the medium affords (addition and comparison of lengths).
Fraction sticks differ in one dimension—their length. One way to transform a medium
into a different medium is just to extend a dimension. A simple dimension to extend is a
physical dimension. Here, we look at what happens when we extend the width. The sticks
are converted into rectangular blocks that share the same height, but differ in length and
width. Whereas sticks afford thinking about length, rectangles afford thinking about area,
the product of the length and the width.
When Bruner (1966) uses these kind of rectangles, he does not map fractions to the
dimensions, but chooses to use whole numbers and algebraic symbols (see Figure 3). Since
areas of rectangles are determined by the products of their length and width, these alge-
bra rectangles afford multiplication. 4 can be represented by four 1-by-1 rectangles. The
symbol x can be represented by an x-by-1 rectangle. The symbol x
can be represented
by an x-by-x rectangle. In Bruner’s use, the learners are not concerned with solving for x,
but rather in using it symbolically. So they can construct equations visually that can then
be converted to algebraic equations (see Figure 4). As area is retained when the shapes
are rearranged, these rectangles afford thinking about equalities. Using Bruner’s algebra
rectangles, learners can explore algebraic equalities.
Rick & Lamberty Draft: Do Not Distribute
Mapping numbers and algebraic symbols to the length and width of these rectangles
creates a useful medium that affords learners to encounter important learning goals of
algebra—how simple algebraic expressions can be multiplied. The algebraic theme works
well as there is a clear and simple mapping between the learning goals (multiplication of
simple algebraic expressions) and what the medium affords (rearrangement and decompo-
sition of rectangles by area).
Though these two media (fraction sticks and algebra rectangles) are very similar mor-
phologically (the one extends the other into another physical dimension), they are quite dif-
ferent in their learning goals. Fractions sticks afford learning about the addition of fractions.
Algebra rectangles afford learning about the multiplication of simple algebraic expressions.
Not only are these different learning goals, but the learning goals are of interest to a dif-
ferent audience. Traditionally, schools reserve learning about algebra for a later grade than
learning about fractions. So, when extending a medium, MBD designers may find that their
ELE addresses different learning goals and a different group of learners for whom those
goals are important. Designing with a medium for learning necessitates focusing on the
solution space (the medium), as the problem space (learning goals and learners) can change
quite drastically. Trying to impose a certain learning goal or audience on a medium is futile
if the medium does not afford it. So, MBD necessarily must be a process that starts in the
solution space.
4.2 Example 2: Turtle(s)
In the previous case, we focused on two similar physical manipulatives that demonstrate that
an extension to a medium can transform that medium’s learning affordances drastically, to
the point that new learning goals are addressed by the environment. In this example, we
aim to demonstrate this effect again in a different setting—programming languages. We will
describe Logo and StarLogo, show how the one extends the other, and how this drastically
affects the learning goals of the two ELEs.
Logo is a widely distributed programming language designed for children (Papert,
1993b). With Logo’s turtle-graphics, children can program a virtual turtle, shown as a tri-
angle (or as a bitmap turtle in some versions of Logo), to move around the screen, thereby
drawing lines and shapes (Papert, 1987). For example, when FORWARD 100 (or FD 100)
is executed, the turtle steps forward 100 units, leaving behind a straight line of 100 units
in length. When RIGHT 90 (or RT 90) is executed, the turtle turns 90 degrees to the right.
Repeating this sequence four times results in a square with sides of 100 units (left part of
Figure 5). Unlike previous programming languages’ graphic systems, turtle graphics is not
based on drawing lines from one coordinate point to another coordinate point. Instead, it is
based on a body syntonic model—the programmer can imagine themselves in place of the
turtle when trying to create a shape (i.e. “How would I move in a circle?”). In the case of
the circle, it is easy to realize that all it takes is repeatedly going forward a bit and turning
a bit (right part of Figure 5).
Because the turtle can make the same movements as the learner (go forward, go back,
turn right, etc.), turtle graphics afford learners mapping their understanding of movement
Rick & Lamberty Draft: Do Not Distribute
// Drawing a Circle
REPEAT 360 [FD 1 RT 1]
// Drawing a Square
REPEAT 4 [FD 100 RT 90]
Figure 5: Using Logo to draw a square and a circle
to the virtual turtle. Because the virtual turtle can move precisely and leave a trail showing
where it has been, turtle graphics afford drawing simple geometric shapes, such as squares
and circles. Thus, by playing with turtle graphics, learners can explore geometry in a way
that leverages their understanding of movement. Viewed from McLuhan’s (1964) perspec-
tive that media extend man, Logo’s turtle is a medium that extends man’s legs. The turtle
movement is similar to the movement that our legs afford us. As the turtle medium addi-
tionally affords moving precisely and keeping a useful record of that movement, it affords
us understanding geometry better than simply moving with our legs.
In the previous case, we showed how extending a physical dimension (width) changed a
medium. In this example, we show how extending quantity can change a medium. Instead
of the learner programming one turtle, what happens when the learner is programming a
massive number of turtles? StarLogo is a programming language that allows learners to do
this (Resnick, 1994). As a programming language, StarLogo’s syntax is nearly identical to
Logo’s. The main difference is that users program one turtle in Logo and program massive
amounts of turtles in StarLogo. In addition, StarLogo turtles have better senses and their
environment, a grid of patches, also has computational power (Resnick, 1996). So, for
instance, a StarLogo turtle may pass on commands to the patch below it; the turtle senses
the patch below it and can send a command to that patch, which can then, for instance,
change one of its variables. When Resnick (StarLogo’s designer) uses StarLogo, he is
not using it to explore geometry, but to develop simulations of decentralized behavior, as
displayed by slime mold, termite colonies, and traffic jams (Resnick, 1994).
Figure 6 shows several frames of such a simulation—an ant colony foraging for food.
Initially, all of the ants are located in the anthill (the large mound in the center) and four
mounds of food are placed at different distances from the anthill (Frame 1). When the simu-
lation starts, the ants scurry out of their anthill. According to their programs, unless sensing
pheromone nearby, the ants just move around randomly (Frame 2). At some point, some
of the ants will come upon some food (located in the patch below) by chance. When this
happens, their behavior alters. They pick up the food, decrementing the food variable in the
Rick & Lamberty Draft: Do Not Distribute
1. 2. 3.
4. 5.
Figure 6: Frames from an ant-colony simulation in StarLogo
patch below, and take it back to the anthill, dropping pheromone to the patches below them
along the way. The patches will slowly diffuse and dissipate this pheromone. When ants
(who are not carrying food) realize that there is pheromone in their immediate proximity,
they behave differently. They start moving in the direction of the strongest pheromone. At
this stage, they will find food in that direction and help to enforce that pheromone trail by
bringing the food back to the anthill. Strong pheromone trails will develop to the food,
allowing most of the ants to become a productive member of one of these trails, bringing
food back to the anthill and enforcing the pheromone trail (Frame 3). After a short while,
the food at the end of these trails will have been exhausted. But, the ants still try to find
food at the end of the trail, as the pheromone is still strongest there (Frame 4). Eventually,
the pheromone is dissipated by the patches and the ants scatter, reverting to their earlier
behavior (like Frame 2). New pheromone trails will develop (like Frame 3) and food there
will be exhausted (like Frame 4). Eventually, all four food mounds will have been taken
back to the anthill and the ant movement will simply be random (Frame 5). The simulation
is over.
Though the ants in this simulation are efficient at finding food and bringing it back
to the nest, they are not being commanded by a centralized force. There is no queen ant
giving orders. Instead, this anthill is a decentralized, yet organized, system. All ants follow
fairly simple procedures, based on their local information. Yet, the emergent behavior
seems organized (particularly in Frame 3). Resnick (1996), in his studies with StarLogo,
found that most people have a hard time attributing organized behavior to decentralized
causes. They get stuck in a centralized mindset, mistakenly attributing emergent behavior to
centralized causes. StarLogo provides a medium for learners to explore and even construct
such decentralized systems (Resnick et al., 1996).
Though these two media (Logo’s turtle-graphics and StarLogo) are similar morpholog-
ically (the one extends the other by increasing the quantity of turtles to program) and one
Rick & Lamberty Draft: Do Not Distribute
was even developed from the other (Resnick, 1996), they are quite different in their af-
fordances for learning. Logo’s turtle-graphics focuses on geometry. StarLogo focuses on
decentralized systems. Like fractions and algebra in the previous example, these different
learning goals are appropriate for a different group of learners. So, this comparison gives a
further example of how extending a medium (working in the solution space) can drastically
affect the learning goals and audience addressed (the problem space).
In addition, it gives a great example of how starting in the solution space might have
additional advantages to starting in the problem space. Papert argues that StarLogo is a clear
example of the solution leading the problem (Resnick, 1994, foreword). Learning about
decentralized systems is not something that you will see in school curricula. So, educational
designers constrained to a certain problem space (addressing an important problem in a
certain curriculum) would never have created StarLogo. Yet, decentralized systems are
important. StarLogo offers a learning environment where these important concepts can be
explored in ways that were not nearly as accessible to learners before its creation.
A problem with a design approach, like MBD, that starts in the solution space is that
the learning environment designed may not address any important learning goals. While
we acknowledge this as a real concern, we feel that the advantages of having a formalized
process for creating an ELE outweigh this advantage of a design approach that starts in
the problem space. Yet, as StarLogo demonstrates, an approach that starts in the solution
space may actually have a corresponding advantage—the solution (the learning environ-
ment) could solve a problem (an important learning goal) that had not been acknowledged
before then.
5 Using Medium-Based Design
The previous section detailed how extending media is a powerful design method; however,
it did not show that this method was actually used in the design of these ELEs. The analysis
was based on the product and not the process of design. One cannot deduce process from
product (Goel & Pirolli, 1992). As such, we are not claiming that fraction sticks, algebra
rectangles, Logo, or StarLogo were designed using MBD. As these learning environments
were designed before MBD was articulated, the designers clearly did not design under the
MBD name, even if their processes were similar.
In this section, we concern ourselves with the design process. We examine two learning
environments, AudioExplorer and DigiQuilt, which were developed by the authors using
MBD. Both are similar in scope, yet different in approach. Both were implemented in
Squeak (Guzdial & Rose, 2002), a system purposefully engineered for creating these kinds
of environments. Both are similar in complexity. Where they differ is that AudioExplorer
is an inquiry tool and DigiQuilt is a constructionist medium; these are two large categories
of ELEs (Hay & Barab, 2001). Comparing across these major categories should make the
conclusions applicable to the design of a broad range of ELEs.
Rick & Lamberty Draft: Do Not Distribute
Figure 7: Using AudioExplorer to investigate the first six harmonics of a C note
5.1 AudioExplorer: an Inquiry Tool
AudioExplorer is a computer-based inquiry tool to explore the physics of sound by exam-
ining the frequency domain (Rick, 2002a). The frequency domain is a transformation of
the sound signal into its frequency components. Since our ear perceives frequencies, ex-
amining the frequency domain is a useful way to understand the properties of sound, such
as the harmonic sound that is essential to music. The learning system consists of a music
keyboard transmitting sound input to a computer (Figure 7); the AudioExplorer software
displays the signal on the screen, which can then be analyzed by the learners.
The idea for AudioExplorer came from a wish to explain why there are twelve half-
tones in an octave. Given Rick’s background in digital signal processing, he knew that this
could be found out by investigating the frequency signal of harmonic sound. So, the solu-
tion was a medium for investigating the frequency signal. Rick started playing around with
Squeak’s sound capabilities and various sound inputs. Widgets for displaying sonogram
(frequency signal over time) and spectrum (frequency signal at one time) graphs already
existed. These widgets were tested with various sound inputs, including singing, playing
music off a CD, whistling, etc. To capture lower frequencies with enough resolution, the
size of the fast Fourier transform (FFT) was increased. In addition, both the spectrum and
the sonogram graph were placed on the screen at the same time (this involved rotating the
spectrum graph by 90 degrees). By adding a musical keyboard as input, the harmonics of
the individual notes could best be analyzed. From there, an analysis tool was added so
that the user(s) could investigate the frequency of individual harmonics and that frequency
would be directly linked with the closest note on the keyboard. Because of the knowledge
of signal processing, Rick knew that AudioExplorer could demonstrate the relationship be-
tween the linear (harmonics have frequencies that are integral multiples of the fundamental)
and exponential (frequency of the fundamental increases exponentially with each octave)
Rick & Lamberty Draft: Do Not Distribute
properties of harmonic sound, as can be found in western music. Yet, it was not clear how
learners would use the system.
Creating a tool that the designer can use to demonstrate the learning concepts might
be a necessary first step, but it does not create an effective ELE. Personal connections are
necessary; others would have to be able to engage the medium in a useful way. To test out
how others could engage the system, Rick, inspired by reflection-in-action (Schön, 1987),
conducted informal user testing. He sat down with several people and showed them the
system, letting them play around and helping them when necessary. In the end, this format
proved awkward. Users were more interested in Rick demonstrating the concepts than in
being guided towards figuring them out for themselves. Yet, it demonstrated that only a
little bit of appropriate guidance was necessary to allow users to engage the system.
Based on a suggestion from a colleague, the possibility for using the system to inves-
tigate the differences between instruments was researched and proved viable. In this way,
AudioExplorer became a system that linked multiple representations of the sound input. Be-
cause the analysis tools (that showed the exact frequency) proved extraneous for this task,
two usage modes were integrated into the system; the novice mode simply hid the analysis
tools. This shows how the technique of extending a medium works on a small scale. In
novice mode, AudioExplorer is a medium to explore the differences between instruments.
In expert mode, AudioExplorer is a medium to explore the mathematical workings of har-
monic sound. By extending the novice AudioExplorer with analysis tools, it became a new
medium with a new message.
By this time, the learning goals were clearly established. Based on Roschelle’s work
on dyads using an inquiry tool together (Roschelle, 1996), the learning situation became
having two novices work together on an open-ended
laboratory assignment. Finally, an
appropriate learning situation for using the ELE was found: a “Physics of Music” class
at Georgia Tech. For formative evaluation, two groups used AudioExplorer; both groups
achieved a good understanding of the learning goals (Rick, 2002a). Based on this, several
interface changes were made to the AudioExplorer software to increase usability.
5.2 DigiQuilt: a Constructionist Medium
DigiQuilt is a computer-based construction kit for learning about math and art by designing
patchwork quilt blocks (Lamberty & Kolodner, 2002). Increasing in size, DigiQuilt is
made of pieces, patches, and blocks (which can be put together into quilts outside of the
system). Users create quilt blocks by selecting colored shapes (pieces) from a palette and
placing them into patches. The software offers learners 2-by-2, 3-by-3, or 4-by-4 patch
quilt-blocks with a variety of grids that can be imposed on them, facilities for saving and
loading designs, buttons for clearing the grid and stepping forward or backward through a
design, a palette with shapes and buttons to change their colors, and facilities for rotating
shapes or patches and copying or swapping patch-level designs so that they can be easily
Participants were asked to complete such open-ended assignments as comparing several instruments to
each other and writing down any five interesting observations that they made about the differences.
Rick & Lamberty Draft: Do Not Distribute
Figure 8: Using DigiQuilt to create a quilt block design (actual 3rd grader’s design)
repeated or changed (Figure 8). DigiQuilt has the learner move shapes to create the designs;
much like a physical manipulative. The designs can be printed or saved easily, and provide
a context for discussions of the targeted concepts among the learners.
Designing DigiQuilt came from a desire to highlight the interesting math found in the
world of patchwork quilt design. The designer, Lamberty, noticed that patchwork repre-
sented a potential medium to explore many mathematical ideas: fractions, symmetry, etc.
Quilting could be an interesting interdisciplinary link for math, art, and possibly history.
The choice to focus on math and art rather than history was strongly influenced by the fact
that the math and art connections lie within the medium and can be learned through design.
The history connections can be made in other ways by providing students with resource
materials including storybooks and books featuring antique quilts. Since elementary-aged
students generally enjoy doing art projects, Lamberty thought patchwork design would pro-
vide a steppingstone from art to math for most, and from math to art for some. Either way,
the hope was to increase the pool of interested students by combining two often disparate
Initially, the idea was explored using a paper version of the quilt-design manipulative
that was designed based on Lamberty’s experiences with other manipulatives to see what
sorts of designs could be created and how to structure the activities. Early field tests showed
that the kids had a tough time rotating the paper triangles the right amount to fit them
into the square spaces of the grid. Since students could place shapes that overlapped the
grid lines (or each other) in random ways, the resultant designs were difficult to talk about
mathematically. If the learning goals were different, that might not be a problem. But, since
random overlapping created designs that were not so easy to talk about in terms of fractions,
it had to be prevented. Thus, when Lamberty moved on to designing a software version of
the manipulatives, the shapes could be manipulated only in very specific ways. The shapes
Rick & Lamberty Draft: Do Not Distribute
can only be turned 90 degrees, and they snap in place in the grid.
This design decision highlights two important aspects of MBD. First, it demonstrates
the importance of strictures. In DigiQuilt, the learner is constrained to rotating pieces by
90 degrees. Instead of being a burden, this stricture makes the learning goals more salient.
Second, it demonstrates the power of the computer as a tool for realizing MBD. Compared
to the paper prototypes, restricting rotation to only 90 degrees was easy to implement in
software and feels much more natural to the learner. In MBD, the computer allows for the
easy creation of powerful strictures and structures that make the learning goals more salient.
Learners cannot place shapes that overlap grid lines using the software like they can in
the paper version. Note, though, that given a different audience or set of learning goals,
this might not have posed the same problem. The inspirational medium (patchwork) does
not necessarily need to change with the audience, but the tools and constraints do need to
change (since the learning goals do).
One feature of DigiQuilt was designed specifically to afford the learning of a listed stan-
dard for third graders: the ability to recognize the same shape in a different orientation. By
presenting each shape on the palette in only one orientation and requiring students to rotate
them to fit their needs and solve the challenge, the software affords learning to recognize
the same shape turned in different ways. With a physical manipulative, where shapes are
scattered on the floor or table, it is possible to select a shape without this recognition.
This design decision highlights another important aspect of designing learning environ-
ments: Design for learning is different than design for use. If the goal of DigiQuilt was quilt
creation (a user goal), having different orientations of the same shape is useful; it saves the
designer valuable time. But, the goal of DigiQuilt is to foster mathematical understanding
through quilt creation (a learning goal). As such, having only one orientation for the same
shape is useful; it makes the learner engage the learning goals.
For the first classroom trials of DigiQuilt, students used both the paper and computer
versions of the manipulatives. In the study, student use of the paper version of the manip-
ulative helped uncover several changes that might improve DigiQuilt and specify or extend
the learning goals it could address. Using the paper version, it was easy to create and test a
variety of methods for supporting students as they explored fractions and symmetry through
their designs. One tool that came from this exploration was the select-a-grid tool, that al-
lowed the overlaying of various grids on top of the block (Figure 8 shows a grid with three
vertical and three horizontal bars).
Challenges that involved equivalent fractions were difficult for students to understand.
For example, when approaching the challenge, “create a quilt block that is
blue,” students ran into several difficulties. Some students did not understand that the
three fractions added up to 1 and would result in the whole quilt block being filled. Some
students began by filling the quilt block with
yellow, and then filled
of the remaining
space with red. Since this didn’t leave 8 squares behind, some students simply put a blue
piece in the design. Lamberty thought that perhaps what was happening was that students
needed to understand that each fraction referred to the whole quilt block. In order to help
the students refocus their attention, she drew heavy lines on a paper 4-by-4 patch-grid: first
Rick & Lamberty Draft: Do Not Distribute
just one heavy line to emphasize
, then another line to emphasize
, and finally enough
lines to break the design into 8 equal parts so the students could finish the design with the
blue. The select-a-grid tool adds support for student learning by highlighting connections
between fractions in the challenges and the designs the students make. The main purpose
of this tool was to help learners refocus their attention on the “whole” as they work on
different aspects of the challenge, but the grids are also useful reference points for students
as they attempt to solve challenges that involve symmetry.
Since that initial classroom fieldwork, several changes to the software have been made
and tested in the classroom. Research on DigiQuilt focuses on studying student engagement
over time in order to understand more about what kinds of tools are best introduced at what
times, and describing the results of different attempts to get and keep students engaged.
Over time, changes made to the system tend to be adding or improving learning supports (in
response to user-testing) to highlight interesting connections between the medium (quilting)
and the targeted message (fractions), or adding new features that support the activity of
design by simplifying the process of making changes or adding more options to the system.
5.3 Product (ELE) Revisited
In the previous sections, we described AudioExplorer and DigiQuilt, two ELEs created
using MBD. In this section, we aim to show how these environments meet the characteristics
of ELEs. Thereby, we confirm that these environments are ELEs and provide concrete
examples of the characteristics of ELEs in context.
An ELE is empty of content. AudioExplorer acts as a simple inquiry tool, such as a
voltmeter or scale. Learners provide content, by playing on the keyboard, and then use
AudioExplorer to analyze it. DigiQuilt acts as a constructionist medium. It is up to the
learners to construct the content, by creating a digital quilt.
The interaction facilitated by ELEs is open-ended. In both AudioExplorer and Digi-
Quilt, the environment itself is open-ended (learners decide how they use the environment),
but, in use, the learner control is more guided (i.e. not unrestricted). In AudioExplorer use,
a laboratory assignment suggests what is worth investigating. In DigiQuilt use, learners are
given specific challenges, such as “cover
of the quilt block with one color and
another color. In both cases, the environment is open-ended and the challenges focus that
open-ended inquiry to better address learning goals. Both of their use is still open-ended,
as learners can meaningfully approach these challenges in different ways.
An ELE is structurally simplistic. AudioExplorer is simple; it focuses on one input—
musical sound coming from a keyboard. It harnesses the computational power of the com-
puter to transform that input into multiple representations. Each of those representations
allows learners to focus on a specific feature of the musical sound. DigiQuilt is simple; it
focuses on the construction of a quilt block with a limited amount of pieces. Those pieces
can only be placed into the quilt block in a fairly constrained way. The computational
power of the computer enables DigiQuilt to create structures and strictures that go beyond
the capabilities of a physical construction kit.
Rick & Lamberty Draft: Do Not Distribute
An ELE offers concrete connections to the important concepts. AudioExplorer extends
the piano, a medium that has concrete connections to western music. It extends the piano to
display the frequency domain, which has concrete connections to how we perceive sound.
The learner is able to connect the concrete (musical sound) to the concepts (structure of
harmonics). As such, they can explore how and why western music works (e.g. why does
a chord sound good?). DigiQuilt extends patchwork quilting to focus the creation process
on mathematical concepts (fractions, symmetry, etc.). When a learner adds to a quilt block,
the fraction of the quilt block that is covered by a color is displayed next to that color (see
Figure 8). So, the learner is able to connect the concrete (the fractions of the quilt block) to
the important concepts (mathematical fractions).
An effective ELE has personal connections. As the example at the beginning of this
article demonstrates, physics-of-music students are able to leverage their previous knowl-
edge of music to better understand what AudioExplorer is displaying. Music is important
to our culture; as such, there is an intrinsic interest in trying to understand it. AudioEx-
plorer leverages that interest to motivate learners. Art, such as patchwork quilting, also has
a strong cultural meaning. DigiQuilt leverages that interest to motivate learners. In addi-
tion, they are creating a personally meaningful artifact (a digital quilt), which can be highly
motivating (Papert, 1991).
An effective ELE has epistemological connections. AudioExplorer connects to the
harmonic properties of music. DigiQuilt connects to mathematical and artistic concepts.
For both of these environments, the concepts addressed are relevant to their audiences.
For physics-of-music students, understanding why music works is important. For fourth
graders, learning about fractions and symmetry is important.
5.4 Process (MBD) Revisited
The previous section was concerned with product—how AudioExplorer and DigiQuilt re-
flect the characteristics of ELEs. This section is concerned with process—how the design
of AudioExplorer and DigiQuilt reflect the phases of MBD. We address this to confirm that
these environments were created using MBD and to give concrete examples of the phases
of MBD in action.
Begin by choosing a medium you know well, care about, and feel has learning potential.
For both Rick and Lamberty, the media they started off with (harmonic sound and patch-
work quilting) were far from arbitrary. Both care deeply about their medium, showing an
interest and fascination with that medium long before the design process started. For a soft-
style method of design, it is advantageous to have an affinity (or closeness) to the object to
be manipulated. Through these ELEs, the designers were able to share their enthusiasm for
the medium with others.
Explore the medium to understand its affordances for promoting learning. If the strength
of the hard-style problem-based approach is knowledge of the problem space, the strength
of the MBD solution-based approach is knowledge of the solution space (what does the
medium afford). Both design methods involve reflective exploration of a search space and
are therefore not arbitrary. For AudioExplorer, early versions were demonstrated to many
Rick & Lamberty Draft: Do Not Distribute
knowledgeable users, including experts in the domain (audio, music and signal processing)
and educational technology. The system already addressed the original problem and, by the
help of others, the solution space was better illuminated. For DigiQuilt, physical manipu-
latives were used to explore the design space. Because of their flexibility, adaptability, and
familiarity, physical versions accelerated the exploration of the solution space and allowed
for the search to be highly iterative.
Extend the medium to change those affordances. It is not coincidental that both Digi-
Quilt and AudioExplorer have different user modes. The different modes accentuate (or
extend) different parts of the medium and thus address different learning goals. In each,
the tools that were built to manipulate the medium addressed specific learning goals. In
AudioExplorer, the analysis tools allowed learners to move from investigating instruments
to investigating the mathematical properties of harmonics. In DigiQuilt, the rotate and copy
tools helped children meet listed standards.
Transition from the solution (the medium) to the problem (a clear message). AudioEx-
plorer design began by investigating the affordances of the frequency domain. AudioEx-
plorer evolved into a medium to understand how and why music worked. Finally, an ap-
propriate audience was found. The solution (AudioExplorer) addresses an important prob-
lem (how and why music works) for that audience (physics-of-music students). DigiQuilt
started off investigating the affordances of patchwork quilting. It evolved into a medium
to connect mathematical concepts to art construction. Finally, an appropriate audience was
found. The solution (DigiQuilt) addresses important problems (fraction understanding) for
that audience (fourth graders).
Though neither design started off with an intended audience or clear problem that it
was trying to address, both ended up addressing important problems for a real audience of
learners. Though neither design started off from specific user needs, both design processes
were user-based and met both the usability and learning goal needs of their audience.
6 Assessing Medium-Based Design
In the previous section, we detailed two learning environments, AudioExplorer and Digi-
Quilt, created using MBD. We hope to thereby provide a rich description of how and why
MBD allows for the creation of ELEs. Because these environments were designed by us,
the authors, the conclusions are limited. At best, we can conclude that MBD can be effec-
tively used by its originators to create ELEs. To be truly effective, a design method needs
to be useful to others as well. It also needs to be better in some way to alternative methods;
otherwise, there is no advantage to using it.
In this section, we address these issues empirically. We seek to show that others can use
MBD in the creation of ELEs. In particular, we look to confirm our hypothesis that MBD
is more suited towards the creation of ELEs than a hard-style design (HSD). This led us to
focus on how the products of MBD and HSD differ. If the hypothesis held, the products
of MBD could better be characterized as ELEs. To address this topic, we conducted a
comparative study of MBD and HSD. In this section, we detail this study. First, we describe
Rick & Lamberty Draft: Do Not Distribute
how the study was conducted. Then, we examine the results to test our hypothesis. Finally,
we address the challenges of studying design methods and confront some of the limitations
of our study.
6.1 The Study Setup
To test our hypothesis we carried out a study in a computer science class on educational
technology. In that class, students had to work in groups on a large (both in terms of time
commitment and point value) design project. The class was split up into design groups
(2-3 members per group). Six of the groups were given a MBD guideline and the other
six were given a HSD guideline to follow. These guidelines scaffolded each phase of the
design process. The groups went through four phases of the design. The phases in the
MBD guideline roughly approximated the phases of MBD: 1) pick an interesting medium;
2) build the prototype to explore the problem space; 3) trial the prototype in user testing;
4) get feedback from classmates and finish the design. The phases in the HSD guideline
roughly correspond to: 1) pick an important problem; 2) explore the problem space through
user interviews; 3) prototype the design; 4) get feedback from classmates and finish the
design. Due to the limitations of the class, none of the groups actually fully realized their
design, though all had to at least create a prototype. As we are trying to determine the
effectiveness of the design method, this may actually be beneficial: These prototypes may
be more representative of the design methods (MBD and HSD) than a finished learning
environment, as they come early in the process when the design is still largely governed by
the design method and less by the particulars of the learning context.
After their projects were completed, students were given a survey so we could better
understand their design process and products. The advantage of having students reflect
on their own designs are twofold. First, students are not invested in MBD and are thus
unbiased towards certain conclusions. Second, the students were the designers, so they
have a familiarity with their design (particularly, their process) that an outsider could not
have. To analyze the data, the responses for each group on a 1-5 Likert scale were averaged
and compared across the study condition. Two groups, one from each condition, were
removed from the analysis; these groups failed to follow their respective design guidelines
closely, so both their process and product were not representative of either design process.
This analysis allows us to explore our hypothesis that MBD is more successful than HSD
in the creation of ELEs. The results suggest our hypothesis has merit.
6.2 Results
One of the defining properties of an ELE is that the teacher transitions from being the “sage
on the stage” to being the “guide on the side. In other words, the teacher becomes a guide
to engaging the learning environment, rather than the transmitter of information. The survey
data suggests that this transition occurs more in MBD groups (see Table 1).
Both MBD and
Because of the small sample size (five versus five), statistical testing does not make sense for this analysis.
Rather, this is a suggestive analysis based on trends in the descriptive statistics.
Rick & Lamberty Draft: Do Not Distribute
Question MBD HSD
The teacher plays a large / important role in my group’s design.
3.73 3.73
In my group’s design, learners learned from (either human or elec-
tronic) experts or teachers.
2.83 4.37
Table 1: Sage on the Stage vs. Guide on the Side, where 1 is strongly disagree and 5 is
strongly agree
Question MBD HSD
My group’s design closely matches cognitive apprenticeship.
2.83 3.40
My group’s design closely matches inquiry-based learning, such as
Learning by Design and Problem-based Learning.
3.47 3.23
My group’s design closely matches constructionism.
3.80 3.20
Compared to other groups, my group’s design is more constructivist
(the learners construct their own understanding, as opposed to memo-
rizing material).
3.83 3.43
Table 2: Alignment with Learning Theories, where 1 is strongly disagree and 5 is strongly
HSD groups agreed that the teacher plays an important role in their design. Yet, when asked
whether learners learned from experts or teachers, the two groups diverged. HSD groups
felt more strongly that learners learned from experts or teachers in their designs. For MBD
groups, the teacher was still considered an important part of a learning environment, but
the learners were not learning as much directly from the teacher. This suggests that the
teacher’s role is more like a “guide on the side.
The survey also suggests that the designs created with MBD align better with learning
approaches that are in-line with ELEs, such as constructionism. This prediction proved true
to various degrees (see Table 2). Constructionism, inquiry-based learning and cognitive
apprenticeship were covered extensively in class, so students were aware of what alignment
with those theories meant.
For MBD groups, the most alignment was with constructionism
and the least alignment was with cognitive apprenticeship; for HSD groups, the alignment
results were reversed. So, the results suggest that MBD may be a comparatively useful
method for designing constructionist environments. The survey also indicates that MBD
groups considered their systems to be more constructivist (Table 2).
Finally, how learners learned was compared. We were interested in how students are
In the actual survey, intelligent tutoring systems were included as a fourth category for students to align
with. Since ITSs were only briefly mention in class, it is likely students had a hard time responding to that
question accurately. As such, it was excluded from this analysis.
Rick & Lamberty Draft: Do Not Distribute
Category Difference
. . . from (either human or electronic) experts or teachers.
. . . from their peers / fellow learners.
. . . from external materials, such as reading.
. . . by playing around / trying things out.
Table 3: Learners Learned. . . , based on a 1 (strongly disagree) and 5 (strongly agree) scale
learning in these environments. Are they learning from experts or teachers? Are they
learning from their peers? Are they learning from external materials? Are they learning
from playing around (exploring)? In particular, we are interested in how answers to these
questions compared to each other within groups. To get at this, we normalized the data
around the average answer for each group. In that way, a group was treated the same
whether they were consistently in agreement that learners learned from these sources or
consistently in disagreement that learners learned from these sources. What counted was
how one answer compared to another answer within that group. The prediction was that
MBD designs would get more of their learning potential from playing around and less from
experts or teachers. Again, this prediction proved true to various degrees (see Table 3).
HSD groups were more confident that their learners learned from experts and teachers.
In contrast, MBD groups were more confident that their learners learned from external
materials or by playing around.
6.3 Challenges and Limitations
While we are encouraged by these results, there are a number of limitations to our findings
due to the difficulty of studying design methods. Comparing design methods is always chal-
lenging. Comparing design methods in the complex field of learning environment design is
even more difficult for a number of reasons:
1. Designers must have an adequate background. To engage in the reflection-in-action
process (Schön, 1987), designers will have to be both skilled in education and com-
puter science. Only someone familiar with learning theory can reflect properly on
his or her design. Only someone capable of programming can properly act on that
reflection. There are few who meet both requirements. As such, there is a small pool
from which to recruit designers to take part in such a study.
2. Designing an effective learning environment involves substantial time and effort. At
a minimum, designers must plan their design, execute it, and refine it based on use.
Often, further design steps are necessary. Finally, to demonstrate its effectiveness,
the learning value of the design must be evaluated; this alone is often a challenging
Rick & Lamberty Draft: Do Not Distribute
3. Many factors influence the success of a design project. Some designers are more
experienced or gifted than others. Some ideas simply bear better fruit. Some contexts
of use are more congenial to learning outcomes than others. The number of factors
involved make it difficult to correlate any success (or failure) to the choice of a design
4. Different learning environments are difficult to compare. How do you compare learn-
ing environments that address different learning goals? How does a ten percent pre-
test / post-test gain in factoring Algebraic equations compare to a better understand-
ing of how an ant colony secures food? Unless the learning environments address the
same learning goals they are impossible to compare directly. Even when they address
the same learning goals, environments are difficult to compare.
These reasons make it difficult to study this domain. Given our limited means and the dif-
ficulty of the task, we face several challenges to the validity of our study. First, because
it is difficult to recruit designers, we only have ten (five versus ve) data points. Conse-
quently, there is no way to achieve statistical significance, a common criteria for evaluating
the validity of a study. However, we are encouraged to find that the results support several
of our hypotheses. Second, due to time and effort considerations, the designs were carried
out by novices. It is questionable whether their experiences reflect those of expert design-
ers. Third, all designs were evaluated by novices. Whether their judgment is accurate is
questionable. For instance, survey data on personal and epistemological connections had to
be thrown out, because it was clear from the final exam that many students did not properly
understand these concepts. Fourth, the designers did not fully realize their designs; they
only created paper designs or (at best) prototypes. It is questionable whether these limited
designs reflect the ultimate product of the design process. Fifth, the results are situated
in our context of study (Georgia Tech students, Rick as the instructor of the course, the
specific implementations of MBD and HSD, etc.). As within any such situated study, it is
questionable whether the results transfer to a different situation.
We hope that it is clear from this discussion that comparing design methods in this field
is challenging. Inevitably, it involves limitations. Even if we had outrageous means and
could study many expert designers realizing full designs, we would still face problems.
First, we would have to account for differences in experience and effort—something that
is largely controlled for in our sample of novices. Second, it would be difficult to attribute
success or failure to the design method rather than the particulars of the learning context.
Third, it would still be difficult to impossible to directly compare the effectiveness of the
different environments.
Comparing design methods for creating learning environments is inherently challeng-
ing. Yet, we should not ignore an area of study just because we are unable to obtain the
kind of evidence standard in other domains (Yin, 2003). Otherwise, we are like the man
who prefers to search for his keys under the streetlight where visibility is highest, rather
than the dark alley where he lost them. We will never find the key. In this study, we admit
to searching in the dark alley, where evidence is hard to come by. Yet, we are encouraged
Rick & Lamberty Draft: Do Not Distribute
by our results.
7 Discussion and Conclusions
In this article, we introduce a new approach for creating learning environments, MBD. We
detailed both the goal of creating ELEs and the MBD approach for reaching that goal. Our
hope is that the reader will have a better understanding of ELEs and why MBD is a useful
method for creating ELEs. While we value ELEs and exploratory learning in general, we
do not argue that all learning environments should be ELEs; there are many useful ones
which are not. We also do not argue that MBD is the only approach that enables designers
to create ELEs; designers have created ELEs before we articulated MBD. Yet, we do feel
that MBD is a particularly useful method for creating ELEs. At a minimum, it has worked
for us; furthermore, our study suggests it can work for others.
7.1 MBD and ELEs: Soft-Style Design, Soft-Style Learning
We believe that in articulating and grounding a soft style of design, we can provide an
alternative to designers. As soft-style designers, we have often had our design process
maligned: “Your process is arbitrary and that your design product is useful is largely despite
the process. With this work, we aim to combat that attitude. Turkle and Papert (1991)
revalued the concrete by showing that a soft style was an acceptable, useful learning style.
Here, we revalue the concrete by showing that a soft style of learning environment design
too can be grounded in theory, have common properties (i.e. the process is not arbitrary),
and be successful because of the process. It is probably not incidental that our soft-style
design method (MBD) is useful in creating environments (ELEs) that allow for a soft-style
learning process (learning by exploring concrete environments).
7.2 Limits: Who should use MBD?
MBD is by no means a panacea for the problems of learning environment design. In an
early presentation of MBD, one colleague pointed out that MBD takes time and creativity
in addition to being counter to a suggestion often offered teachers: “Start with learning
goals and develop activities that help students achieve those goals. We agree with that
assessment. While MBD may be a great method for us, we realize it may not be appropriate
for others.
One reason MBD works for us is that we, as computer scientists, are able to harness
the power of the computer to create a new medium. Practically, this requires programming.
Programming applications that learners can use is a challenging, time-consuming task. It
is doubtful that an average teacher has either the time or the programming ability to use
MBD. Even if a teacher had the time and ability to use MBD, they face further challenges.
Teachers are responsible for helping their students meet particular learning goals. In MBD,
the problem and thereby audience addressed by the product of the design process is not
Rick & Lamberty Draft: Do Not Distribute
known from the beginning of the design process. So, even if our “überteacher” created an
ELE using MBD, in the end it might not be useful for his or her students. As such, we do
not feel that MBD is an appropriate design method for most teachers.
While it is probably impractical for teachers to design using MBD, teachers are crucial
to the use of ELEs in the classroom and can play a useful role in MBD. A good teacher can
adapt an ELE to suit the learning needs of their students. As Papert recalls of his experience
with Logo, “I have been extraordinarily gratified to see teachers using Logo as a painter uses
paint, that is, as a medium for creative work, in this case for the creation or enhancement of
learning environments” (Papert, 1993b, p. xiv). A good MBD designer often needs to work
closely with a teacher (particularly in phase 4); the teacher is in a good position to evaluate
the usefulness of the ELE for his or her students for meeting their specific learning goals.
In our experience, in learning environment design, computer programmers are often rel-
egated to simply implementing the software ideas of others. That approach is incompatible
with our notion of MBD. In MBD, programmers must play a primary role in creating ELEs.
Those programmers must have a fairly good understanding of learning theory and be able
to work with other educators (curriculum designers, teachers, etc.) to realize their designs.
This is far from a trivial task. We are fortunate in that we had the flexibility and enough
expertise to carry out MBD. We hope that by pointing out the benefits of our approach that
others may come to value those opportunities and make them available to others.
Bolter, J. D. (2001). Writing space: Computers, hypertext, and the remediation of print
(Second ed.). Mahwah, New Jersey: Lawrence Erlbaum Associates.
Bolter, J. D., & Grusin, R. (1999). Remediation: Understanding new media. Cambridge,
MA: MIT Press.
Bruckman, A. (1998). Community support for constructionist learning. Computer-
Supported Cooperative Work, 7, 47-86.
Bruner, J. S. (1966). Towards a theory of instruction. Cambridge, MA: The Belknap Press
of Harvard University Press.
Carroll, J. (2000). Five reasons for scenario-based design. Interacting with Computers, 13,
Dewey, J. (1938). Experience & education. New York, NY: Simon & Schuster.
diSessa, A. A. (1985). A principled design for an integrated computational environment.
Human-Computer Interaction, 1(1), 1-47.
diSessa, A. A. (1987). Artificial worlds and real experience. In R. Lawler & M. Yazdani
(Eds.), Artificial intelligence and education, volume one: Learning environments &
tutoring systems (p. 55-77). Ablex Publishing Corporation.
diSessa, A. A. (2000). Changing minds: Computers, learning, and literacy. Cambridge,
MA: The MIT Press.
Goel, V., & Pirolli, P. (1992). The structure of design problem spaces. Cognitive Science,
16(3), 395-429.
Rick & Lamberty Draft: Do Not Distribute
Guzdial, M., Rick, J., & Kehoe, C. (2001). Beyond adoption to invention: Teacher-created
collaborative activities in higher education. The Journal of the Learning Sciences,
10(3), 265-279.
Guzdial, M., & Rose, K. (Eds.). (2002). Squeak: Open personal computing and multimedia.
Upper Saddle River, NJ: Prentice Hall.
Hay, K., & Barab, S. (2001). Constructivism in practice: A comparison and contrast of ap-
prenticeship and constructionist learning environments. The Journal of the Learning
Sciences, 10(3), 281-322.
Joseph, D., & Nacu, D. C. (2003). Designing interesting learning environments when the
medium isn’t enough. Convergence, 9(2), 84-115.
Kaput, J. (1989). Linking representations in the symbolic systems of algebra. In S. Wagner
& C. Kieran (Eds.), Research agenda for mathematics education: Research issues in
the learning and teaching of algebra (p. 167-194). Reston, VA: National Council of
Teachers of Mathematics.
Kay, A., & Goldberg, A. (1977). Personal dynamic media. IEEE Computer, March 1977,
Kolodner, J. (1997). Educational implications of analogy: A view from case-based reason-
ing. American Psychologist, 52(1), 57-66.
Lamberty, K. K., & Kolodner, J. L. (2002). Exploring digital quilt design using manipula-
tives as a math learning tool. In P. Bell, R. Stevens, & T. Satwicz (Eds.), Proceedings
of ICLS 2002 (p. 552-553). Mahwah, NJ: Lawrence Erlbaum Associates.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation.
Cambridge, UK: Cambridge University Press.
McLuhan, H. M. (1964). Understanding media: The extensions of man. Cambridge, MA:
The MIT Press.
McLuhan, H. M., & Fiore, Q. (1967). The medium is the massage: An inventory of effects.
Corte Madera, CA: Ginko Press.
Norman, D. A. (1988). The psychology of everyday things. New York, NY: Basic Books.
Papert, S. (1987). Microworlds: Transforming education. In R. Lawler & M. Yazdani
(Eds.), Artificial intelligence and education, volume one: Learning environments &
tutoring systems (p. 79-94). Ablex Publishing Corporation.
Papert, S. (1991). Situating constructionism. In I. Harel & S. Papert (Eds.), Constructionism
(p. 1-13). Ablex.
Papert, S. (1993a). The children’s machine: Rethinking school in the age of the computer.
New York, NY: Basic Books.
Papert, S. (1993b). Mindstorms: Children, computers and powerful ideas (Second ed.).
New York, NY: Basic Books.
Quintana, C., Reiser, B. J., Davis, E. A., Krajcik, J., Fretz, E., Duncan, R. G., et al. (2004).
A scaffolding design framework for software to support science inquiry. The Journal
of the Learning Sciences, 13(3), 337-386.
Reeves, T. (1994). Evaluating what really matters in computer-based education. In M. Wild
& D. Kirkpatrick (Eds.), Computer education: New perspectives (p. 219-246). Perth,
Rick & Lamberty Draft: Do Not Distribute
Australia: Edith Cowan University Press.
Resnick, M. (1994). Turtles, termites, and traffic jams: Explorations in massively parallel
microworlds. Cambridge, MA: MIT Press.
Resnick, M. (1996). Beyond the centralized mindset. The Journal of the Learning Sciences,
5(1), 1-22.
Resnick, M. (1998). Technologies for lifelong kindergarten. Educational Technology
Research & Development, 46(4), 43-55.
Resnick, M., Bruckman, A., & Martin, F. (1996). Pianos not stereos: Creating computa-
tional construction kits. Interactions, 3(6), 41-50.
Rick, J. (2002a). AudioExplorer: Multiple linked representations for convergence. In
G. Stahl (Ed.), Online proceedings of CSCL 2002.
Rick, J. (2002b). Pianos, not orchestras. In G. Stahl (Ed.), Proceedings of CSCL 2002 (p.
635-636). Hillsdale, NJ: Lawrence Erlbaum Associates.
Roschelle, J. (1996). Learning by collaborating: Convergent conceptual change. In
T. Koschman (Ed.), CSCL: Theory and practice of an emerging paradigm (p. 209-
248). Mahwah, NJ: Lawrence Erlbaum Associates.
Scardamalia, M., & Bereiter, C. (1991). Higher levels of agency for children in knowledge
building: A challenge for the design of new knowlege media. The Journal of the
Learning Sciences, 1(1), 37-68.
Schank, R., Fano, A., Bell, B., & Jona, M. (1994). The design of goal-based scenarios. The
Journal of the Learning Sciences, 3(4), 305-345.
Schön, D. A. (1987). Educating the reflective practitioner. San Francisco, CA: Jossey-Bass
Soloway, E., Guzdial, M., & Hay, K. E. (1994, April). Learner-centered design: The
challenge for HCI in the 21st century. Interactions, 1(2), 36-48.
Standing, E. M. (1957). Maria Montessori, her life and work. New York, NY: A Plume
Suthers, D. D., & Hundhausen, C. D. (2003). An experimental study of the effects of
representational guidance on collaborative learning processes. The Journal of the
Learning Sciences, 12(2), 183-219.
Terry, M., Brostow, G. J., Ou, G., Tyman, J., & Gromala, D. (2004). Making space for time
in time-lapse photography. In SIGGRAPH 2004 sketches program.
Turkle, S., & Papert, S. (1991). Epistemological pluralism and the revaluation of the con-
crete. In I. Harel & S. Papert (Eds.), Constructionism (p. 161-191). Ablex Publishing
Wilensky, U. (1991). Abstract meditations on the concrete and concrete implications for
mathematics education. In I. Harel & S. Papert (Eds.), Constructionism (p. 193-204).
Norwood, NJ: Ablex Publishing Corp.
Yin, R. K. (2003). Case study research: Design and methods (third ed.). Thousand Oaks,
CA: SAGE Publications.
Zucchermaglio, C. (1993). Toward a cognititve ergonomics of educational technology. In
T. Duffy, J. Lowyck, & D. Jonassen (Eds.), Designing environments for constructivist
Rick & Lamberty Draft: Do Not Distribute
learning (p. 249-260). Heidelberg, Germany: Springer-Verlag.
... Notwithstanding the importance of this focus, business schools and educators are facing complex teaching and learning challenges in order to deliver these subjects in a way that matters most for their audience, in this case students, who represent future practitioners (Dacre et al., 2019;Ojiako et al., 2011). As such, by bringing together constructivist learning theories like exploratory learning environments (Duckworth, 2006;Rick & Lamberty, 2005), organisational sustainability teaching frameworks (Stubbs & Cocklin, 2008) and empirical research on the use of the Lego Serious Play (LSP) methods in educational settings (James, 2013;Kurkovsky, 2015;Mccusker, 2014) This paper aims to conceptualise LSP as an innovative teaching method that utilises Lego bricks to improve student engagement and participation, create exploratory teaching environments that can support learning, and help shape responsible organisational leaders. Therefore, the focus of this paper offers a departure from prior research focused on what constitutes responsible management education content to provide a framework on how to engage in the teaching and learning process of this content. ...
... With LSP having successfully been utilised in different knowledge domains such as computer science (Kurkovsky, 2015) and management education (Grienitz & Schmidt, 2012) this paper suggests that LSP can enable educators to embed the values of constructivist learning theories into their teaching practices and operationalise exploratory learning environments to enhance student engagement and participation. Rick & Lamberty (2005) define exploratory learning environments as educational arrangements and activities that facilitate the learners' ability to construct knowledge connected to the subject matter through student-led reflective exploration. Exploratory learning activities can promote an increased democratic style of education and, according to Duckworth (2006, p. 67) can guide students towards: ...
Full-text available
This paper conceptualises a Lego Serious Play Wheel framework as a gamification teaching and learning method. It aims to offer a detailed approach from Design and Preparation to Delivery, to engage a broad section of continuing learners and students, which can be easily applied throughout different educational and training contexts. The LSP Wheel refers to the concept of a circular learning journey and draws on a combined autoethnography responsible management research approach. A prominent part of the responsible management literature has hitherto focused on examining whether responsible management modules are inherently considered non-crucial elements of curriculum design. However, there is a paucity of research into applying novel teaching approaches to engage students and promote responsible management education endeavours. This paper therefore contributes to broader pedagogical application and critical responsible management education discourse, by providing educators with an academic gamification framework to support student engagement and co-creation of knowledge, by fostering exploratory learning environments and enriching the practices of active learning communities.
... As such, by bringing together constructivist learning theories like exploratory learning environments (Duckworth, 2006;Rick & Lamberty, 2005), organisational sustainability teaching frameworks (Stubbs & Cocklin, 2008) and empirical research on the use of the Lego Serious Play (LSP) methods in educational settings (James, 2013;Kurkovsky, 2015;Mccusker, 2014), this research will aim to conceptualise LSP as an innovative teaching method that utilises Lego bricks to improve student engagement and participation, create exploratory teaching environments that can support learning, and help shape responsible organisational leaders. ...
Full-text available
Research into responsible management education has largely focused on the merits, attributes, and transformation opportunities to enhance responsible business school education aims. As such, a prominent part of the literature has occupied itself with examining if responsible management modules are inherently considered a non-crucial element of the curriculum and determining the extent to which business schools have introduced such learning content into their curriculum. However, there has been scant research into how to apply novel teaching approaches to engage students and promote responsible management education endeavours. As such, this paper seeks to address this gap through the development of a teaching framework to support educators in designing effective learning environments focused on responsible management education. We will draw on constructivist learning theories and Lego Serious Play (LSP) as a learning enhancement approach to develop a pedagogical framework. LSP is selected due to its increasing application in learning environments to help promote critical discourse, and engage with highly complex problems, whether these are social, economic, environmental, or organisational.
... As such, by bringing together constructivist learning theories like exploratory learning environments (Duckworth, 2006;Rick & Lamberty, 2005), organisational sustainability teaching frameworks (Stubbs & Cocklin, 2008) and empirical research on the use of the Lego Serious Play (LSP) methods in educational settings (James, 2013;Kurkovsky, 2015;Mccusker, 2014), this research will aim to conceptualise LSP as an innovative teaching method that utilises Lego bricks to improve student engagement and participation, create exploratory teaching environments that can support learning, and help shape responsible organisational leaders. ...
Full-text available
Research into responsible management education has largely focused on the merits, attributes, and transformation opportunities to enhance responsible business school education aims. As such, a prominent part of the literature has occupied itself with examining if responsible management modules are inherently considered a non-crucial element of the curriculum and determining the extent to which business schools have introduced such learning content into their curriculum. However, there has been scant research into how to apply novel teaching approaches to engage students and promote responsible management education endeavours. As such, this paper seeks to address this gap through the development of a teaching framework to support educators in designing effective learning environments focused on responsible management education. We will draw on constructivist learning theories and Lego Serious Play (LSP) as a learning enhancement approach to develop a pedagogical framework. LSP is selected due to its increasing application in learning environments to help promote critical discourse, and engage with highly complex problems, whether these are social, economic, environmental, or organisational.
... As such, by bringing together constructivist learning theories like exploratory learning environments (Duckworth, 2006;Rick & Lamberty, 2005), organisational sustainability teaching frameworks (Stubbs & Cocklin, 2008) and empirical research on the use of the Lego Serious Play (LSP) methods in educational settings (James, 2013;Kurkovsky, 2015;Mccusker, 2014), this research will aim to conceptualise LSP as an innovative teaching method that utilises Lego bricks to improve student engagement and participation, create exploratory teaching environments that can support learning, and help shape responsible organisational leaders. ...
Research into responsible management education has largely focused on the merits, attributes, and transformation opportunities to enhance responsible business school education aims. As such, a prominent part of the literature has occupied itself with examining if responsible management modules are inherently considered a non-crucial element of the curriculum and determining the extent to which business schools have introduced such learning content into their curriculum. However, there has been scant research into how to apply novel teaching approaches to engage students and promote responsible management education endeavours. As such, this paper seeks to address this gap through the development of a teaching framework to support educators in designing effective learning environments focused on responsible management education. We will draw on constructivist learning theories and Lego Serious Play (LSP) as a learning enhancement approach to develop a pedagogical framework. LSP is selected due to its increasing application in learning environments to help promote critical discourse, and engage with highly complex problems, whether these are social, economic, environmental, or organisational.
... In general, the design process is considered to be a journey which is embarked upon by the designer (Rick and Lamberty 2005). This journey can be explored in various forms, including architecture, sculpture, painting, fashion, textiles, theatre, literature etc. ...
Full-text available
By the late twentieth century, the use of digital media for the design process had become epidemic amongst the designer communities. This paper aims to explore the place of human creativity. In this way, the present study is an attempt to develop a new medium as a tool for the architectural design process by using Found-Object Art. To identify the opportunities of this medium, an experimental study was carried out in some notable universities in Iran. Findings demonstrate that the metaphor as a meaning-making tool for idea generation and Found-Object Art as a physical tool can provide a creative tool to present the idea in a real form. Findings in the study show that the students tried to think in a different way about the objects surrounding them which had previously been ignored. Through analysis of the design projects, it can be said that this medium is a motivating factor for innovative design.
... Редица автори обаче отбелязват, че използването на "централизирана система" (има се предвид LCMS-система, която се използва от институция и поддържа множество учебни курсове) поражда негативен ефект върху психологията на студентите и ги кара да участват по-плахо в провежданите дискусии. Такова мнение, под една или друга форма, е подкрепено в (Dalsgaard, 2006), (Rick, et al., 2005) (Kim, 2008), от гледна точка на обучение чрез блог (но доближено към вид на форум, защото в разисквания модел самите обучаеми са автори на основните статии в блоговете), на базата на множество разгледани литературни източници е направена сравнителна характеристика на предимствата и недостатъците на LCMS системите (в частност с фокус върху BlackBoard) спрямо употребата на обучение чрез блог. Авторът посочва два литературни източника, в които са проведени експерименти със сходни начални условия (31 на брой студенти, използване на блог като допълнение към традиционна LCMS система, незадължително участие в субект-субектната комуникация в блога), но в които се отчитат коренно противоположни резултати -за една учебна година е имало над 700 публикувани съобщения в блог системата на (Lin, et al., 2006) и само 9 в блог системата на (Divitini, et al., 2005). ...
Full-text available
Research into responsible management education has largely focused on the merits, attributes, and transformation opportunities to enhance responsible business school education aims. As such, a prominent part of the literature has occupied itself with examining if responsible management modules are inherently considered a non-crucial element of the curriculum and determining the extent to which business schools have introduced such learning content into their curriculum. However, there has been scant research into how to apply novel teaching approaches to engage students and promote responsible management education endeavours. As such, this paper seeks to address this gap through the development of a teaching framework to support educators in designing effective learning environments focused on responsible management education. We draw on constructivist learning theories and Lego Serious Play (LSP) as a learning enhancement approach to develop a pedagogical framework titled The Educator’s LSP Journey. LSP is chosen due to its increasing application in learning environments to help promote critical discourse, and engage with highly complex problems, whether these are social, economic, environmental, or organisational. Therefore, this paper contributes to the responsible management education discourse by providing educators with a practical methodology to support student engagement and co-creation of knowledge by fostering exploratory learning environments and enriching the practices of active learning communities.
Full-text available
The constructivist perspectives outlined in this chapter contribute important insights about knowing, learning, and instruction as well as epistemological and theoretical foundations for designing principles-based constructivist learning environments. Findings from learning science research are synthesized and aligned discussing various ways of constructivist thinking: cognitive constructivism, social constructivism, and situativity theory, including selected learning and instruction models with relevance to teacher education. In addition, current critiques of and misconceptions about constructivist perspectives are presented. This chapter also derives common design principles of student-centered learning environments, drawing on findings from several established design frameworks that are based on a situative constructivist view of learning and instruction.
To develop students’ critical thinking abilities, teachers must lead students to engage in discussions and to reason within various points of view, while employing evidence to draw conclusions, make decisions or seek solutions. Computer mediates learning by providing students with visualizations of relevant subject content to facilitate their reflection on these experiences. The purpose of this study is to investigate how technology can enable students to develop critical thinking through technology-enhanced interactivity. A geography teacher and 62 grade seven students from two classes participated in the experiment. One class was taught using traditional methods whereas the instructional strategy of computer-mediated learning was adopted in the other. The findings showed that computer media designed for enhancing interactivity would facilitate teacher-pupil interaction and peer discussion, and consequently contributed to improve students’ comprehension of the geography curriculum content. In addition, the critical thinking ability of the high-achievement students was significantly improved. Key-words: computer-mediated learning; computer-assisted learning; critical thinking; geography education; interactivity.
This thesis uses a systematic understanding of sustainability informed by human needs, learning and design theory to explore ways in which small retail environments can effectively communicate sustainability concepts. The envisioned outcome of successfully communicating and implementing sustainability within retail environments is a lasting change in people’s daily behaviors. The methods of literature review, surveys, human needs investigation and professional validation are used to develop a behavioral change model centered on human needs and learning as well as six communication guidelines. The appendix of this thesis contains a userfriendly pocket guidebook titled The Six Guidelines for Sustainable Retail. The guidebook is designed as a quick-reference tool for retailers, designers and employees. It contains principles, visuals and concepts of sustainability for daily communication and comprehension purposes.
The unimportant and often negative results of the introduction of technology into various educational settings are mainly due to scarce consideration given to a realistic and cognitively based epistemology of learning and to an analysis of educative settings in which technology should be inserted and used. I will explain these points briefly.
The notion of scaffolding learners to help them succeed in solving problems otherwise too difficult for them is an important idea that has extended into the design of scaffolded software tools for learners. However, although there is a growing body of work on scaffolded tools, scaffold design, and the impact of scaffolding, the field has not yet converged on a common theoretical framework that defines rationales and approaches to guide the design of scaffolded tools. In this article, we present a scaffolding design framework addressing scaffolded software tools for science inquiry. Developed through iterative cycles of inductive and theory-based analysis, the framework synthesizes the work of prior design efforts, theoretical arguments, and empirical work in a set of guidelines that are organized around science inquiry practices and the challenges learners face in those practices. The framework can provide a basis for developing a theory of pedagogical support and a mechanism to describe successful scaffolding approaches. It can also guide design, not in a prescriptive manner but by providing designers with heuristics and examples of possible ways to address the challenges learners face.
Educators broadly agree that interest plays an important role in learning. In our work, we develop learning environments that align learner interest and important adult-defined learning objectives. Through this work we have come to recognise the complexity of the enterprise of this kind of learning environment design.1 At this stage, we have a relatively stable design model in the passion curriculum design approach.2 Missing, however, is a basis for analysing the interests and engagement of individual learners as they interact with a learning environment over time. This paper describes the theoretical and design frameworks we use, and recounts our most recent curriculum implementation, Multimedia Studio, and how it exposed this critical gap in the design model. We found that designing for learner interest is an even more complex undertaking than we originally understood. The lessons learned demonstrate the challenges of interest-centred approaches to curriculum design and can inform the work of other learning environment designers and researchers working in similar contexts.
Case-based reasoning (CBR) focuses on analogy in the context of solving real-world problems. Its research methodology of computational modeling is aimed at deriving hypotheses about cognition. CBR's computational models show the roles of encoding, retrieval, and adaptation in analogical reasoning processes. In addition, its algorithms provide insight into what it might take to enhance human cognition. CBR as a plausible cognitive model can thus advise on educational philosophy, educational practice, and design of educational software. (PsycINFO Database Record (c) 2012 APA, all rights reserved)