Conference PaperPDF Available

Mad Mixologist: Exploring How Object Placement in Tangible Play Spaces Affects Collaborative Interaction Strategies

Authors:

Abstract and Figures

Tangible games afford an engaging and often unique form of hybrid play (i.e., physical-digital elements), but there is currently limited work explicitly exploring how these games can be designed to provide spatial affordances that implicitly encourage collaboration. In this paper, we present a novel col-laborative tangible game, titled Mad Mixologist, and investigate how making a simple change in the location of game objects within the tangible play space can lead to significantly different collaborative interaction strategies. The results from our group comparison study indicate that 1) players with exclusively direct access to multiple relevant resources (i.e., a digital instruction and a physical object) were more likely to assume responsibility for completing tasks in a shared play space and 2) distributing these same task-relevant resources across multiple players created ambiguity over whether the player with the digital or physical resource should engage with the shared play space. This study demonstrates one possible way in which the physical design of a tangible game can be arranged to implicitly encourage players to develop more collaborative interaction strategies, specifically by distributing exclusive resources across players. Overall, this study highlights and reinforces the connection between spatial affordances and social interactions via embodied facilitation within the context of collaborative tangible games.
Content may be subject to copyright.
Mad Mixologist: Exploring How Object Placement
in Tangible Play Spaces Affects Collaborative
Interaction Strategies
Katelyn M. Grasse, Marjorie Ann M. Cuerdo and Edward F. Melcer
Alternative Learning Technologies and Games Lab, Computational Media Department
University of California, Santa Cruz
Santa Cruz, CA, USA
{katy, mcuerdo, eddie.melcer}@ucsc.edu
Abstract—Tangible games afford an engaging and often unique
form of hybrid play (i.e., physical-digital elements), but there
is currently limited work explicitly exploring how these games
can be designed to provide spatial affordances that implicitly
encourage collaboration. In this paper, we present a novel col-
laborative tangible game, titled Mad Mixologist, and investigate
how making a simple change in the location of game objects
within the tangible play space can lead to significantly different
collaborative interaction strategies. The results from our group
comparison study indicate that 1) players with exclusively direct
access to multiple relevant resources (i.e., a digital instruction
and a physical object) were more likely to assume responsibility
for completing tasks in a shared play space and 2) distributing
these same task-relevant resources across multiple players created
ambiguity over whether the player with the digital or physical
resource should engage with the shared play space. This study
demonstrates one possible way in which the physical design of a
tangible game can be arranged to implicitly encourage players
to develop more collaborative interaction strategies, specifically
by distributing exclusive resources across players. Overall, this
study highlights and reinforces the connection between spatial
affordances and social interactions via embodied facilitation
within the context of collaborative tangible games.
Index Terms—tangible games; collaborative play; interaction
strategy; embodied facilitation; spatial design; augmented reality
I. INTRODUCTION
Collaborative play is a desirable behavior in games that
require individuals to work together to achieve a common
goal [2], [35]. When designing a tangible game where players
interact with objects in a physical environment, it is important
to consider how the spatial configuration of those objects
within the play space can implicitly encourage players to adopt
a collaborative play style (as opposed to an independent or
competitive style). Eva Hornecker first established a frame-
work to guide the design of tangible interaction for collab-
orative use [13]. One of the four design themes—embodied
facilitation—is a critical concept highlighting that “tangible
interfaces/interaction systems embody facilitation methods and
means by providing structure and rules, both physically and
procedurally. Any application can be understood as offering
structure that implicitly directs user behavior by facilitating
some actions, and prohibiting or hindering others. It thus influ-
ences behavior patterns and emerging social configurations.
[13] Due to this inherent link between physical and social
affordances [14], it is especially important to consider how just
the spatial design of a tangible system can influence a person’s
interaction strategies both with objects and other people. As a
result, designers of physically collaborative games need to be
aware of how embodied facilitation can be used to encourage
players to engage in the intended collaborative behavior. While
existing frameworks are useful for guiding game design [7],
[14], it is rare to find empirical work explicitly demonstrating
how embodied facilitation can impact collaborative strategies
in tangible games [18]. This work aims to address this knowl-
edge gap.
In this paper, we present an empirical study that uses a
novel two-player tangible game—titled Mad Mixologist—to
examine how object placement in a tangible play space can
influence players’ natural tendency to collaborate. Tangible
play spaces can often be delineated into shared and personal
proxemic zones, where objects in the environment are either
physically accessible to a group of players or to a single
individual at a time, respectively. In this case, players’ physical
contributions to group tasks conducted in the shared space are
necessarily limited by their physical access to task-relevant
resources. The Mad Mixologist game is designed in a way
that delineates the play space into both shared (physical) and
personal spaces (one digital and one physical). For this paper,
we hypothesized that affording one player exclusive access
to both personal spaces at a time (i.e., a digital instruction
AND target object) would encourage them to develop 1) less
collaborative and 2) more consistent interaction strategies
than if each of the players was always given access to one
of the exclusive resources at a time (i.e., a digital instruction
OR target object). To test these hypotheses, we designed two
test conditions that either 1) provided one player at a time
with exclusive access to both personal spaces or 2) gave both
players alternating access to one of the personal spaces at a
time throughout the game. We then used video coding and978-1-6654-3886-5/21/$31.00 ©2021 IEEE
timestamp analysis techniques to quantify how much time
each player spent trying to interact with the physical game
objects in order to statistically compare the groups’ physical
interaction strategies. We conclude by discussing these group
comparison results and their implications for the design of
collaborative tangible games [15]. Evaluating the impact of
spatial affordances within the context of collaborative play
should help game developers and researchers understand the
potential of seemingly simple physical design choices to
generate drastically different interaction strategies.
II. RE LATE D WOR K
A. Tangible Games
Tangibles have a longstanding history in digital games—
such as the light gun that accompanied the Magnavox Odyssey,
the first commercial home video game console [4], [8]. These
tangible games are playful experiences created using a hybrid
of physical and digital interactive elements, and are becoming
more prevalent and sophisticated with modern emerging tech-
nologies such as AR, VR, and wearables [16], [17]. Notably,
the hybrid interfaces of tangible games can be designed in
a variety of unique ways depending on how the application
intersects with current technological capabilities (e.g., tabletop
interfaces or AR), which can lead to a diverse range of tangible
games [33]—such as hybrid board and card games [10], [19],
tangible tabletop games [24], [25], alternative controller games
[12], [30], hybrid AR/tangible games [1], [22], [23], and
hybrid VR tangible games [3].
B. Embodied Facilitation Guides Actions and Interactions
Hornecker’s design framework identifies four interrelated
themes: tangible manipulation, spatial interaction, embodied
facilitation and expressive representation. The present paper
focuses particularly on exploring the understudied connection
between spatial interaction and embodied facilitation. The
spatial interaction theme states that, because movement and
perception are tightly coupled, a person perceives their func-
tional role based on the spatial properties of the environment
(e.g., “why do I act?”). Embodied facilitation focuses instead
on describing how the properties of a tangible space can
be configured to guide—and even restrict—both individual
interaction strategies and emerging group behaviors (e.g.,
“how do I act?”) [6], [14]. Critically, Hornecker highlighted
that “the support of social interaction and collaboration might
be the most important ... feature of tangible interaction,
but this issue has attracted little explicit attention.” While
this observation was published over 15 years ago, it is still
rare to find tangible game studies that explicitly demonstrate
how embodied facilitation can impact collaborative interaction
strategies in a tangible play space [18]. This is especially true
with respect to the configuration of physical resources in the
tangible play space.
C. Facilitating Collaboration in Tangible Games
When a tangible space allows users to interact with the en-
vironment and each other in flexible ways, it affords users op-
portunities to perform the same activity using vastly different
interpersonal strategies (e.g., collaborative versus competitive
play styles) [20], [23], [31]. Conversely, a tangible space can
employ embodied facilitation (i.e., by guiding or restricting
user actions) to encourage a more specific interactive experi-
ence [14], [21], [34]. A prime example of embodied facili-
tation is when objects are spread far apart across a tangible
environment, forcibly limiting players’ physical interactions to
nearby sub-spaces [32]. Sub-spaces (whether social or phys-
ical) are an important characteristic of tangible environments
to consider because they create perceived boundaries between
people and/or objects [26], [27]. These imaginary boundaries
can then influence players’ perceived connections between
various elements of the interactive environment, indicating the
potential for sub-spaces to affect a user’s interaction strategy.
To date, various aspects of tangible games have been evalu-
ated in order to inform game design that drives specific inter-
personal behaviors like collaboration. Some of these studies
have examined the influence upon collaboration of things like
replacing a traditional keyboard/mouse interface with tangible
objects [23], adding tangible tools to a tabletop interface [29],
enabling an environment to be configurable [9], and changing
the shape of the surrounding space [5]. The current study
aims to add to this list of documented game design aspects by
evaluating how the configuration of objects within a tangible
play space, and consequently the distribution of exclusively
accessible resources between players, can significantly influ-
ence collaborative interaction strategies.
D. Definition of Collaboration
In evaluating Futura, an interactive tabletop game for
collaborative learning, Antle et. al. [2] defined collaborative
play as “a coordinated, synchronous activity that is the result
of a continued attempt to construct and maintain a shared
conception of a problem.” While providing design considera-
tions for collaborative board games, Zagal et. al. [35] added
that “collaborative players have only one goal and share the
rewards or penalties of their decisions.
III. DESIGN OF Mad Mixologist
Mad Mixologist is a tangible collaborative game that
utilizes augmented reality (AR) to swap the vision of its two
players. Upon donning the headsets, the players learn that
the main objective is to use the objects/ingredients arranged
on the table between them to mix a drink. Swapping the
players’ perspectives decouples their visual and proprioceptive
feedback (i.e., hand-eye coordination), which makes execution
of these “simple” tasks much more challenging, especially
if both players want to perform actions at the same time.
Figure 1 shows an example of how the game is set up
and what the players’ perspectives look like. Assuming the
players want to avoid making a mess (an implicit goal of the
game), the swapped perspective design necessitates at least
some collaborative effort because each player can only see
from the other’s field of view. This means that the player
pair must communicate with each other in order to give
feedback about the utility of the visual information provided
Fig. 1. Players (B) receive the others’ viewpoint as well as instructions through an augmented reality headset. The video that P1 (red headset) sees is shown
in A, and P2’s (purple) view is shown in C. The featured game is sampled from the Unmatched group, where the bell colors do not match the nearest headsets.
by the other player. Ultimately, the players are not required
to synchronize physical interactions, but this communication
requirement ensures that the game enforces collaboration.
In addition to swapping perspectives, the headsets are used
to present the players with a series of instructions (see Figure 1
for an example of how instruction presentation looks from the
two players’ points of view). For this study, the game’s design
was comprised of six instructions that gradually increase in
difficulty and culminate in the creation of a non-alcoholic
“Blushing Arnold Palmer” drink. Instructions are presented
in an alternating fashion to one player at a time (see Figure
2). Player 1 (P1), wearing the red headset, receives the odd
numbered instructions (i.e., 1, 3, 5), while Player 2 (P2),
wearing the purple headset, sees the rest (i.e., 2, 4, 6). The
player not currently receiving an instruction is instead provided
with a statement intended to encourage collaboration, i.e.,
“Help the other player! ASK THEM HOW!”
Fig. 2. The six instructions provided to each player during the game. The font
style of the instructions emphasizes the importance of color words—in this
case, the word “red” uniquely uses red, capitalized font, while the rest of the
font is a neutral blue. Each of the objects on the table are able to be identified
via corresponding distinct shape-color combinations such as GREEN pitcher,
SILVER cylinder, and so forth (see Figure 3).
The game starts when the red headset presents the first
instruction to P1: “Ring the RED bell to start.” The tasks listed
in the instructions can be divided into two distinct categories:
bell ringing and drink mixing. The first two instructions only
involve bell ringing and are considered “practice levels” for
those that come after, providing players with a single simple
task to help orient themselves to this new visual perspective
and form of collaboration. Each of the four remaining instruc-
tions involve ringing a bell as well, but only after completing
a more challenging drink mixing task. All of the drink mixing
objects are located in a shared interactive space (i.e., within
reach of both players). In contrast, one bell is placed within
each players’ personal space (i.e., out of their partner’s reach).
Fig. 3. Space definitions and object locations. One of two bells is located
within each player’s personal physical space. The drink mixing objects are
located within the shared physical space (top to bottom: purple cup, yellow
straw, silver cylinder, green pitcher, purple cylinder, blue straw, purple cup).
Aside from the player-specific constraints imposed by the
headsets and the location of objects in the play space, both
players are essentially afforded complete freedom to accom-
plish the instructions however possible. In other words, the
game does not provide any explicit restrictions to indicate
which player(s) should complete each task/instruction. This
observation is important because the design of tangible in-
teractive spaces can naturally facilitate (or restrict) users’
interaction strategies within that environment. When a game
does not assign players specific roles, they must instead infer
appropriate interaction strategies based on the physical and
social affordances imposed by the design of that interactive
space. Prior research has shown that working within a com-
pletely shared interactive space is effective for encouraging
cooperative experiences [23]. However, in addition to a
shared space, this game’s design also affords each player both
physical and digital forms of personal spaces (see Figure 3)
that function as exclusive access points to information (instruc-
tions) and objects (bells). Compared to an interactive space
that is completely accessible to both players, the inclusion
of personal spaces has the potential to alter or even disrupt
cooperative interaction strategies [28].
Importantly, each player’s physical personal space contains
only one of the bells, which implicitly encourages both players
to adopt distinct functional roles with respect to those objects.
Each player also has a digital personal space provided by the
headsets, where only one player at a time has visual access
to the current instruction (although players may convey the
instruction to their partner).
IV. MET HO DS
For this study, we created two versions of the Mad
Mixologist game that were identical except for the contents of
each player’s physical personal space. More specifically, we
switched the locations of the two bells between the two test
conditions—e.g., the Matched condition placed the red bell
next to the red player (P1), while the Unmatched condition
placed the red bell next to the purple player (P2). In doing
so, we aimed to evaluate how a simple change in the spatial
configuration of objects in a collaborative tangible game would
influence players’ interaction strategies. We hypothesized that
switching the positions of the two bells between the personal
spaces would cause the players to adopt significantly different
interaction strategies with respect to the objects in the shared
space.
A. Procedure
This study was approved by the Institutional Review Board
of the University of California, Santa Cruz. All testing was
conducted in-person during a series of video game playtesting
events hosted in a large event space on the university campus.
Two play conditions were tested and video recorded (one
condition per night): 1) the red bell was placed closer to P1
while the purple bell was placed closer to P2, and 2) the bell
positions were switched (see Figure 5). All other gameplay
conditions were the same between the two groups. For each
game, a member of the research team closely monitored
the play session. When the players successfully completed
a task and rang the correct bell, the researcher operated the
game’s control software to manually advance the game to the
next instruction. The game software documented these manual
commands by saving a timestamp file.
B. Participant Recruitment
Each participant was recruited during one of two game
playtesting events. All participants volunteered (i.e., they were
invited but not requested) to attend the event. Once at the
event, self-selected pairs of participants were recruited to play
the game by the research team. Before playing, participants
signed forms indicating that they agreed to participate in the
study and to be audio and video recorded during all gameplay.
They also completed a short demographics survey.
C. Definition of Interaction Strategy
While the Mad Mixologist game facilitates various forms
of interaction strategies, including both social (user-user)
and functional (user-object) behaviors, this paper focuses
exclusively on understanding how the design of the tangible
environment affects players’ functional interactions with the
physical game objects. More specifically, this study aims to
examine the consequences of a small change in the spatial
configuration of the tangible environment upon the amount of
time each player spends physically completing tasks. There-
fore, for the purpose of this study, the definition of an “interac-
tion” or “interaction strategy” will be restricted to behavioral
data involving players physically interacting with any of the
game objects (unless stated otherwise). The following section
explains how this specific type of interactive behavior was
identified and quantified for analysis.
D. Behavioral Data Coding
After the conclusion of the study, the recorded videos were
manually coded for behavioral data by one of the authors using
the BORIS event-logging software [11]. Every participant
received a unique identifier (e.g., G1-P1) during the coding
process. Actions were defined as “any movement of the upper
limbs” and were sub-categorized to be either interactive (i.e.,
functional) or gestural (i.e., emotive), and only interactive
actions were statistically analyzed. Interactive actions occurred
when a player clearly attempted to interact with any of
the game objects placed on the table. More specifically, a
START code indicated that a player started moving with
apparent intent to interact with an object, and a STOP code
indicated the functional conclusion of that movement (i.e., the
player returned the limb close to its original neutral position).
Due to feelings of disorientation, players frequently initiated
interaction with objects but ceased movement during the action
for extended periods of time (e.g., hovering over a bell or
holding a container). These periods of time were included as
part of the interaction whenever they clearly contributed to
task completion. “Incorrect” actions (e.g., ringing the wrong
bell) were also included as valid interactions.
E. Statistics
For each game, the total time spent completing each in-
struction was calculated using the timestamp data generated
by the game’s software. Each player’s total interaction time
for each instruction was calculated using the timestamp data
generated by the BORIS program. Proportion of play time
for each instruction was calculated by dividing a participant’s
total interaction time by the duration of time spent completing
the instruction. Because this paper focuses mainly on compar-
ing collaborative strategies between two test conditions, we
generated a “percent total effort” metric to indicate whether
one player spent more interaction time attempting to complete
an instruction compared to the other player. Each player’s
percent total effort for each instruction was calculated by
dividing their total interaction time by the sum of both players’
total interaction time for that instruction. For this metric,
a score of 0 or 100% for either player indicated that only
one of the players contributed to the total interaction time
for an instruction. Conversely, a score of 50% indicated that
both players contributed equal amounts of interaction time
for an instruction. This metric generated complementary data
between the two players (e.g., 75% effort for P1 implies 25%
for P2). Therefore, in order to simplify the presentation of
group comparison results, the statistics for percent total effort
data is only reported for P1. Finally, we created a metric
to assess the total proportion of time that each player spent
interacting with objects during the three instructions involving
the same bell. This “total percent interaction time” metric was
calculated by summing the total interaction time a player spent
on the three instructions involving the same bell divided by
the total duration of those instructions.
All data was evaluated using Matlab and is presented in the
text as mean ±standard deviation (SD). For each instruction
within each of the test conditions, non-parametric two-sided
Wilcoxon sign rank tests with a 95% confidence interval
were used to determine whether P1 on average contributed
significantly more or less than half (i.e., 50%) of the total
effort. Comparisons across test groups were conducted using
non-parametric two-sided Wilcoxon rank sum tests with a 95%
confidence interval.
V. RE SU LTS
A. Participants
A total of 32 participants were recruited for this study (21
male, 9 female, 2 non-binary). Eighteen participants (9 pairs)
played the game on the first round of testing (i.e., “Matched”
bell-headset color condition), and fourteen (7 pairs) played
on the second (i.e., “Unmatched” condition). The two groups
were comparable in age (Unpaired t-test: Matched = 25.9±3.2,
Unmatched = 28.1±7.0, p = 0.25) and prior experience playing
AR/VR games (Wilcoxon rank sum test: p = 0.46).
B. Group Comparison of Level Completion Times
We compared the average length of time that each test
group spent completing the whole game and each of the six
instructions (see Figure 4). With one exception, the Matched
and Unmatched groups spent statistically comparable amounts
of time completing the individual instructions (first instruc-
tion: p = 0.03; other instructions: p >0.07). However, the
Unmatched group on average always took longer to complete
each instruction. Consequently, we found that the Unmatched
group spent significantly more time playing the entire game
than the Matched group (Wilcoxon rank sum test: p = 0.03).
In order to control for these differences in play duration, the
rest of this study’s results compare the amount of interaction
time that P1 spent with respect to P2 within each game—
creating a proportion of total effort for each pair with respect to
each instruction. As such, the proportion of total effort metric
describes the players’ interaction strategies in a way that is
independent of the total play time.
C. Matched Group Interaction Strategies
All data and statistical tests comparing the Matched players’
average percent total effort to 50% are reported in Table
I (see Matched columns). In the Matched game condition,
P1 received three instructions involving the red bell, which
was located in that same player’s personal space. A series
of Wilcoxon sign rank tests revealed that, for two of these
instructions, P1 on average provided significantly more than
Fig. 4. Duration (in minutes) player pairs took to complete the whole game
(left) and each instruction. Error bars represent group standard deviation.
Small dots represent individual games. * indicates p <0.05.
half (50%) of the total effort that both players spent com-
pleting the tasks. While the third red bell instruction was not
significant, P1 still spent on average more than three times the
amount of effort as P2. These results indicated that when P1
was afforded access to both the red bell and its corresponding
instructions, that player was significantly more likely to spend
effort towards completing the tasks in those instructions.
We then conducted the same analysis for the Matched
P2, who received the other three instructions that instead
involved the nearby target purple bell. A series of Wilcoxon
sign rank tests showed that P2 consistently spent significantly
more than half the total effort needed to complete each of
these instructions. This pattern, combined with that from
P1’s red bell results, strongly suggested that both players
on average chose to adopt similar complementary interaction
strategies throughout the game (i.e., the players “took turns” to
independently complete “their” instructions). Altogether, these
results provided convincing evidence that the players with
simultaneous access to both the physical and digital personal
spaces were consistently significantly more likely to interact
with the any of the task-relevant objects. Even though most
of the interactions occurred in a shared space, both players on
average reliably opted to take responsibility for their “own”
tasks.
D. Unmatched Group Interaction Strategies
Statistical data for the Unmatched participants is also pro-
vided in Table I. In this test condition, P1 always received
instructions about the red bell, but this target bell was inacces-
sible to P1 because it was instead positioned in P2’s personal
space. We found that the Unmatched P2 was on average more
likely to interact with objects listed in any of the three in-
structions involving the red bell. However, Wilcoxon sign rank
tests showed that only one of these results was significantly
different from 50%. These results indicated that the player
next to the target bell was on average consistently more likely
to spend time trying to complete any tasks involving the target
red bell, even though all these instructions were only provided
to P1 (who necessarily communicated them to P2).
The instructions involving the purple bell were provided
only to P2, but that target bell was located next to P1.
The Unmatched P1 assumed the majority of responsibility
for trying to complete any instructions involving the target
Fig. 5. Interaction strategy diagrams and corresponding average percent total effort data for players under each test condition. After the first two “practice
levels” (in which both of the players located their respective bells), the players did not typically require much time or effort to repeat the bell ringing task.
Consequently, the vast majority of the percent effort that each player provided in the final four levels was spent trying to interact with the novel objects in
the shared play space. For both groups, the player with the target bell was consistently (i.e., for each instruction) on average more likely to complete tasks
in the shared space. However, the Matched group (A) developed consistent (i.e., statistically significant) interaction strategies for more instructions than the
Unmatched group (B). Error bars are not shown, but total effort SDs are provided in Table I. The total length of the bar(s) on either side of the zero mark
represents the group average for a player’s percent total effort, where the total bar length for both players together always equals 100%. Levels in which
either player on average spent significantly more than half the total effort are marked as * <0.05. Significance is determined using Wilcoxon sign rank tests.
purple bell, complementing the results witnessed for their
partner (P2) on the red bell tasks. However, Wilcoxon sign
rank tests again revealed that only one of these results was
significantly different from 50%. Altogether, the results from
the participants in the Unmatched group indicated that the
player with physical access to the target bell was more likely
to try to complete any instructions that included that bell, even
though they did not have direct access to those instructions.
E. Group Comparison of Interaction Strategies
Switching the position of the bells was the only difference
between the Matched and Unmatched game design conditions.
To determine whether P1’s interaction strategy for this task
changed along with the position of the bell, we compared P1’s
percent total effort between the two test conditions. A series of
Wilcoxon rank sum tests revealed that the percent total effort
that P1 spent for each of the six instructions was significantly
different between the two test conditions (Table I; all p<0.05
significance level). Overall, these group comparison results
indicated that players’ interaction strategies were consistently
significantly different between the two test conditions.
VI. DISCUSSION AND DESIGN IMPLICATIONS
A growing variety of tangible games have been shown to
benefit spatial reasoning and learning by leveraging theories of
embodied interaction [6], [18]. The vast majority of these stud-
ies demonstrate a diverse range of advantages tangible games
can offer compared to more traditional interactive formats,
but there are comparatively few experiments exploring how
the finer aspects of a system can be designed to encourage
specific interactive behaviors—especially collaboration. This
TABLE I
MEA N AND S TANDA RD D EVI ATIO NS FO R GRO UP AVE RAGE P ER CEN T
TOTAL E FFO RT PROV IDE D BY P1, WITH SIGNIFICANT DIFFERENCES BASED
ON WI LCOX ON S IGN R AN K (50%) OR RA NK S UM TE ST S (SI G).
P1 Percent Total Effort
Matched Unmatched Sig
Instruction µ σ 50% µ σ 50% p
1-P1: Red bell 77 29 0.03 39 36 0.55 0.03
2-P2: Purple bell 11 22 0.008 100 0 0.02 0.0002
3-P1: Task1 + Red bell 84 30 0.03 40 43 0.36 0.03
4-P2: Task2 + Purple bell 11 29 0.008 70 40 0.11 0.001
5-P1: Task3 + Red bell 70 37 0.12 17 29 0.03 0.02
6-P2: Task4 + Purple bell 17 35 0.02 75 35 0.08 0.002
paper establishes empirical evidence to support prior claims
of the importance of spatial design to implicitly encourage
collaborative interactions through embodied facilitation [14].
Max Mixologist does not provide any explicit restrictions
to indicate which player(s) should complete the tasks listed
in each instruction. Aside from the game’s design imposing
digital and physical personal spaces (i.e., exclusive access to
an instruction and/or target bell, respectively), the players had
complete freedom of choice in their interaction strategies. Ul-
timately, we observed that the novelty of the game’s swapped
perspective design made physical collaboration (i.e., simul-
taneous execution of any tasks located in the shared space)
especially challenging, which resulted in the players from both
test groups typically developing variations of “turn-based”
interaction strategies. Due to these game design features, we
expected embodied facilitation to implicitly guide players’
interactions with the objects and with each other. We compared
interaction strategies that resulted from two versions of the
game that were identical except for swapping the positions
of the two bells. For the Matched condition, the exclusively
accessible resources were only afforded to one player at a
time (i.e., the instruction AND target bell). For the Unmatched
condition, both of the players had exclusive access to only one
of these resources at a time (i.e., the instruction OR target bell).
Using this experimental design, we tested two hypotheses:
Hypothesis 1: Distributing the exclusively accessible
resources between the players would encourage more
interpersonal collaboration.
Hypothesis 2: Concentrating the exclusively accessible
resources to one player at a time would result in more
consistent interaction strategies.
A. Distributing Exclusive Resources Facilitates Collaboration
In the Matched group, the game afforded exclusive access
to both the instructions and the target bell for only one of the
players at a time (see Figure 5A). For the first two instructions
that only involved bell-ringing, each player was informed to
execute a task that only they could complete. In this way, the
game’s design set a standard by implying to the players that
they were not required to involve their partner to complete
“their” instructions. Correspondingly, both Matched players
overwhelmingly opted to continue this trend in subsequent
instructions by taking responsibility for any tasks that were
paired with their nearby bell, even though those extra tasks
were located in the shared space that was accessible to both
players. In contrast, each of the Unmatched players always
either had access to only the instruction or the target bell (see
Figure 5B). This design meant that the game always required
the players to communicate at least part of their instructions
(i.e., bell ringing tasks) to their partner. Similar to the Matched
group, we found that the Unmatched players most often opted
to take responsibility for completing any tasks in the shared
space that were paired with the bell in their personal space. In
other words, even though the player receiving the instruction
could have opted to complete the task in the shared space,
they tended to inform/allow their partner to do it instead.
These group trends indicate that the first two bell-ringing
practice levels “trained” both players to develop interaction
strategies that persisted for the remainder of the game. While
speech data was not reported in this study, these interaction
strategy results overall indicate that the Unmatched players
engaged in a greater level of interpersonal collaboration via
communicating instructions to the other player.
B. Concentrating Exclusive Resources Facilitates Consistent
Interaction Strategies
Regardless of the test condition, our results strongly indi-
cated that whichever player had access to the bell specified
in the instruction was consistently (i.e., for every instruction)
on average more likely to provide effort to complete the
entire instruction (see Figure 5). However, this trend was
almost always statistically significant for the Matched group,
and almost never so for the Unmatched group. Since both
groups tended to play the game in a turn-based fashion,
these results indicate that the Unmatched players were more
likely to divide execution of tasks for each instruction. More
specifically, the Unmatched player with the instruction would
occasionally opt to complete the drink mixing task in the
shared space, and then their partner would finish the instruction
by ringing the bell. Ultimately, distributing the exclusive
resources between the players afforded greater ambiguity over
which player should take responsibility for tasks in the shared
space. In comparison, the Matched players with the instruction
were more likely to complete tasks in the shared space, so
they consequently demonstrated a more consistent interaction
strategy both within and across individual games.
C. Design Implications
It can be difficult to generalize results between tangible
game designs because of the diverse possible design choices
and combinations of digital and physical elements [18]. How-
ever, this paper is relevant for tangible games that incorporate
multiple resources that are exclusively accessible to one player
at a time (e.g., a tabletop game with physical and digital
elements). This group comparison study offers an example
showing how distributing exclusively accessible resources,
whether real or digital, across multiple players can encourage
more collaboration between players. By extension, increasing
collaboration through embodied facilitation can also afford
players greater freedom of interpretation over how to collab-
orate, which can lead to greater variability in their chosen
interaction strategies. When creating a tangible game with
exclusively accessible components, designers should take care
to understand how the distribution of exclusively accessible
resources between players can implicitly guide their natural
tendencies to both communicate and physically collaborate.
D. Study Limitations
The consistency of the interaction strategies witnessed in
this study indicate that the observed turn-based patterns are
most likely real and not due to chance. Even so, it is important
to note that the sample sizes used in this study are small
(less than 10 games in each group), and the number of
games sampled for each group are not equal (although close).
Additionally, a mixed-method approach (i.e., interviewing
players about their experience) would have further bolstered
our quantitative analysis.
E. Future Work
Future work will include testing further interactions with
embodied facilitation, including modulating the location of
objects in either the shared or personal spaces. For example,
additional object configuration conditions can include making
the bells equidistant from both players or, alternatively, placing
some of the drink mixing objects within personal spaces. We
are also interested in statistically evaluating social aspects of
play. Comparing the amount of verbal communication between
groups would better indicate how object placement impacts
players’ social (rather than just physical) interaction strategies.
VII. CONCLUSION
When playful collaboration is desired, tangible game de-
signers can utilize principles of embodied facilitation to pur-
posefully shape both physical and interpersonal interaction
strategies. The results of this group comparison study pro-
vide an explicit example of how embodied facilitation can
implicitly guide both physical and interpersonal interaction
strategies within the context of a collaborative tangible game.
Our results highlight how making a subtle change in the spatial
configuration of exclusively accessible objects in a tangible
game environment can cause players to naturally adopt widely
different interaction strategies. This work provides implica-
tions for the importance of considering spatial design in a
tangible game environment and the potential of such choices
to drastically impact collaborative behaviors.
REFERENCES
[1] Sarah AlDakhil, Ebtehal Al Taleb, Munirah Al Ghamlas, and Shiroq
Al-Megren. Assessing the usability of a tangible educational game for
children. In 2nd International Conference on Computer Applications &
Information Security (ICCAIS), pages 1–5. IEEE, 2019.
[2] Alissa N. Antle, Allen Bevans, Theresa Jean Tanenbaum, Katie Seaborn,
and Sijie Wang. Futura: Design for collaborative learning and game play
on a multi-touch digital tabletop. In Proceedings of the 5th International
Conference on Tangible, Embedded, and Embodied Interaction (TEI),
pages 93–100, 2010.
[3] Jack Shen-Kuen Chang, Alison F. Doucette, Georgina Yeboah, Timothy
Welsh, Michael Nitsche, and Ali Mazalek. A tangible VR game designed
for spatial penetrative thinking ability. In Extended Abstracts of the 2018
CHI Conference on Human Factors in Computing Systems, pages 1–4,
2018.
[4] Alastair H. Cummings. The evolution of game controllers and control
schemes and their effect on their games. In The 17th Annual University
of Southampton Multimedia Systems Conference, volume 21. Citeseer,
2007.
[5] Jared Donovan and Robb Mitchell. Social contraptions and embodied
interaction. In Proceedings of the 12th ACM International Conference
Adjunct Papers on Ubiquitous Computing, pages 403–404, 2010.
[6] Paul Dourish. Where the action is: The foundations of embodied
interaction. MIT Press, 2004.
[7] Augusto Esteves. Designing tangible interaction for embodied facilita-
tion. In Proceedings of the 6th International Conference on Tangible,
Embedded and Embodied Interaction (TEI), page 395–396, 2012.
[8] Nick Evanson. Controlling the universe. Physics Education, 39(5):407–
414, 2004.
[9] William Farr, Nicola Yuill, Eric Harris, and Steve Hinske. In my
own words: Configuration of tangibles, object interaction and children
with autism. In Proceedings of the 9th International Conference on
Interaction Design and Children (IDC), pages 30–38, 2010.
[10] Ylva Fernaeus and Jakob Tholander. Finding design qualities in a
tangible programming space. In Proceedings of the SIGCHI Conference
on Human Factors in Computing Systems, pages 447–456, 2006.
[11] Olivier Friard and Marco Gamba. BORIS: A free, versatile open-source
event-logging software for video/audio coding and live observations.
Methods in Ecology and Evolution, 7(11):1325–1330, 2016.
[12] Katelyn M. Grasse and Edward F. Melcer. Generation: A novel
fabrication game for simulating evolution and natural selection. In
Extended Abstracts of the 2020 Annual Symposium on Computer-Human
Interaction in Play (CHI Play), pages 38–42, 2020.
[13] Eva Hornecker. A design theme for tangible interaction: Embodied fa-
cilitation. In Proceedings of the 9th European Conference on Computer-
Supported Cooperative Work (ECSCW), pages 23–43, 2005.
[14] Eva Hornecker and Jacob Buur. Getting a grip on tangible interaction:
A framework on physical space and social interaction. In Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems,
pages 437–446, 2006.
[15] Katherine Isbister and Florian “Floyd” Mueller. Guidelines for the
design of movement-based games and their relevance to HCI. Human–
Computer Interaction, 30(3-4):366–399, 2015.
[16] Hiroshi Ishii. Tangible bits: Beyond pixels. In Proceedings of the 2nd
International Conference on Tangible and Embedded Interaction (TEI),
pages 15–25, 2008.
[17] Hiroshi Ishii, D´
avid Lakatos, Leonardo Bonanni, and Jean-Baptiste
Labrune. Radical atoms: Beyond tangible bits, toward transformable
materials. Interactions, 19(1):38–51, 2012.
[18] Serena Lee-Cultura and Michail Giannakos. Embodied interaction and
spatial skills: A systematic review of empirical studies. Interacting with
Computers, 32(4):331–366, 2020.
[19] Regan L. Mandryk, Diego S. Maranan, and Kori M. Inkpen. False
prophets: Exploring hybrid board/video games. In Extended Abstracts
on Human Factors in Computing Systems (CHI), pages 640–641, 2002.
[20] Paul Marshall, Rowanne Fleck, Amanda Harris, Jochen Rick, Eva Hor-
necker, Yvonne Rogers, Nicola Yuill, and Nick Sheep Dalton. Fighting
for control: Children’s embodied interactions when using physical and
digital representations. In Proceedings of the SIGCHI Conference on
Human Factors in Computing Systems, pages 2149–2152, 2009.
[21] Edward Melcer and Katherine Isbister. Embodiment, collaboration, and
challenge in educational programming games: Exploring use of tangibles
and mouse. In Proceedings of the 12th International Conference on the
Foundations of Digital Games (FDG), pages 1–6, 2017.
[22] Edward F. Melcer, Victoria Hollis, and Katherine Isbister. Tangibles
vs. mouse in educational programming games: Influences on enjoyment
and self-beliefs. In Proceedings of the 2017 CHI Conference Extended
Abstracts on Human Factors in Computing Systems, pages 1901–08,
2017.
[23] Edward F. Melcer and Katherine Isbister. Bots & (main) frames:
Exploring the impact of tangible blocks and collaborative play in an
educational programming game. In Proceedings of the 2018 CHI
Conference on Human Factors in Computing Systems, pages 1–14, 2018.
[24] Joan Mora-Guiard and Narcis Pares. “Child as the measure of all things”:
The body as a referent in designing a museum exhibit to understand the
nanoscale. In Proceedings of the 2014 Conference on Interaction Design
and Children (IDC), pages 27–36, 2014.
[25] Sara Price and Carey Jewitt. A multimodal approach to examining
‘embodiment’ in tangible learning environments. In Proceedings of
the 7th International Conference on Tangible, Embedded and Embodied
Interaction (TEI), pages 43–50, 2013.
[26] Jochen Rick, Amanda Harris, Paul Marshall, Rowanne Fleck, Nicola
Yuill, and Yvonne Rogers. Children designing together on a multi-touch
tabletop: An analysis of spatial orientation and user interactions. In
Proceedings of the 8th International Conference on Interaction Design
and Children (IDC), pages 106–114, 2009.
[27] Stacey D. Scott, M. Sheelagh T. Carpendale, and Kori M. Inkpen.
Territoriality in collaborative tabletop workspaces. In Proceedings of
the 2004 ACM Conference on Computer Supported Cooperative Work
(CSCW), pages 294–303, 2004.
[28] Stacey D. Scott, Karen D. Grant, and Regan L. Mandryk. System
guidelines for co-located, collaborative work on a tabletop display. In
Proceedings of the 8th European Conference on Computer-Supported
Cooperative Work (CSCW), pages 159–178, 2003.
[29] Tess Speelpenning, Alissa N. Antle, Tanja Doering, and Elise van den
Hoven. Exploring how tangible tools enable collaboration in a multi-
touch tabletop game. In IFIP Conference on Human-Computer Interac-
tion, pages 605–621, 2011.
[30] Anne Sullivan and Gillian Smith. Designing craft games. Interactions,
24(1):38–41, 2016.
[31] Mike Tissenbaum, Matthew Berland, and Leilah Lyons. DCLM frame-
work: Understanding collaboration in open-ended tabletop learning en-
vironments. International Journal of Computer-Supported Collaborative
Learning, 12(1):35–64, 2017.
[32] Aaron Toney and Bruce H. Thomas. Applying reach in direct manipu-
lation user interfaces. In Proceedings of the 18th Australia Conference
on Computer-Human Interaction: Design: Activities, Artefacts and En-
vironments, pages 393–396, 2006.
[33] Elise van den Hoven and Ali Mazalek. Tangible play: Research and
design for tangible and tabletop games. In Proceedings of the 12th
International Conference on Intelligent User Interfaces, volume 28,
pages 10–14, 2007.
[34] Peta Wyeth. How young children learn to program with sensor, action,
and logic blocks. The Journal of the Learning Sciences, 17(4):517–550,
2008.
[35] Jos´
e P. Zagal, Jochen Rick, and Idris Hsi. Collaborative games: Lessons
learned from board games. Simulation & Gaming, 37(1):24–40, 2006.
... The authors used hexagonal tiles (hex tiles) which can be tangibly rearranged at each game round to yield a desired workspace shape and configuration, allowing tabletop mobile robots to move continuously within each new workspace. Mad Mixologist is another tangible collaborative game that used augmented reality (AR) to swap the vision of its two players according to Grasse et al. (2021). Mad Mixologist works by wearing headsets, then players learn what the main objective of the game is, for example, to use the objects/ingredients arranged on the table between them to mix a drink. ...
... Mad Mixologist works by wearing headsets, then players learn what the main objective of the game is, for example, to use the objects/ingredients arranged on the table between them to mix a drink. Swapping the players' perspectives separate their visual and physical feedback (i.e., hand-eye coordination), which makes the execution of these "simple" tasks much more challenging, especially if both players want to perform actions at the same time (Grasse et al., 2021). Continuing, tangible gaming has evolved to the use of video-conferencing platforms to facilitate collaborative storytelling using physical objects as researched by Harley et al. (2022). ...
Article
Full-text available
TANKS is a tangible coding game that introduced coding to South African 10-14-year-old learners. It is aimed to initiate and stimulate learning using facilitators. Facilitators use untraditional teaching methods because they are influential on training success. Also, facilitators undertook face-to-face contact sessions, since TANKS requires the use of a smartphone and tokens. A phenomenological qualitative study design approach of six facilitators of a face-to-face interview using an adapted semi-structured questionnaire was used. All interviews were audio-taped, transcribed verbatim, and thematically analyzed using an explanation-building strategy applied. Results identified themes where each theme was further explained in the study. Limitations included small interviewee numbers because of a nationwide COVID-19 restrictive lockdown. The feed-forward learning flow model concluded that supportive school leadership by principals or teachers creates an academic coding culture. This culture will necessitate learners' competitive behavior by adopting a performance-based appraisal practice, to improve the coding interest and TANKS performance by learners.
Article
Full-text available
Embodied interaction describes the interplay between the brain and the body and its influence on the sharing, creation and manipulation of meaningful interactions with technology. Spatial skills entail the acquisition, organization, utilization and revision of knowledge about spatial environments. Embodied interaction is a rapidly growing topic in human-computer interaction with the potential to amplify human interaction and communication capacities, while spatial skills are regarded as key enablers for the successful management of cognitive tasks. This work provides a systematic review of empirical studies focused on embodied interaction and spatial skills. Thirty-six peer-reviewed articles were systematically collected and analysed according to their main elements. The results summarize and distil the developments concerning embodied interaction and spatial skills over the past decade. We identify embodied interaction capacities found in the literature review that help us to enhance and develop spatial skills. Lastly, we discuss implications for research and practice and highlight directions for future work. RESEARCH HIGHLIGHTS • Systematically reviewed 36 studies to identify aspects of embodied interaction and spatial skills convergence that have been the focus of publications between 2008 and 2018. • Assessed embodied interaction-based spatial skills interventions, paying specific attention to the employed technologies, targeted spatial skills and main thematic categorizations of the research questions. • Discuss three capacities of embodied interaction that might catalyse the development and enhancement of spatial skills engagement: namely, enrichment, transferability and convergent smart physicality.
Conference Paper
Full-text available
3D printers are becoming increasingly accessible to the average consumer, however their potential utility within games has yet to be fully explored. Integrating 3D printer fabrication technology within game design presents a novel means for engaging players and providing them with tangible representations of gameplay elements. This in turn could be employed to increase embodied gameplay and even embodied learning for the player. In this paper, we present a novel "fabrication game" designed to teach basic evolutionary concepts. In the game, players take turns physically assembling components 3D printed in real-time to iteratively evolve their creatures and observe the impact of their evolutionary choices on a digital population simulation. We discuss the potential of this game's unique design in leveraging real-time fabrication of tangibles to enhance a player's understanding of principles of evolution and natural selection.
Conference Paper
Full-text available
While recent work has begun to evaluate the efficacy of educational programming games, many common design decisions in these games (e.g., single player gameplay using touchpad or mouse) have not been explored for learning outcomes. For instance, alternative design approaches such as collaborative play and embodied interaction with tangibles may also provide important benefits to learners. To better understand how these design decisions impact learning and related factors, we created an educational programming game that allows for systematically varying input method and mode of play. In this paper, we describe design rationale for mouse and tangible versions of our game, and report a 2x2 factorial experiment comparing efficacy of mouse and tangible input methods with individual and collaborative modes of play. Results indicate tangibles have a greater positive impact on learning, situational interest, enjoyment, and programming self-beliefs. We also found collaborative play helps further reduce programming anxiety over individual play.
Conference Paper
Full-text available
While there are common design decisions in existing games for teaching Computer Science (single player puzzle based games for the touchpad/keyboard and mouse), recent work has suggested that alternative approaches such as collaborative play and physically embodied designs may also provide important benefits to learners. In order to explore how making interactions with an educational programming game more physically embodied could impact collaborative play, we created an educational programming game called Bots & (Main)Frames. We then conducted a preliminary study to examine if the level designs achieved desired challenge and explore how two versions of the game with different forms of physical embodiment/input (e.g., mouse vs. tangible programming blocks) impacted player interactions underlying collaboration. We found that game levels seem to provide desired increasing challenge, and that players often used the mouse and tangible programming blocks to aid communication/collaboration in distinctly different ways.
Conference Paper
Full-text available
Computer Science (CS) and related skills such as programming and Computational Thinking (CT) have recently become topics of global interest, with a large number of programming games created to engage and educate players. However, there has been relatively limited work to assess 1) the efficacy of such games with respect to critical educational factors such as enjoyment and programming self-beliefs; and 2) whether there are advantages to alternative, physically embodied design approaches (e.g., tangibles as input). To better explore the benefits of a tangible approach, we built and tested two versions of an educational programming game that were identical in design except for the form of interaction (tangible programming blocks vs. mouse input). After testing 34 participants, results showed that while both game versions were successful at improving programming self-beliefs, the tangible version corresponded to higher self-reports of player enjoyment. Overall, this paper presents a comparison between the efficacy of tangible and mouse design approaches for improving key learning factors in educational programming games.
Article
Full-text available
There is a growing understanding of the unique ways that tabletops support effective collaboration; however, this research mostly focuses on environments in which learners work towards a single shared goal. Underpinning this perspective, either implicitly or explicitly, is the theory that collaborative learning is a process of attaining convergent conceptual change. However, this model of collaboration may not apply to all scenarios where learners are working together. In particular, informal, open-ended exploratory environments support (and often promote) shared activities where the goal may not be for all participants to emerge with a single, shared understanding. There is increased interest in understanding the efficacy of designs that support (and encourage) learners to collaborate while seeking divergent goals, ideas, and conceptions. This paper advances a framework (Divergent Collaboration Learning Mechanisms - DCLM) for recognizing and coding collaboration and divergent learning in such environments. We apply the DCLM framework to an informal tabletop environment (Oztoc) as a means of highlighting how DCLM may reveal new productive interactions environments that support divergent forms of collaboration, mentorship, and learning. Analysis of participants’ interactions within Oztoc revealed that participants who have non-aligned goals can still productively collaborate, and in many cases can provide insight and feedback that would not be possible in shared-goal activities. We conclude with an examination of how open-ended exploratory environments can support communities of practice and legitimate peripheral participation, and the importance of divergent inquiry and divergent conceptual change across a range of learning environments.
Article
Full-text available
Quantitative aspects of the study of animal and human behaviour are increasingly relevant to test hypotheses and find empirical support for them. At the same time, photo and video cameras can store a large number of video recordings and are often used to monitor the subjects remotely. Researchers frequently face the need to code considerable quantities of video recordings with relatively flexible software, often constrained by species-specific options or exact settings. BORIS is a free, open-source and multiplatform standalone program that allows a user-specific coding environment to be set for a computer-based review of previously recorded videos or live observations. Being open to user-specific settings, the program allows a project-based ethogram to be defined that can then be shared with collaborators, or can be imported or modified. Projects created in BORIS can include a list of observations, and each observation may include one or two videos (e.g. simultaneous screening of visual stimuli and the subject being tested; recordings from different sides of an aquarium). Once the user has set an ethogram, including state or point events or both, coding can be performed using previously assigned keys on the computer keyboard. BORIS allows definition of an unlimited number of events (states/point events) and subjects. Once the coding process is completed, the program can extract a time-budget or single or grouped observations automatically and present an at-a-glance summary of the main behavioural features. The observation data and time-budget analysis can be exported in many common formats (TSV, CSV, ODF, XLS, SQL and JSON). The observed events can be plotted and exported in various graphic formats (SVG, PNG, JPG, TIFF, EPS and PDF).
Conference Paper
Longitudinal and large-scale studies have shown a strong link between spatial abilities and success in STEM learning and careers. However, many spatial learning materials are either still 2D-based, or not optimized to engage students. In addition, research in spatial cognition distinguishes many spatial abilities as independent from one another. This implies the need to design content and interactions that target individual spatial abilities in order to further research and support training. Following this approach, we describe the design and implementation for a VR system, "Keep the Ball Rolling", controlled with tangible inputs and played as a multi-level game. The system is designed around one central spatial ability: penetrative thinking, which is important in geology, biology, human anatomy, and dentistry among other areas.