Learning by Doing with Shareable Interfaces

Abstract

New technologies, such as multi-touch tables, increasingly provide shareable interfaces where multiple people can simultaneously interact, enabling co-located groups to collaborate more flexibly than using single personal computers. Soon, these technologies will make their way into the classroom. However, little is known about what kinds of learning activities they will effectively support that other technologies, such as mobile devices, whiteboards, and personal computers, are currently unable to do. We suggest that one of the most promising uses of shareable interfaces is to support learning through exploration and creation. We present our work on DigiTile as a case study of how shareable interfaces can enable these forms of learning by doing. We demonstrate how DigiTile supports collaboration, present a field study on its learning benefits, and show how it can fit into a larger computing ecology.

Full text

Open Research Online
The Open University’s repository of research publications
and other research outputs
Learning by doing with shareable interfaces
Journal Article
How to cite:
Rick, Jochen; Rogers, Yvonne; Haig, Caroline and Yuill, Nicola (2009). Learning by doing with share-
able interfaces. Children, Youth and Environments, 19(1), pp. 321–342.
For guidance on citations see FAQs.
c
2009 Children, Youth and Environments
Version: Version of Record
Link(s) to article on publisher’s website:
http://thunder1.cudenver.edu/cye journal/abstract.pl?n=1900
Copyright and Moral Rights for the articles on this site are retained by the individual authors and/or other copy-
right owners. For more information on Open Research Online’s data policy on reuse of materials please consult
the policies page.
oro.open.ac.uk

Children, Youth and Environments 19(1), 2009
Learning by Doing with Shareable Interfaces
Jochen Rick
Yvonne Rogers
Department of Computing, The Open University
Milton Keynes, UK
Caroline Haig
Nicola Yuill
Psychology Department, University of Sussex
Brighton, UK
Citation: Rick, Jochen, Yvonne Rogers, Caroline Haig and Nicola Yuill (2009).
“Learning by Doing with Shareable Interfaces.” Children, Youth and
Environments 19(1): 320-341. Retrieved [date] from
http://www.colorado.edu/journals/cye.
Abstract
New technologies, such as multi-touch tables, increasingly provide shareable
interfaces where multiple people can simultaneously interact, enabling co-located
groups to collaborate more flexibly than using single personal computers. Soon,
these technologies will make their way into the classroom. However, little is known
about what kinds of learning activities they will effectively support that other
technologies, such as mobile devices, whiteboards, and personal computers, are
currently unable to do. We suggest that one of the most promising uses of
shareable interfaces is to support learning through exploration and creation. We
present our work on DigiTile as a case study of how shareable interfaces can enable
these forms of learning by doing. We demonstrate how DigiTile supports
collaboration, present a field study on its learning benefits, and show how it can fit
into a larger computing ecology.
Keywords: shareable interfaces, multi-touch tables, learning by doing,
collaborative learning, constructionism
© 2009 Children, Youth and Environments

Learning by Doing with Shareable Interfaces 321
Vision and Background
Personal computers (PCs) are becoming commonplace in school and leisure
settings. How to incorporate them effectively in modern education curricula is now
a central attainment target (Littleton and Light 1999). A new generation of
technologies is about to make its way into the classroom. Many of these are
designed to provide shareable interfaces, where multiple users can interact
simultaneously. For instance, SMART Technologies is planning to market multi-
touch tables for classroom use in 2009. How can these technologies be put to good
pedagogical use? In particular, what kinds of classroom learning activities will they
support and how will they compare with other technologies, such as mobile devices
and PCs?
Since early research can often affect how a technology is understood and utilized
(Bijker 1995), it is timely for developers, educators, and researchers to come
together to determine how this new technology can benefit learning. In this article,
we present our vision of shareable interfaces supporting learning by doing, using
our work on DigiTile as a case study.
Learning by Doing
A core challenge for constructivist educators is designing environments that enable
learning by doing, where learners actively engage with a domain through
exploration and creation (Bruner 1966). One approach is to provide inquiry tools;
learners can actively explore the domain concepts themselves, rather than being
told about them. In constructionism, learners create personally meaningful public
artifacts with tools that allow them to engage important ideas (Papert 1991).
Because the artifacts are personally meaningful, learners are able to forge
meaningful connections with the ideas underlying them. This emphasis on
environments and tools, rather than on subject matter and lesson plans, can be
traced back to Montessori’s prepared environment and Fröbel’s manipulatives
(Standing 1957). The computer can be a powerful tool for creating these systems
(Kay and Goldberg 1977; Rick and Lamberty 2005).
However, it can be difficult to actively engage students in this manner in whole
classroom settings, especially given current teacher-to-student ratios. In contrast,
collaborative work in small groups allows students to actively work with each other
(Cohen 1994; Webb and Palincsar 1996). Moreover, there is growing recognition
that computers can be used to support collaborative learning (Scardamalia and
Bereiter 1991; Dillenbourg 1999). The development of educational software to
support co-located collaboration, however, has until now been constrained by the
available single-user technology (e.g., PDAs, PCs). Shareable interfaces offer much
potential for new possibilities, enabling students to work together more actively for
a number of domains. For example, they allow co-located students to construct
digital content together that are coupled with other learning activities, such as
math. This form of co-construction allows learners to share, discuss, and reflect
upon their own and each others’ ideas.

Learning by Doing with Shareable Interfaces 322
Shareable Interfaces
There are essentially two forms of shareable interfaces: distributed systems and
single display groupware. In distributed systems, multiple components are used in
tandem to create a distributed interface to a single system. While any particular
component may only support a single user, the integration of the components
makes the interface shareable. Because of their distributed nature, these systems
can often adapt to large groups, including whole-class work. Technologies that
support distributed systems include mobile devices and PCs. Single display
groupware, on the other hand, support multiple users interacting simultaneously at
the same display (Stewart, Bederson, and Druin 1999). Since the displays are
limited in size, these systems tend to support small groups (i.e., two to four
learners), typically working together on the same task. Technologies that support
single groupware include whiteboards and tabletops.
In the Ambient Wood project, groups of 10- to 12-year-olds used a variety of
shared technologies to discover more about invisible ecological processes while
exploring a physical woodland (Rogers et al. 2004b). These included mobile devices
and shared displays that were customized to support the learning activity of
scientific inquiry. In RoomQuake (Moher 2006), the classroom itself essentially
became a shareable interface through a combination of Pocket PCs strategically
placed throughout the classroom at stations. Throughout a normal school day,
seismic events were programmed to randomly occur to make the simulation more
authentic. When they occurred, the whole class broke up into small groups to
record dynamic readings of the simulated earthquakes. The recorded readings were
then used to create a physical model of the earthquake; string and Styrofoam balls
were hung from the classroom ceiling to physically show the epicenter of the
digitally recorded earthquakes. In a follow-up project, WallCology (Moher et al.
2008), wall displays were designed to simulate windows as part of the classroom
wall, with simulated life forms living beneath. Classrooms of children collaborated to
track the creatures and understand what conditions allowed them to prosper.
Early work on single groupware focused on using multiple computer mice or pens as
input to traditional desktop PCs. Learning applications that were developed included
collaborative storytelling (Benford et al. 2000) and learning about prime numbers
(Scott, Mandryk, and Inkpen 2002). This early work demonstrated how working
simultaneously could be more enjoyable and effective than taking turns (Inkpen et
al. 1999). A more recent input development for single display groupware is tangible
interaction, where users physically manipulate tangible artifacts to interact with
digital content. Learning applications developed for this approach include reading
(Sluis et al. 2004), rhetorical skills (Stringer et al. 2004), programming (Gallardo,
Julià, and Jordà 2008), and dynamic systems (Zuckerman, Arida, and Resnick
2005).
One of the most promising technologies to support single display groupware is the
multi-touch table, a horizontal display that can detect multiple concurrent touches.
Because the display is also the touch interface, users can manipulate the object
directly, creating a more appealing and natural means of input (Shen, Everitt, and
Ryall 2003). Groups can “dive in” to a task, since communication, such as pointing

Learning by Doing with Shareable Interfaces 323
out an object to a group and seeing what others are doing, is improved (Rogers,
Lim, and Hazlewood 2006). Compared to PCs, tabletops are more likely to elicit
contributions from all members of a group and encourage more equal decision-
making and problem-solving (Rogers et al. 2004a). Multi-touch tables have been
developed so far to support language learning (Morris et al. 2005) and social skills
development for autistic children (Piper et al. 2006).
Case Study: DigiTile
To show how shareable interfaces can support learning by doing, we present our
work on DigiTile as a case study (c.f. Yin 2003). DigiTile is being developed and
studied as part of the ShareIT Project, which aims to understand how new
technologies can support co-located collaboration. The case study is presented in
four sections. First, we introduce DigiTile, showing how we adapted a learning
application originally developed for a desktop PC to run on a multi-touch table.
Second, we describe how children collaborate using the shareable version of
DigiTile. Third, we present a field study on DigiTile’s learning benefits. Fourth, we
demonstrate how DigiTile can be situated in a more informal learning space outside
of the classroom, as part of a larger computing ecology.
From DigiQuilt to DigiTile
DigiTile is an adaptation of DigiQuilt (Lamberty and Kolodner 2002) that runs on a
DiamondTouch multi-touch table (Rick and Rogers 2008). Like DigiQuilt, DigiTile is
a construction kit (Resnick, Bruckman, and Martin 1996) for learning about math
and art by designing colorful mosaic tiles. In addition to being aesthetically
pleasing, these tiles lend themselves to mathematical analysis. The designs embody
fraction concepts and are often symmetric. For instance, the design in Figure 1 is
half red and half yellow; it is also diagonally symmetric. When using DigiQuilt or
DigiTile, learners are given increasingly difficult challenges to accomplish, such as
creating a design that is half red or creating a design that is horizontally symmetric.
DigiTile is intended to enable two co-located learners to place pieces simultaneously
on the tile using touch input.
While DigiTile uses an interactive tabletop to support two concurrent users,
DigiQuilt uses a conventional PC to support a single user. While DigiQuilt’s
interface was not intended to support collaboration, its users did frequently
collaborate. Users often made comments on others’ designs and occasionally
sought to take over the controls to demonstrate something to a peer (Lamberty
2007). This led us to assume that it would be well-suited to being further developed
as a collaborative tool. The design of DigiQuilt was also inspired by Roschelle’s
(1996) theory of convergent conceptual change—when two learners work together
with a reflective tool (a tool that responds to user input to reflect the embedded
domain concepts), they tend to converge on an understanding that is better than
either would achieve independently. When learners work together on a challenge,
they need to articulate their approach for solving the challenge based on their
current understanding. If their strategies clash, it causes a conflict. For Piaget
(1978), socio-cognitive conflict is the central mechanism for the development of
individual knowledge; the conflict has a disequilibrating effect on the individual and
hence he begins to test out new ideas. The reflective tool enables learners to

Learning by Doing with Shareable Interfaces 324
demonstrate or test their understanding. Thus, the conflict can be addressed and
the individuals may develop deeper understanding of the subject.
Figure 1. Two users simultaneously placing pieces on a 4-by-4 tile
As they work on the challenge together, learners’ conceptual understandings do not
just converge with each other, but also with the domain concepts embodied in the
tool (Roschelle 1996). Like DigiQuilt, DigiTile is a reflective tool, giving feedback
when changes are made (e.g., the fraction updates when a new piece is placed). In
addition, visual representations are particularly well-suited for discussion, so
conflicts are easier to resolve (Jehng and Chan 1998). Our aim is that two learners
working together with DigiTile will engage in convergent conceptual change. In this
way, DigiTile has the potential to support collaborative learning.
DigiTile’s interface is split into four main areas (Figure 1). The central tile [1] is a
square grid (2-by-2, 3-by-3, 4-by-4, or 5-by-5) of snaps. Pieces can be dragged
into these snaps to create a colorful tile. Users are each provided with their own
palette area [2] on the left or right side to choose pieces of different colors and
shapes. Pressing on a color button changes the color of the palette pieces. To
provide feedback on fraction tasks, each color button displays the fraction of the
central tile corresponding to that color. Users can drag pieces to one of the five
work snaps [3]. There, multiple pieces can be assembled and rotated. When pieces
are dragged out of a work snap, a copy is automatically created; thus, the
assembly and rotation work does not have to be repeated. To clear a work snap,
the user presses the eraser button below that snap. A scrollable bar contains a

Learning by Doing with Shareable Interfaces 325
graphical history [4] of how the tile was created. By clicking on a thumbnail, users
can revert to an older version of the tile. Using the grid menu, users can overlay
different lines on their tile; this menu also serves as a feedback mechanism,
indicating whether the current design is symmetric along the respective lines.
To further utilize convergent conceptual change, DigiTile allows users to change the
representation of the color fraction from reduced fraction to least-common-divisor
fraction to percentage to visual pie chart, as chosen by the buttons on top of Area 2
(Figure 1). One learner can display his or her palette with percentages, while the
other displays reduced fractions. This allows us to create challenges like “a tile that
is ½ red and 50% yellow.” This challenge allows the learners to discover that one-
half and 50 percent are the same. Using multiple-linked representations to
represent a mathematical concept leads to deeper understanding (Kaput 1989).
This strategy would not work as well in the single-user DigiQuilt.
Adapting a single-user desktop application to a multi-touch table is not
straightforward. There are many design decisions to take into consideration. These
include group size, table size (Ryall et al. 2004), display orientation, and
territoriality (Scott et al. 2004).
Based on convergent conceptual change, dyads seemed an ideal grouping for
learners to challenge each other, while not making the collaboration too difficult.
We wanted the learners to collaborate on one design, so we positioned the quilt
block in the middle of the table (Figure 1); thus, learners can reach it from any
location. The size of the table also needs to allow intended users to reach where
they want to. In our case, the DiamondTouch table, measuring 32 inches
diagonally, proved to be adequate for our intended users.
Depending on where users are sitting, the orientation of the display will affect how
it is used. Unlike a desktop display, users sitting on different sides of the table will
have different perspectives on a tabletop display. They could sit across from each
other, around the corner, or next to each other. This makes it difficult to maintain
the desktop orientation, where eyes up equals the top of the screen and eyes down
means the bottom of the screen (Tang et al. 2006). Developing software without a
dominant orientation can be tricky (Shen et al. 2004). For DigiTile, different
positions would change how learners view the tile. If seated around a corner,
horizontal symmetry for one partner would be vertical symmetry for the other. If
seated across from each other, the tile would appear upside down. For many
patterned designs, this would not be a problem; however, many DigiQuilt designs
were based on real-life objects that have a preferred orientation, such as a stick
figure (Lamberty 2007). In addition, while adults tend to favor working across from
each other, children often prefer working next to each other (Scott, Grant, and
Mandryk 2003). So, we positioned the learners next to each other to share the
same orientation. This configuration (two learners next to each other) allows
DigiTile to maintain a desktop orientation (Figure 1).
Territoriality refers to who “owns” what parts of the shared display and has been
observed to occur in user studies of group interactions at tabletops. To encourage

Learning by Doing with Shareable Interfaces 326
individual ownership, we created a left and a right palette area, one for each user.
To encourage sharing, we intentionally used an odd number of work snaps, so that
one would be in the middle, implying no left or right ownership.
After some informal user testing, it became obvious that dropping a piece
accidentally was fairly common in the multi-touch case. In addition, partners often
disapproved of a change. For these reasons, navigating the history was more
important for DigiTile than DigiQuilt. While undo and redo buttons sufficed to
navigate the DigiQuilt history, we developed a graphical history for DigiTile that
allows users to more easily back up to a previous design point.
DigiTile in Action
In the previous section, we described our vision of DigiTile use, based on DigiQuilt
experience and learning theory. To get a better sense of how children actually use
DigiTile, we conducted a small user study with pairs of 9 and 10 year olds. This
formative study enabled us to determine what level of difficulty to set for
challenging design tasks that would engender collaboration. Below, we detail one
exchange to illustrate DigiTile use.
To familiarize study participants with DigiTile and its touch interface, the pairs were
asked to design a “good looking” tile. During this familiarization phase, the children
commented on each other’s designs, sometimes leading to conflicts. It can be
difficult to compromise in a way that does not leave one participant unhappy with
the results. Since there was no right or wrong choice, participants often clashed
about the design choices. Occasionally, we reminded them of the need to work
together.
After the familiarization phase, we assigned the pairs mathematical challenges. The
goal of this phase was to see how pairs would collaborate when faced with a
difficult challenge—one that required significant thought and work. As observed
with DigiQuilt (Lamberty 2007), when given a challenge, users shift their focus to
solving the challenge, rather than creating an aesthetically interesting tile.
Consequently, aesthetic-based conflicts disappear. The initial challenges proved
trivial and were accomplished in less than a minute with minimum planning and
negotiation between the participants. Hence, we increased the difficulty until the
task became challenging. At that point, the pairs had a more difficult time forming
and trialing approaches. Because both participants could interact simultaneously, it
was necessary for one participant to recruit the other in trialing an approach.
Below is an excerpt from one pair using DigiTile to address the challenge of creating
a tile that is half red. While this is usually an easy challenge, it proved to be quite
difficult as the participants were using a 5-by-5 grid. In this case, the total number
of grid squares (25) is not evenly divisible by two, so the task cannot be
accomplished with whole-square pieces alone. After several failed attempts, the
pair was able to create a solution (Figure 2) without the help of the researchers.

Learning by Doing with Shareable Interfaces 327
Figure 2. Solution to the half-red challenge
Researcher: [After getting a bit sidetracked, the researcher reminds them of
their original mission.] Remember that you have to get the half red.
Girl: Oh. I know. [She places two black whole-square pieces on the right
side.]
Boy: Oh. I get it. I get it. I’ll do the red ones. I do the red.
Girl: Yeah. But, we need... Wait! There— [She gestures towards the middle
of the tile.] you see that line? Don’t do any over the gray part of that
line. [Because of their previous efforts, the right side of the middle
column is grey and the left side is white.] I’ve just got a really good
idea. [She fills in some black on her side, while he fills in two red on
his side.]
Boy: What’d you do? Oh ya. [He gestures to divide the middle column.] We
need a half, don’t we?
Girl: That’s what I’m going to do. [She works to orient the half-sized-
rectangular black piece to fill in her half of the middle column.
Meanwhile, he fills the left two columns with whole-square red pieces.
See Figure 3.] And, then, you’re going to need to use... [She starts to
point at the half-sized rectangular piece on his side. He cuts her off.]
Boy: Yeah, I know. I am. [He orients the half-sized red piece correctly and
fills in his side of the middle column. Meanwhile, she fills the remaining
empty space on her side with whole-square black pieces. They finish at
roughly the same time.]
Girl: Done.
The pair worked together to accomplish the challenge. While the girl came up with
the plan, the boy was able to follow her lead and grasp her strategy. They both
placed an equal number of pieces to accomplish the task.

Learning by Doing with Shareable Interfaces 328
Figure 3. Working together with DigiTile
A Field Study
To get a better sense of DigiTile’s value for promoting collaborative learning, we
conducted a field study. While collaboration can promote learning, even groups with
capable members often find it difficult to productively collaborate (Barron 2003).
We provided tasks involving fraction understanding (rather than symmetry, shape
orientation, or equivalence across representations). We simplified the interface to
six colors and three shapes (Figure 4). Our hypotheses going in were that DigiTile
would enable collaboration and increase understanding of fractions.
In addition, we were interested in how interface features can encourage
collaboration and learning. Different access to an interface can elicit cooperation
between the participants (Kerawalla et al. 2008). To investigate this possibility, we
set up two conditions. In the split palette condition (Figure 4a), the colors were split
between the two palettes. The left palette contained red, orange, and green; the
right palette contained brown, yellow, and blue. In the shared palette condition
(Figure 4b), both palettes contained all six colors. Consulting with educational
experts in the area of mathematics education, we designed a sequence of three
tasks that were appropriately difficult for the target group. Each task was designed
to utilize the split palettes equally.
Task 1 was to create the half red, half yellow pattern shown in Figure 4a. The
children were given a printout of the pattern, including grid lines, to guide their
work. Task 2 was to create a 4-by-4 tile that was three-eighths orange and three-
eighths brown. Participants were informed that not all of the tile had to be filled in.

Learning by Doing with Shareable Interfaces 329
Task 3 was to create a 5-by-5 tile that was one-tenth red, four-tenths green, three-
tenths yellow, and two-tenths blue; an example of a successful solution is shown in
Figure 4b.
Figure 4. Simplified DigiTile setup
Procedure
21 participants (9 to 11 years old) took part in the study. Children who provided a
consent form (to be videotaped) were randomly allocated to an experimental group
(four pairs in the shared palette condition, four pairs in the split palette condition)
and paired randomly. The five students who did not provide a consent form were
allocated to the control group.
All participants undertook a pre-test. Two similar fraction tests were created using a
“test-based key stage two” computer program. Half the participants received
version one and half received version two. The classroom teacher, who was also the
school’s mathematics coordinator, helped in the formulation of the tests. The tests
contained ten fraction-related questions. Participants were instructed to work in
silence on the test for 25 minutes.
The following week, the DigiTile sessions were conducted in the back of the
classroom during normal class time. Before starting on the first task, the researcher
demonstrated how to move and rotate pieces, change their color, and use the
graphical history. The participants were informed that they were going to work on
three tasks and that they were permitted to talk to each other freely. Each task was
read to the participants; the tasks were also provided on an instruction sheet.
This first task allowed participants to become familiar with DigiTile. At its
completion, the researcher asked, “what fraction of the square is red and what
fraction is yellow?” Most groups were able to come up with the correct answer on
the first try; if not, the researcher helped them to realize it was one half. As a
follow up, the researcher asked, “what is the percentage of that fraction?” These
questions were asked to draw the participants’ attention to the fractions on the

Learning by Doing with Shareable Interfaces 330
color buttons and the different ways that they were represented (least-common-
denominator fraction, reduced fraction, and decimal). The second and third tasks
were given to children without any prompting or questions. The researcher neither
assisted participants with the task, nor revealed solutions. Each session of three
tasks lasted approximately 30 minutes; if children could not complete a task within
a reasonable time frame, they were asked to move to the next task. After 30
minutes, the session was ended. At the end, participants were thanked for their
efforts.
Three days after the last DigiTile session, the participants undertook a post-test.
Those who previously received version one of the test completed version two, and
vice versa. Two of the children who took part in a DigiTile session were absent from
one of the tests; consequently, their data has been excluded from the analysis.
After the post-test, the researcher demonstrated how to complete the three tasks
in DigiTile.
Results
Seven of the eight groups successfully completed both Task 1 and Task 2; one
group was unable to complete either task. While Task 1 was primarily designed to
familiarize participants with DigiTile, it proved to be relatively challenging. In
particular, rotating the triangles into the correct orientation was not trivial. It was
common for participants to drag a rotated piece to a location, realize then that it
did not match their intentions, and drag it back to a work area for further rotation.
Task 2 was usually completed faster than Task 1. Task 3 proved to be quite
difficult. Only one group finished it. One group produced a correct solution at one
time, but failed to notice it was correct. The remaining six groups never completed
the task, although all made significant progress. While most groups were
unsuccessful in finishing the task, they did engage the mathematical concepts in
their efforts.
Several common misconceptions about fractions were noticeable during the
sessions. Occasionally, fraction equivalence was a problem. For example, a group
would work on two-tenths, reach one-fifth, and not realize that these two fractions
are equivalent. Most groups had problems realizing that higher denominators
indicated a smaller fraction, which caused problems. For instance, groups would
stumble onto three-sixteenths along their path to three-eighths. Because they felt
the former was larger than the latter, they would start to remove pieces—an
incorrect strategy. The presence of the small triangle often exacerbated this
problem by creating fractions with large denominators. Several times, a group
would make good progress, reach an impasse, resort to the small triangle as a
possible solution, and fail to make any progress for some time.
All groups used the fraction representations displayed on the color buttons
extensively for working on the second and third tasks. The prominence of the use of
this feature was particularly noticeable since most had overlooked it until the
researcher pointed it out after the first task. Often, one partner would add pieces,
while the other kept a watch on the fraction representation.

Learning by Doing with Shareable Interfaces 331
Figure 5 shows the mean test scores for the two experimental conditions and the
control group. A one-way (Group: Shared Palette vs. Split Palette vs. Control)
independent ANCOVA (post-test scores) was conducted in order to discover
whether DigiTile improved learning of fraction knowledge. Pre-test scores were
added as the covariate. After controlling for the effect of pre-test scores, there was
a significant main effect on the experimental group, F(3, 15) = 3.45, p < .05.
Contrasts indicated that the experimental groups who underwent the DigiTile
session had significantly higher scores on the post-test compared to the controls, p
< .05. There was no significant difference between the post-test results of the
shared and split palette groups, p = ns.
Figure 5. Pre- and post-test results by condition
In creating the two experimental conditions, we considered the split palette
condition would improve collaboration, as participants would have to collaborate to
complete the tasks. The results showed, however, that a lack of collaboration was
not a problem in any of the groups. Irrespective of condition, the children worked
well together. Because the second and third challenges were relatively difficult and
moves often affected each other, the partners quickly figured out that they had to
work together. On both the second and third task, the vast majority of the time was
spent working jointly on a single color, rather than trying to work simultaneously on
multiple colors.
The children also had no qualms about reaching across the table to pick up pieces
from the other palette. While adults are often sensitive about their personal space,
children are not. Reaching across to the other palette was also observed in the
shared palette condition. In those cases, the move often modeled a behavior to the
other participant (e.g., “see, you can put it here”). In one split palette group, the
participants elected to switch places, so that one of them could more easily
demonstrate an idea.

Learning by Doing with Shareable Interfaces 332
Discussion
This study demonstrates the benefit of collaboration for learning. Participants were
able to work together (e.g., one places pieces while the other watches the fraction
representation), model behavior for their partners, and articulate strategies and
concepts for each other. With almost no guidance from the researcher or teacher,
the children showed significant improvement in their understanding of fractions.
While we are encouraged by the results, there are several caveats to consider.
First, since participation was based on returning a consent form, condition
assignment was not truly random. However, since the pre-test scores for the
control condition was similar to the other conditions, no overt sampling bias was
obvious. Second, while the researcher did not assist on the task, her presence and
the presence of the cameras was certainly an incentive for the children to stay on
task and be on their best behavior.
While the results of the study were positive, this learning scenario was less than
ideal. Thirty minutes is not a long time to spend on task. The three tasks certainly
did not exhaust the potential of DigiTile. To truly integrate DigiTile into a class
curriculum, more time and a wider variety of tasks could be useful. In order to
make the experience as similar as possible across groups, the researcher kept the
session on a strict task progression. Because of this, one group was unable to finish
any of the tasks; consequently, their self-confidence waned (as one child
concluded, “this is hard”). While that group was slower than the other groups, they
were making progress. It is likely that they would have completed the first two
tasks if given the full 30 minutes on just those two tasks. Ideally, the task difficulty
could be adjusted to the children’s abilities, providing each group with tasks that
are challenging yet ultimately doable.
While children were able to use DigiTile, there were several recurring problems with
the hardware. When children bumped into the table, the DiamondTouch moved,
throwing off the alignment with the projected image. This caused problems with the
interface. For instance, pressing on a piece would not work as the mouse pointer
would be located at a different point than the finger touch. It often took the
researcher a few minutes to catch this problem and rectify it. In addition, the mats
that received the signal often interfered with each other, which can happen on
conductive floors. This caused problems, e.g., on a single touch, a piece in a work
space would rotate by 180 degrees, rather than the usual 90, since the software
received two touches (one from each pad). While a layer of non-conductive material
placed underneath the pad could have solved this problem, the researcher was
unaware of this.
Beyond the Tabletop
While the tabletop can be a compelling learning environment by itself, it can also fit
into a larger computing ecology. To illustrate this potential, we detail our
experience of using DigiTile at a large science festival aimed at families. The one-
day festival featured interactive exhibits in different classrooms of the host school;
hundreds of children, often accompanied by their parents, went from room to room
throughout the day to interact with the various exhibits and tools. DigiTile was

Learning by Doing with Shareable Interfaces 333
networked using two computers. One ran the DigiTile application and the other was
used for printing and displaying the tiles completed by the children.
When saving their designs, children named them and identified themselves as
authors. The researcher at the second computer then printed a nametag featuring
the design and first name of its creator for the children to wear. This served two
purposes. First, it turned their virtual design into a physical artifact. The nametag
allowed children to display their design prominently to their friends and family.
Second, it served as an advertisement for our exhibition as children would ask
previous participants where they got their nametags. The nametags were very
popular, with several requests for second printings. A core emphasis of
constructionism is on creating personally meaningful public artifacts. This example
demonstrates how those artifacts, when used effectively, can motivate learners,
embody domain concepts, and promote a culture of learning.
The other function of the second computer was to drive a large plasma screen,
positioned by the door. The display program picked a design created that day at
random. It ran through the history to show how the design was created. The name
of the design and its creators were displayed below the tile. After arriving at the
end of the history, the grid disappeared and the final design was displayed for 30
seconds. Like the nametags, the large display served as both public artifact and
advertisement. Several children waited to see their designs get the big screen
treatment before leaving the room. Another purpose for the big display was to
emphasize the design challenges.
Because the children visiting the exhibition varied in age and ability, we did not
directly assign any challenges. We did, however, post four challenges (half red,
one-third blue, vertical line of symmetry, and diagonal line of symmetry) that
participants could tackle. Some of the older participants found these interesting.
When a final design was displayed on the large screen, star-shaped awards
appeared above the tile for the respective challenges it met. These awards served
as an incentive for new1 participants to meet these challenges. It also became a
starting point for observers to talk about the target concepts—fractions and
symmetry. Towards the end of the day, a group of boys came in and spent over 30
minutes just looking at the designs that had been created during the day (Figure
6). As expected, they were excited about seeing their own designs; they also
started talking about others’ designs. They discussed why they liked certain designs
(colors, shapes, symmetry, etc.). They also noticed the awards and started pointing
out the symmetry and how the design process led to that result. It became a
natural opportunity to engage the target concepts.
This example demonstrates how different computing components can work
together. The display components enhanced the meaning of the created artifact. It
allowed the constructed artifact to be more public (the display) and personal (the
1Due to limited resources, we were unable to allow visitors to complete more than one
design, even though many would have liked to.

Learning by Doing with Shareable Interfaces 334
name tags). It served as advertisement for others to create and became a source of
discussion. In the classroom, an interactive whiteboard could allow students to
present their work to each other; the teacher could then lead the kids to reflect on
the target concepts, based on these displays. Outside the classroom, a public
display could show the work of the students inside that classroom to others in the
hallway. As our example demonstrated, this can lead to useful discussion.
Figure 6. A group discusses a design that met three design challenges
Discussion
Previously, we provided a case of shareable interfaces supporting learning by doing.
We envision that this sort of constructive use will become more commonplace in
education, as shareable interfaces become part of classrooms. However, significant
challenges lie ahead for this vision to be realized on a large scale. In this section,
we discuss the challenges and suggest ways to overcome them.
Design and Usability Challenges
Multi-touch hardware and software support is still in its early stages. The hardware
is just starting to become commercially available. As was illustrated in our field
study, there are still hardware obstacles and concomitant usability issues to
overcome. For example, nudging the table can cause the projected image to move
out of alignment. The conductive mats are sensitive to a conductive floor or non-
conductive shoes (we had to ask several participants to take off their shoes for the
table to properly register their touches). Other multi-touch technologies work
differently and thus have different problems. Infrared light based tables, for

Learning by Doing with Shareable Interfaces 335
instance, have problems working in the presence of sunlight. Software support, too,
poses a difficult interface and programming challenge. Simultaneous users and
touch input are more complex than the one-user mouse input for which desktop
toolkits are built. One cannot simply use existing software with the new interface.2
DigiTile is a relatively simple application, where translating the touch input into
discreet mouse points works well. Other applications may benefit from more
sophisticated input afforded by touch, such as pinching or putting a fist down.
Developing toolkits that take advantage of touch input is still an ongoing challenge.
On the positive side, there is a growing interest among commercial companies in
supporting touch. For example, Apple’s iPhone supports a multi-touch interface, and
Microsoft has announced that the next version of Windows will support multi-touch
input.
Fundamental research questions of how this kind of shareable technology can
support co-located collaboration still need to be explored. Learning is a particularly
challenging area, since techniques that work well with one age group may be
inappropriate for others. For example, certain tabletop learning applications may
not be suitable for 4 to 6 year olds because they often lack the dexterity to
smoothly drag their fingers across the table without lifting them (Mansor, De
Angeli, and De Bruijn 2008). Even older children, such as those in our studies, will
lift their fingers while dragging digital objects, such as a DigiTile piece. These
“accidental hops” can cause the objects to be unintentionally dropped in the wrong
place. While these mistakes are easy to recover from they can be frustrating.
Classroom Challenges
When a new technology arrives on the scene, there is often much excitement about
its ability to fundamentally change human processes and to realize a new
pedagogical vision on a large scale (e.g., Postman and Weingartner 1969). Visions,
however, rarely become reality. In the 1980s, Papert (1993) cast the personal
computer as the device that would allow learners to take charge of their own
learning. While that vision was compelling, it did not come to fruition for a number
of reasons (Noss and Hoyles 1996). In the 1990s, the U.S. put a major focus on
networking the classrooms and connecting it to the wealth of information and
resources that is the Internet. Again, this vision has fallen short of revolutionizing
learning (Bruckman 1999). More recently, there has been a big push to bring
interactive whiteboards into UK schools. But, as an early progress report noted, this
technology can “reinforce a transmission style of whole-class teaching in which the
contents of the board multiply and go faster, whilst pupils are increasingly reduced
to a largely spectator role” (Moss et al. 2007).
While most grand visions of technology-driven pedagogic change have not been
realized on a large scale, there are reasons to be optimistic about shareable
2 To be fair, touch input can be mapped to traditional mouse input, but then the multi-user
and multi-touch aspects are lost. While it is possible to map simultaneous touch input to
traditional input (Tse, Greenberg, and Shen 2006), that approach is still quite limited.

Learning by Doing with Shareable Interfaces 336
interfaces, such as multi-touch tables. First, the applications created for them are
most compelling when multiple people use them together. This can encourage
small-group learning in the classroom. Second, the form factors prevent them from
being used like whiteboards: while it is easy for a teacher to appropriate an
interactive whiteboard for a lecture, gathering the whole class around a (small)
table is more awkward. Third, many shareable interfaces can run different
applications to engage a variety of learning domains. Fourth, shareable interfaces
build on one of the real strengths of the classroom—the potential for collaboration
between peers. Hence, shareable technologies have the potential to succeed in
changing the learning process to be more active in the classroom where others
have failed.
If applications like DigiTile are to make it into a normal classroom, however, the
classroom computing ecology needs to support it. One challenge for small group
technology is that it needs to be replicated for each group. If an entire class wants
to work with DigiTile at the same time, that classroom would need one multi-touch
table for every two students. While this may seem prohibitively expensive now, it is
becoming cheaper each year to embed computing technology in everyday things
(Greenfield 2006). Another model to consider for future classroom ecology is
Montessori’s prepared environment, where different sections of the classroom are
set up for different activities and students choose where they want to be. A single
multi-touch table may be set up for DigiTile, while other engaging activities fill the
rest of the classroom. The Montessori method is generally reserved for younger
children, but newer technologies could make such a style of free explorations viable
for learning domains generally reserved for older students (Zuckerman et al. 2005).
Acknowledgements
This work is part of the ShareIT project (http://shareitproject.org) funded by the EPSRC,
grant number EP/F017324/1. We would like to thank our other collaborators in that project
for their helpful feedback: (alphabetically) Nick Dalton, William Farr, Rowanne Fleck,
Amanda Harris, Eva Hornecker, Paul Marshall, Shems Marzouq, Richard Morris, Nadia
Pantidi, and Nicola Yuill. Special thanks go to K.K. Lamberty for supporting us in adapting
DigiQuilt and Kim Bryant for creating the sketched figures. We thank MERL for loaning us
the DiamondTouch tables.
Jochen Rick is a research fellow at the Open University, working on the ShareIT
Project. His primary role on the project is as technologist—designing and
implementing novel pervasive computing applications. His background is in
educational and collaborative technologies. He was the lead developer of Swiki, the
first wiki developed for educational research. For his dissertation, he designed and
implemented AniAniWeb, a personal home page system built on a wiki foundation,
and studied its adoption by graduate students, a particularly relevant group of early
adopters. In May 2007, he received a Ph.D. in Computer Science from the Georgia
Institute of Technology.
Since 2006, Yvonne Rogers has been a professor of Human-Computer Interaction
in the Computing Department at the Open University, where she directs the

Learning by Doing with Shareable Interfaces 337
Pervasive Interaction Lab. From 2003–2006, she had a joint appointment in the
schools of Informatics and Information Science at Indiana University (where she
continues to be a visiting professor). Prior to that, she was a professor in the
former School of Cognitive and Computing Sciences at Sussex University. She has
also been a Visiting Professor at Stanford, Apple, Queensland University, and UCSD.
At an early stage of her career, she worked at Alcatel.
Caroline Haig completed a Masters degree in Psychology at the University of
Sussex in 2008.
Nicola Yuill is a senior lecturer in developmental psychology at the University of
Sussex and runs the Sussex Children and Technology (ChaT) Lab. Her research
interests lie at the interface of social cognitive development and educational
technology: for example, how technology can support children's learning about
language through mediated discussion, the role of mothers' conversations in
supporting social cognitive development through childhood, and the role of
mediated technology in supporting playful learning in children with typical and
atypical development.
References
Barron, B. (2003). “When Smart Groups Fail.” The Journal of the Learning
Sciences 12(3): 307–359.
Benford, S., B.B. Bederson, K.-P. Akesson, V. Bayon, A. Druin, P. Hansson,
et al. (2000). “Designing Storytelling Technologies to Encouraging Collaboration
between Young Children.” In Proceedings of CHI ’00. New York: ACM Press, 556–
563.
Bijker, W.E. (1995). Of Bicycles, Bakelites, and Bulbs: Toward a Theory of
Sociotechnical Change. Cambridge, MA: The MIT Press.
Bruckman, A. (1999). “The Day after Net Day: Approaches to Educational Use of
the Internet.” Convergence 5(1): 24–46.
Bruner, J.S. (1966). Towards a Theory of Instruction. Cambridge, MA: The
Belknap Press of Harvard University Press.
Cohen, E.G. (1994). “Restructuring the Classroom: Conditions for Productive Small
Groups.” Review of Educational Research 64(1): 1–35.
Dillenbourg, P. (1999). “What Do You Mean by Collaborative Learning?” In
Dillenbourg, P., ed. Collaborative-Learning: Cognitive and Computational
Approaches. Oxford: Elsevier, 1–19.

Learning by Doing with Shareable Interfaces 338
Gallardo, D., C.F. Julià, and S. Jordà (2008). “Turtan: A Tangible Programming
Language for Creative Exploration.” In Proceedings of TABLETOP ’08. Los Alamitos,
CA: IEEE, 95–98.
Greenfield, A. (2006). Everyware: The Dawning Age of Ubiquitous Computing.
Berkeley, CA: New Riders.
Inkpen, K.M., W. Ho-Ching, O. Kuederle, S.D. Scott, and G.B.D. Shoemaker
(1999). “‘This is fun! We’re all best friends and we’re all playing’: Supporting
Children’s Synchronous Collaboration.” In Proceedings of CSCL ’99. Mahwah, NJ:
Lawrence Erlbaum Associates, 252–259.
Jehng, J.-C.J. and T.-W. Chan (1998). “Designing Computer Support for
Collaborative Visual Learning in the Domain of Computer Programming.” Computers
in Human Behavior 14(3): 429–448.
Kaput, J. (1989). “Linking Representations in the Symbolic Systems of Algebra.” In
Wagner, S. and C. Kieran, eds. Research Agenda for Mathematics Education:
Research Issues in the Learning and Teaching of Algebra. Reston, VA: National
Council of Teachers of Mathematics, 167–194.
Kay, A. and A. Goldberg (1977). “Personal Dynamic Media.” IEEE Computer
(March): 31–41.
Kerawalla, L., D. Pearce, N. Yuill, R. Luckin, and A. Harris (2008). “‘I’m
keeping those there, are you?’ The Role of a New User Interface Paradigm—
Separate Control of Shared Space (SCOSS)—in the Collaborative Decision-Making
Process.” Computers and Education 50(1): 193–206.
Lamberty, K.K. (2007). Getting and Keeping Children Engaged with a
Constructionist Design Tool for Craft and Math. Unpublished doctoral dissertation,
Georgia Institute of Technology, Atlanta, GA.
Lamberty, K.K., and J.L. Kolodner (2002). “Exploring Digital Quilt Design Using
Manipulatives as a Math Learning Tool.” In Bell, P., R. Stevens, and T. Satwicz, eds.
Proceedings of ICLS 2002. Mahwah, NJ: Lawrence Erlbaum Associates, 552–553.
Littleton, K. and P. Light, eds. (1999). Learning with Computers: Analysing
Productive Interactions. London: Routledge.
Mansor, E.I., A. De Angeli, and O. De Bruijn (2008). “Little Fingers on the
Tabletop: A Usability Evaluation in the Kindergarten.” In Proceedings of TABLETOP
’08. Los Alamitos, CA: IEEE, 99–102.
Moher, T. (2006). “Embedded Phenomena: Supporting Science Learning with
Classroom-Sized Distributed Simulations.” In Proceedings of CHI 2006. New York:
ACM Press, 691–700.

Learning by Doing with Shareable Interfaces 339
Moher, T., B. Uphoff, D. Bhatt, B.L. Silva, and P. Malcolm (2008). WallCology:
Designing Interaction Affordances for Learner Engagement in Authentic Science
Inquiry.” In Proceedings of CHI 2008. New York: ACM Press, 163–172.
Morris, M.R., A.M. Piper, A. Cassanego, and T. Winograd (2005). Supporting
Cooperative Language Learning: Issues in Interface Design for an Interactive Table.
Research Report No. 8. Palo Alto, CA: Department of Computer Science, Stanford
University.
Moss, G., C. Jewitt, R. Levačić, V. Armstrong, A. Cardini, and F. Castle
(2007). The Interactive Whiteboards, Pedagogy and Pupil Performance Evaluation:
An Evaluation of the Schools Whiteboard Expansion (SWE) Project: London
Challenge. Research Report No. 816. London: Department for Education and Skills,
Institute of Education.
Noss, R., and C. Hoyles (1996). “Cultures and Change.” In Windows on
Mathematical Meanings: Learning Cultures and Computers. Boston: Kluwer.
Papert, S. (1991). “Situating Constructionism.” In Harel, I. and S. Papert, eds.
Constructionism. Ablex, 1–13.
Papert, S. (1993). Mindstorms: Children, Computers and Powerful Ideas. Second
ed. New York: Basic.
Piaget, J. (1978). Success and Understanding. London: Routledge and Kegan Paul.
Piper, A.M., E. O’Brien, M.R. Morris, and T. Winograd (2006). “SIDES: A
Cooperative Tabletop Computer Game for Social Skills Development.” In
Proceedings of CSCW ’06. New York: ACM Press, 1–10.
Postman, N. and C. Weingartner (1969). Teaching as a Subversive Activity. New
York: Dell.
Resnick, M., A. Bruckman, and F. Martin (1996). “Pianos Not Stereos: Creating
Computational Construction Kits.” Interactions 3(6): 41–50.
Rick, J. and K.K. Lamberty (2005). “Medium-Based Design: Extending a Medium
to Create and Exploratory Learning Environment.” Interactive Learning
Environments 13(3): 179–212.
Rick, J. and Y. Rogers (2008). “From DigiQuilt to DigiTile: Adapting Educational
Technology to a Multi-Touch Table.” In Proceedings of TABLETOP ’08. Los Alamitos,
CA: IEEE, 79–86.
Rogers, Y., W. Hazlewood, E. Blevis, and Y.-K. Lim (2004a). “Finger Talk:
Collaborative Decision-Making Using Talk and Fingertip Interaction around a
Tabletop Display.” In Proceedings of CHI ’04 Extended Abstracts. New York: ACM
Press, 1271–1274.

Learning by Doing with Shareable Interfaces 340
Rogers, Y., S. Price, G. Fitzpatrick, R. Fleck, E. Harris, H. Smith, et al.
(2004b). “Ambient Wood: Designing New Forms of Digital Augmentation for
Learning Outdoors.” In Proceedings of IDC ’04. New York: ACM Press, 3–10.
Rogers, Y., Y.-K. Lim, and W.R. Hazlewood (2006). “Extending Tabletops to
Support Flexible Collaborative Interactions.” In Proceedings of TABLETOP ’06.
Washington, D.C.: IEEE Computer Society, 71–78.
Roschelle, J. (1996). “Learning by Collaborating: Convergent Conceptual Change.”
In Koschman, T., ed. CSCL: Theory and Practice of an Emerging Paradigm.
Mahwah, NJ: Lawrence Erlbaum Associates, 209–248.
Ryall, K., C. Forlines, C. Shen, and M.R. Morris (2004). “Exploring the Effects of
Group Size and Table Size on Interactions with Tabletop Shared-Display
Groupware.” In Proceedings of CSCW ’04. New York: ACM Press, 284–293.
Scardamalia, M. and C. Bereiter (1991). “Higher Levels of Agency for Children in
Knowledge Building: A Challenge for the Design of New Knowledge Media.” The
Journal of the Learning Sciences 1(1): 37–68.
Scott, S.D., K.D. Grant, and R.L. Mandryk (2003). “System Guidelines for Co-
located, Collaborative Work on a Tabletop Display.” In Kuutti, K., E.H. Karsten, G.
Fitzpatrick, P. Dourish, and K. Schmidt, eds. Proceedings of ECSCW 2003.
Netherlands: Kluwer, 159–178.
Scott, S.D., R.L. Mandryk, and K.M. Inkpen (2002). “Understanding Children’s
Interactions in Synchronous Shared Environments.” In Stahl, G., ed. Proceedings of
CSCL 2002. Mahwah, NJ: Lawrence Erlbaum Associates.
Scott, S.D., M. Sheelagh, T. Carpendale, and K.M. Inkpen (2004).
“Territoriality in Collaborative Tabletop Workspaces.” In Proceedings of CSCW ’04.
New York: ACM Press, 294–303.
Shen, C., K. Everitt, and K. Ryall (2003). “UbiTable: Impromptu Face-to-Face
Collaboration on Horizontal Interactive Surfaces.” Lecture Notes in Computer
Science 2864: 281–288.
Shen, C., F.D. Vernier, C. Forlines, and M. Ringel (2004). “DiamondSpin: An
Extensible Toolkit for Around-the-Table Interaction.” In Proceedings of CHI ’04. New
York: ACM Press, 167–174.
Sluis, R.J.W., I. Weevers, C.H.G.J. van Schijndel, L. Kolos-Mazuryk, S.
Fitrianie, and J. B. O.S. Martens (2004). “Read-It: Five-to-Seven-Year-Old
Children Learn to Read in a Tabletop Environment.” In Proceedings of IDC ’04. New
York: ACM Press, 73–80.
Standing, E.M. (1957). Maria Montessori: Her Life and Work. New York: Plume.

Learning by Doing with Shareable Interfaces 341
Stewart, J., B.B. Bederson, and A. Druin (1999). “Single Display Groupware: A
Model for Co-Present Collaboration.” In Proceedings of CHI ’99. New York: ACM
Press, 286–293.
Stringer, M., E.F. Toye, J.A. Rode, and A.F. Blackwell (2004). “Teaching
Rhetorical Skills with a Tangible User Interface.” In Proceedings of IDC ’04. New
York: ACM Press, 11–18.
Tang, A., M. Tory, B. Po, P. Neumann, and S. Carpendale (2006).
“Collaborative Coupling over Tabletop Displays.” In Proceedings of CHI 2006. New
York: ACM Press.
Tse, E., S. Greenberg, and C. Shen (2006). “GSI Demo: Multiuser Gesture /
Speech Interaction over Digital Tables by Wrapping Single User Applications. In
Proceedings of ICMI ’06. New York: ACM Press, 76–83.
Webb, N.M. and A.S. Palincsar (1996). “Group Processes in the Classroom.” In
Berliner, D.C. and R.C. Calfee, eds. Handbook of Educational Psychology. Simon
and Schuster, 841–873.
Yin, R.K. (2003). Case Study Research: Design and Methods. 3rd ed. Thousand
Oaks, CA: Sage.
Zuckerman, O., S. Arida, and M. Resnick (2005). “Extending Tangible Interfaces
for Education: Digital Montessori-Inspired Manipulatives.” In Proceedings of CHI
’05. New York: ACM Press, 859–868.