How Augmented Reality Enables Conceptual Understanding of Challenging Science Content

Abstract

Research on learning about science has revealed that students often hold robust misconceptions about a number of scientific ideas. Digital simulation and dynamic visualization tools have helped to ameliorate these learning challenges by providing scaffolding to understand various aspects of the phenomenon. In this study we hypothesize that students acquire a more accurate understanding of the Bernoulli's principle, a challenging science concept, by interacting with an augmented reality (AR) device. We show that even given a short period for investigation in a science museum, students in the AR condition demonstrate significantly greater gains in knowledge over students in the non-AR condition. Through interview responses, we further show that the AR affords greater ability to visualize details and hidden information to help students learn the science.

Full text

Yoon, S., Anderson, E., Lin, J., & Elinich, K. (2017). How Augmented Reality Enables Conceptual Understanding of
Challenging Science Content. Educational Technology & Society, 20 (1), 156–168.
156
ISSN 1436-4522 (online) and 1176-3647 (print). This article of the Journal of Educational Technology & Society is available under Creative Commons CC-BY-ND-NC
3.0 license (https://creativecommons.org/licenses/by-nc-nd/3.0/). For further queries, please contact Journal Editors at ets-editors@ifets.info.
How Augmented Reality Enables Conceptual Understanding of
Challenging Science Content
Susan Yoon1*, Emma Anderson1, Joyce Lin2 and Karen Elinich3
1Graduate School of Education University of Pennsylvania, Philadelphia PA, USA // 2Knowles Science Teaching
Foundation, Moorestown, NJ, USA // 3The Franklin Institute, Philadelphia, PA, USA // yoonsa@gse.upenn.edu
// emmaa@gse.upenn.edu // joyce.lin@kstf.org // kelinich@fi.edu
*Corresponding author
(Submitted November 19, 2015; Revised January 28, 2016; Accepted July 28, 2016)
ABSTRACT
Research on learning about science has revealed that students often hold robust misconceptions about a
number of scientific ideas. Digital simulation and dynamic visualization tools have helped to ameliorate
these learning challenges by providing scaffolding to understand various aspects of the phenomenon. In this
study we hypothesize that students acquire a more accurate understanding of the Bernoulli’s principle, a
challenging science concept, by interacting with an augmented reality (AR) device. We show that even
given a short period for investigation in a science museum, students in the AR condition demonstrate
significantly greater gains in knowledge over students in the non-AR condition. Through interview
responses, we further show that the AR affords greater ability to visualize details and hidden information to
help students learn the science.
Keywords
Augmented reality, Challenging science content, Bernoulli’s principle
Introduction
Research on learning about science has revealed that students often hold robust misconceptions or naïve
conceptions about a number of scientific ideas (Chi, 2005; Bransford, Brown, & Cocking, 1999). For example,
studies on student understanding of the Bernoulli’s principle, which is the subject of our exploration, have shown
that students find learning the content challenging due to, among other things, the counterintuitive experiences of
pressure-related events observed in the real world (Stepans, 2003).
Digital simulation and dynamic visualization tools have helped to ameliorate these learning challenges by
providing scaffolding (Honey & Hilton, 2011; Kim & Hannafin, 2011) to understand various aspects of
phenomenon that may contribute to misconceptions. Related to this, a recent focus in the learning sciences has
investigated how augmented reality (AR) tools can support science learning (Dunleavy, Dede, & Mitchell, 2009;
Dunleavy & Dede, 2014; Klopfer & Squire, 2008). At its simplest, augmented reality describes systems that
integrate computer-generated virtual elements or information (known as “digital augmentations”) with the real
world environment (Zhou et al., 2008). By superimposing virtual elements onto the real world environments, AR
allows users to experience and perceive the newly incorporated information as part of their present world,
thereby enhancing their perception of the real world (Kirkley & Kirkley, 2004; New Media Consortium, 2012).
Everyday examples of AR include Google Effects in Hangouts, AR games for Nintendo 3DS, and Webcam
Greeting cards from Hallmark.
Over the last 4 years, our project, Augmented Reality for Interpretive and Experiential Learning (ARIEL), has
investigated optimal uses of AR in science museums (e.g., Yoon, Elinich, Wang, Steinmeier, & Tucker, 2012a;
Yoon, Elinich, Wang, Van Schooneveld, & Anderson, 2013; Yoon & Wang, 2014), where misconceptions about
science are rarely addressed. In this study, we hypothesized that the use of AR, because it provides a
visualization of the underlying causal mechanisms, can assist students in developing a more accurate conception
of Bernoulli’s principle. We found that after participating in brief, informal investigations of the principle at a
science museum, students who interacted with an exhibit using AR were better able to understand the science
than students in a non-AR condition. Findings from our interviews and surveys suggest that the tool supported
students’ learning by revealing typically invisible features of the phenomenon.

157
Theoretical considerations
Common misconceptions and challenges associated with understanding Bernoulli’s principle
Bernoulli’s principle states that “when an incompressible, smoothly flowing fluid gains speed, internal pressure
in the fluid decreases, and vice versa” (Hewitt, 2004). In other words, there is an inversely proportional
relationship between fluid speed and pressure. When the fluid’s speed increases, the pressure drops. As it turns
out, this is a conceptually challenging and counterintuitive idea to understand for students, who typically believe
that when speed increases, so does the pressure (Faulkner & Ytreberg, 2011). Stepans (2003) explains,
Children learn from experience that when they blow on something – like a bubble or dandelion plume – it goes
away. These experiences make it difficult to make sense of the fact that when you blow on a surface, it comes
toward you, or that when you blow between things, they come together. These experiences make it difficult to
accept the concept of Bernoulli’s Principle. (p. 46)
In a test given to private and public 6th, 7th, and 8th grade Turkish students on the outcomes of discrepant events
related to Bernoulli’s principle, Bulunuz, Jarrett, and Bulunuz (2009) found that the majority of students held
incorrect conceptions of the phenomenon. Researchers gave a similar test to a group of pre-service elementary
teachers, and results indicated that less than 50% of the teachers gave correct responses (Bulunuz & Jarrett,
2009). Even physics teachers claim to have unclear understandings of Bernoulli’s principle and sometimes avoid
teaching the concept altogether to their students (Hewitt, 2004).
In examining misconceptions of pressure-related concepts, Basca and Grotzer (2001) organize the conceptual
challenges into four categories, two of which have important implications for our research. The first difficulty
that children have is that they tend to reason using obvious, rather than nonobvious, variables when determining
the causes of pressure-related events. For example, de Berg (1995) found that when a syringe was compressed,
17- and 18-year-old students felt, and accurately identified, that the pressure in the system increased. However, a
majority of them also thought that the enclosed air did not exert any pressure if the syringe was not being
compressed. In other words, children tend to associate pressure with movement; in the absence of detectable
movement, they assume that there is also no pressure (Glough & Driver, 1985). Similarly, Séré (1982), found
that 11- to 13-year-old children explained the movement of air in relation to another movement. Consequently if
a system was at equilibrium, they believed that there were no forces being exerted. Séré (1982) concluded,
[Children] thus lack the knowledge of atmospheric pressure as a state of reference in order to understand that
air—even when immobile—exists, is present, and acts. This state of reference is needed to recognize the effects
of pressure and to attribute a pressure to any quantity of air, even when it is not in movement. (p. 308)
The unobvious and undetectable nature of “stationary” air explains why students are not always aware that
pressure contributes to an effect. Therefore, they turn to more obvious and concrete but inaccurate explanations
(Kariotoglou & Psillos, 1993).
A second challenge that inhibits students’ understanding of pressure-related concepts is that they tend to reason
linearly rather than systemically and relationally (Basca & Grotzer, 2001). Children often attribute causes and
effects to a linear, unidirectional model, instead of considering more complex variables, and the relationships
between the variables, that might offer a better model of the scientific phenomenon (Grotzer & Tutwiler, 2014).
For example, Glough and Driver (1985) found that 12- to 16-year-old children explained drinking through a
straw as simply pulling or sucking rather than the result of a pressure differential between air pressure inside and
outside of the straw. Oftentimes, relational causality, which refers to the interaction between causes and effects,
more accurately explains scientific phenomena. As depicted in the second image of Figure 1, the cause of the
floating ball is the result of the interaction of and relationship between the two air pressures—not one (or two)
pressures acting disparately. Although a linear model is more conspicuous and accessible, it often does not fully
explain the complexity of the phenomenon (Basca & Grotzer, 2001).
These prevailing misconceptions and challenges that prevent children and adults from accurately understanding
pressure-related topics in the physical world motivate our study. Because Bernoulli’s principle is illustrated in
museums all over the world yet is a conceptually challenging topic to grasp, we hypothesized that the addition of
AR could help visitors build better knowledge of the science behind the floating ball. In the following section,
we describe previous studies of AR in science learning environments that show promising evidence to support its
use in improving science content learning.

158
Augmented reality to scaffold science learning
Augmented reality (AR) technologies have been highlighted for their enormous potential to enable people to
construct new understanding (New Media Consortium, 2014). By layering digital displays (known as “digital
augmentations”) over real-world environments, the hybrid display of phenomena provides scaffolds for users to
experience and perceive virtual elements as part of their present world (New Media Consortium, 2014; Kirkley
& Kirkley, 2004). In so doing, the augmentations help users explore aspects of the world in more concrete ways
than might otherwise be possible (Yoon & Wang, 2014).
This potential to augment users’ interactions, engagement, and experiences has revealed numerous affordances
of AR for science learning. These include supporting students’ scientific spatial ability, by (a) allowing them to
manipulate and learn content in three-dimensional perspectives (Kerawalla, Luckin, Seljeflot, & Woolard, 2006;
Martín-Gutiérrez et al., 2010); (b) engaging them in scientific inquiry by encouraging them to make
observations, ask questions, collaborate with others, and investigate and interpret data (e.g., Dunleavy et al.,
2009; Rosenbaum, Klopfer, & Perry, 2007; Squire & Jan, 2007; Squire & Klopfer, 2007); and (c) enhancing their
conceptual understanding by enabling them to visualize invisible or abstract concepts or events (e.g., Clark,
Dunser, & Grasset, 2011; Dunleavy et al., 2009; Dunleavy, 2014). For further descriptions of the features and
affordances of AR for educational purposes, see Cheng and Tsai (2013), Wu, Lee, Chang, and Liang (2013), and
Dunleavy and Dede (2014). These studies demonstrate that, compared with traditional teaching methods,
students who use AR applications tend to demonstrate higher academic achievement levels (Ibañez, Di Serio,
Villarán, & Delgado Kloos, 2014; Kamarainen et al., 2013; Lin, Duh, Li, Wang, & Tsai, 2013).
AR technology is also starting to slowly extend into museum spaces. However, as most of these technologies are
prototypes and still in the development stages, research on their use in museums is largely concerned with their
design, evaluation, and usability (Bell, Lewenstein, Shouse, & Feder, 2009). Some studies have investigated the
development of guidebooks to support visitors’ navigation of AR displays and their interactions with the displays
throughout the museum (e.g., Damala, Cubaud, Bationo, Houlier, & Marchal, 2008; Szymanski et al., 2008),
while others have studied the technological design, architecture, and implementation of an AR system (e.g.,
Koleva et al., 2009; Wojciechowski, Walczak, White, & Cellary, 2004). Although these studies do not
specifically examine the impacts on visitor learning, they do offer important insight into the general effects AR
has on visitors’ behavior. For instance, Asai, Sugimoto, and Billinghurst (2010) reported that an AR lunar surface
navigation system implemented at a science museum exhibit encouraged more collaborative interactions
between parents and their children. Szymanski and colleagues (2008) revealed that electronic guidebooks
increased visitors’ exploration of the objects being augmented, and Hall and Bannon (2006) demonstrated that
children’s engagement and interest increased when they interacted with several museum artifacts that were
augmented.
There are only a few studies that look at museum visitors’ knowledge and use of AR. Chang et al. (2014)
investigated college students’ appreciation of art by comparing the use of an AR enhanced guide, an audio guide,
or no guide. Students who experienced the art museum through the AR enhanced guide showed greater art
appreciation compared to the audio and non-guide experience. The behavior and amount of time with the
paintings was not significantly different between the audio and AR guided students. The AR guide was credited
with having more easily digestible information compared to the audio guide due to the use of visuals. Similarly,
Sommerauer and Müller (2014) explored how AR contributed to visitors’ mathematics knowledge in a museum
mathematics exhibition. They found that visitors who interacted with the AR enhanced exhibit performed
significantly better on knowledge acquisition and retention tests.
These studies demonstrate that AR has the potential to support learning. From conveying spatial information
about scientific elements essential to understanding and visualizing phenomena to increasing collaboration and
engagement among its users, AR technology offers promise for transforming science learning. However,
particularly for informal environments such as science museums, more empirical research is needed to determine
whether and how AR supports visitors’ conceptual understanding of science ideas.
ARIEL studies in a science museum
Building on the research described in the previous section, over the last several years we have used the ARIEL
project to investigate how augmented reality and various forms of learning scaffolds can improve visitors’
scientific knowledge in an informal science museum setting. To date, three pre-existing exhibit devices have
been modified to include digital augmentations. These devices were selected by museum staff because of their

159
prevalence in science museums and centers worldwide. The first device, “Be The Path,” was augmented to show
the flow of electricity when visitors completed an open circuit with their bodies. The second device, “Magnetic
Maps,” was augmented to visualize the magnetic field surrounding two bar magnets. And the third device,
“Bernoulli Blower” (depicted in Figure 1), was augmented to feature the interactions between two types of air to
keep a plastic ball afloat. Jonassen and colleagues (1994) proposed that when investigating the role of media in
student learning we should examine the process of learning first, then the role that context plays in understanding
the kinds of cognitive tools and their affordances needed to support learning. The ARIEL project has conducted
research examining both the learning afforded by AR in the context of an informal learning environment. Our
previous research has shown that learning is largely influenced by collaboration among peers while using the AR
device (Yoon, Elinich, Wang, Steinmeier, & Van Schooneveld, 2012b), all the while preserving core aspects of
informal participation, such as self-directed experimentation (Yoon et al., 2013). In terms of the cognitive tools,
results from experiments with our first two augmented devices demonstrate that AR can increase conceptual
(content) understanding (Yoon et al., 2012a) and cognitive (theorizing) skills (Yoon et al., 2012b). For
understanding the affordances of the media we have shown that learning is supported through the device’s
dynamic visualization capabilities (Yoon & Wang, 2014).
Figure 1. Images of the Bernoulli blower device with digital augmentation

160
It is important to note that the ARIEL studies take place in a science museum where learning is characterized by
free choice, individual motivation, and open-ended playful exploration (Bell, Lewenstein, Shouse, & Feder,
2009). In this informal learning environment there is no set curriculum, there is no instructor and the only change
to the learning environment is the addition of AR to the Bernoulli Blower. Here the instructional method is
intertwined with the media of AR. It is this change in instruction enabled by the dynamic visualization of hidden
information in real time that this particular study aims to explore.
In this study, we examine how the digital augmentations in “Bernoulli Blower” can serve as a scaffold for
learning about Bernoulli’s principle. Briefly, the exhibit features a physical plastic ball that is able to float in
midair because it is caught between the fast moving air coming from a blower attached to the exhibit and the
slow moving air in the room. The digital augmentation is produced on a screen that depicts the fast moving air
through arrows that point diagonally up and curve around a real time image of the physical plastic ball. At the
same time the screen displays the slow-moving air from the room by depicting shorter arrows that point in and at
the real-time image of the plastic ball. Although the normal room air moves at a lower speed than the faster
moving blown air, the room air exerts greater pressure on the ball and is therefore able to keep the ball floating in
the stream of fast-moving air instead of being blown away. Thus the speed and pressure of flowing air are
inversely proportional.
Methods
Participants and context
This study was conducted at a large, well-established science museum in a northeastern U.S. city. The students
who participated in this study were selected by their teachers, who themselves responded to a mass email
invitation sent to middle school (6th to 8th grade; 11 years old to 14 years old) science teachers in the surrounding
area. In total, 58 students (41% male, 59% female) from five schools (three charter, two community public)
participated (see Table 1 for other demographic data). We specifically targeted students in this grade band
because the concept of air pressure is first introduced in 5th grade in our state’s standards; therefore, all students
would have some prior knowledge of the science concepts illustrated by the device. This study was embedded
within an all-day school field trip to the museum, and participating students were given free general admission to
the exhibits. The total amount of time it took to participate in the research was approximately 1 hour.
Table 1. Student demographic data
School
Economically disadvantaged
Percent non-white
Public A
75%
31%
Public B
83%
84%
Charter A
84%
99%
Charter B
82%
100%
Charter C
80%
100%
On the day before the students’ field trip, researchers went to the schools to collect consent forms and to
administer pre-intervention surveys of students’ knowledge of Bernoulli’s principle. On the day of the field trip,
each chaperoned group (assigned by the teacher, with roughly nine students in each group) was given a specific
time to report to the research area, a space commonly used for museum workshops and classes. (Outside of this
time slot, students were free to explore the museum per their teacher’s instructions.) When students arrived at the
research area, they were randomly assigned to one of two conditions: the non-AR condition (device with no
digital augmentation) or the AR condition 2 (device with digital augmentation). In groups of three, students were
invited to the research area and shown the device. Depending on the condition that the students were assigned to,
the computer screen, which displayed the augmentations, would either be turned on (AR condition) or off (non-
AR condition) and the red ball would be lying on the table. The students were told “see if you can make the red
ball float” and asked to play with it as if they had found it on the museum floor. After students signaled that they
were finished, they were individually asked a set of interview questions about their experience with the device.
Their responses were audio recorded and later transcribed. The day after the field trip, researchers went back to
the schools to administer post-intervention surveys of student knowledge. In total, the non-AR condition had 29
students (55% female, 45% male, 55% 7th graders, 45% 8th graders) who spent on average 11 minutes and 43
seconds interacting with the device. The AR condition had 29 students (62% female, 38% male, 66% 7th graders,
34% 8th graders) who spent on average 9 minutes and 41 seconds interacting with the device.

161
Data sources and analyses
Two qualitative data sets were collected, coded, and analyzed to determine how AR impacted students’
conceptual knowledge of Bernoulli’s principle.
Pre- and post-intervention surveys of student knowledge
The pre- and post-intervention surveys consisted of four multiple-choice (MC) questions and one open-ended
(OE) response question. These questions were constructed by a team of researchers and are modeled on similar
questions found in middle school science textbooks. Three of the MC questions could be considered near-
transfer questions, as the correct answers could be directly accessed from the exhibit device itself. The fourth
MC question could be considered a far-transfer question, as it asked students to select a real-world situation that
illustrated Bernoulli’s principle. The OE response question depicted a similarly constructed device using
common household materials and asked, Why do you think the plastic ball floats in the stream of fast-moving
air? The complete intervention survey can be found in the Appendix. Students’ responses were coded using a
previously validated categorization manual on a six-point Likert scale ranging from limited understanding (1) to
complete understanding (6) (Wang, 2014). Refer to Table 2 for a description of the levels of understanding. Two
analyses of covariance (ANCOVAs) were separately conducted for the MC responses and the OE responses. For
the MC responses, the independent variable in the ANCOVA was the condition students participated in and the
dependent variable was the students’ post-intervention MC scores. The covariate was the students’ pre-
intervention MC scores. Similarly for the ANCOVA on the OE responses, the independent variable was the
students’ condition and the dependent variable was the students’ post-intervention OE responses. The covariate
was the students’ pre-intervention OE scores.
Table 2. Levels of content understanding
Level
Description
Sample response
1 – Little
Understanding
Identifies the air from the blower as the only
factor making the ball float. May discuss
various aspects of the air that contribute to the
phenomenon, but the focus is on the air itself.
The air pressure, if you have too much air
pressure, it’ll just push it away. But if you
have just the right amount right in the line
of symmetry, it will stay right in place.
2 – Emergent
Understanding
Considers features of the ball that may impact
its ability to float. Discussion of the air from
the blower may or may not be expressly stated.
It was able to float because it was light and
it’s plastic so it has air already in it. Since
it’s light enough, the air pushing on it
won’t make it move around because it’s
not just solid. It has enough air in it to
make it move.
3 – Partial
Understanding
Acknowledges other sources of forces—forces
not associated with the air from the blower—
that are involved, such as gravity and the
normal air in the room. May describe how the
forces interact (e.g., pushing up/down) with
each other to keep the ball floating.
The air, the actual air, is pushing down I
think. And the one from the tube is going
up so then it’s making it float cause it was
pushing on the sides.
4 – Basic
Understanding
Begins to consider that the ball floats because
of differences in the pressure/force of the
various airs involved. These differences may
be described in the amount of pressure/force
exerted on the ball or simply identified as
high(er) or low(er) pressure. Although they
may incorrectly describe these differences,
they acknowledges that the interaction of the
varying amounts of pressures is integral to the
phenomenon. (Note: Simply acknowledging
that pressures are involved is not enough).
The air was pushing it up so it would stay
away from the tube and the air was also
going around it so it could stay stabilized.
And also like, slow moving air could
stabilize it too…but it wouldn’t push it that
far. It would just like…it wouldn’t make a
big impact.
5 – Adequate
Understanding
Considers both the speed AND pressure of the
airs involved understanding there is a
relationship between the speeds and pressures,
though the relationship may be incomplete or
inaccurate. May only address the speed and
pressure of just one kind of air.
So the air from the tube is pushing it up
and the air around it’s pushing it in to
make sure that it doesn’t fall off the air
flow that’s keeping it up...The slow air has
the lower pressure and the fast air has the
higher pressure.

162
6 – Complete
Understanding
Understands that the ball floats because the
high pressure from the slow-moving room air
keeps the ball in the low-pressure, fast-moving
air stream. Both kinds of airs must be explicitly
addressed. These responses cannot contain any
misconceptions.
The faster air was pushing it up and the
slower air had more…pressure so it was
pushing it up and down and keeping it in
the same spot so it couldn’t really go out of
the air circle. (About the slow air) It’s just
like…like not moving air has higher
pressure than really fast air.
Interviews
Post-intervention interviews were administered to understand how students interpreted the concept illustrated by
the device. Interview questions included the following:
What did you learn from playing with this device?
How did you learn this?
For the AR condition students we asked a follow up question that probed whether they felt the AR supported
their learning: Do you think you would have learned this without the digital augmentations?
Responses were qualitatively mined both for content understanding (using the same categorization manual that
we used for the OE post-intervention survey questions) and for the affordances of AR as a visualization tool
(Yoon & Wang, 2014).
For both the OE survey and interview data sets, two graduate students were trained on the categorization manual
and scored 20% of the OE survey data (12 out of 59), yielding 83% agreement (on 10 out of 12 responses). The
discrepancies on codes were negotiated until one code was assigned. After this step, one researcher coded the
rest of the OE survey and interview data.
Results
Table 3 shows that students in the AR condition scored significantly higher on the MC portion of the knowledge
survey than the non-AR students, F(1, 55) = 8.600, p = .005, effect size (Cohen’s d) = .802. However, the
difference in means between conditions was not found to be significant for the OE responses on the knowledge
survey, F(1, 55) = 2.679, p = .107.
Table 3. Mean scores on student knowledge of Bernoulli’s principle
Student knowledge
Mean Pre (SD)
Mean Post (SD)
N
MC Condition 1
1.517 (.726)
2.000 (.598)
29
MC Condition 2
1.793 (.789)
2.517 (.688)
29
OE Condition 1
1.862 (.785)
1.931 (1.193)
29
OE Condition 2
1.759 (.773)
2.276 (1.750)
29
Interview Condition 1
n/a
2.660 (.550)
29
Interview Condition 2
n/a
3.100 (.770)
29
Regarding students’ interview responses, an independent samples t-test showed that there was a significant
difference between the knowledge scores of students in the non-AR condition and students in the AR condition,
t(56) = -2.543, p = .014, effect size (Cohen’s d) = .658 where more AR condition students scored in the higher
levels of understanding. In the non-AR condition, 38% of the students had a Level 2 understanding (defined as
describing observations or listing objects or concepts presented), and 59% had a Level 3 understanding (defined
as identifying a relationship between two of three variables – air speed, air pressure, and the floating ball). In
contrast, in the AR condition (device with digital augmentation), although there was a similar frequency of
students with a Level 3 understanding, only 17% had a Level 2 understanding, and 21% reached a Level 4
(defined as identifying the involvement of both air speed and pressure) or 5 (defined as recognizing a
relationship between varying air speeds and pressures) understanding.
Collectively, these results suggest that the digital augmentation had a positive effect on students’ content
knowledge. A perusal of student responses illustrates how the AR influenced their understanding. For example,
one student (ID6) in the AR condition who scored a Level 5, said,

163
It helped you see the air currents that [were] coming from the tube and it helped you see the high pressure air
that was coming in from below and above. If air is moving quickly, it has low pressure. If it's moving slowly, it
has high pressure.
This student went on to explain that the activity was different from how they normally learned in school because
of the screen and the display where she could “experience what it was instead of reading about it in a textbook.”
She and two other girls also “tried to play a game” in which they had to “get the ball to move around without
completely cutting off the air current.” Here we can see that the student was able to build an accurate
understanding of the phenomena while at the same time engaging in self-directed experimentation, which is an
archetypal characteristic of informal participation.
Other student responses illustrated how the AR acted as a scaffold for more accurate understanding the
phenomenon. Students commented on the affordance of visualizing scientific details:
…what kind of air like how many, how much pressure do you need to put on it or what kind of pressure do you
need to put on it slower, quicker, faster. (ID186)
Students also commented on the affordance of visualizing hidden information:
That the ball was being caught in the air pressure between the outside and what’s coming from the tube… In
school, we don’t really have the, I would say the visual learning of it. We just picture it in our minds. (ID17)
These responses demonstrate the affordances of the AR as well as students’ ability to acquire an accurate
understanding of how Bernoulli’s principle works in the brief time they were exposed to the exhibit during their
museum visit.
Discussion
In this study, we hypothesized that augmented reality, because it enables the visualization of typically invisible
causal mechanisms that underlie complex phenomena, could be used as an effective scaffold to help visitors
learn about challenging scientific content in a museum. Through the multiple-choice portion of the knowledge
survey and interviews, our results indicate that students in the AR condition significantly improved in their
understanding of Bernoulli’s principle and showed greater gains compared with students in the non-AR
condition. That the results from the open-ended response portion of the knowledge survey did not show a
significant difference between conditions may be attributed to the fact that those kinds of surveys draw on
limited information while interviews provide the opportunity for students to be probed for deeper understanding.
Student interviews also showed that the AR served as a valuable learning scaffold by enabling students to
visualize scientific details, recognize and make sense of hidden information, and gain a more accurate
understanding of the science. Elsewhere we have documented similar dynamic visualization affordances of
another, conceptually less-challenging AR device in our ARIEL series of experiments (Yoon & Wang, 2014),
which points to the continuing validity and reliability of these results.
These results also support the viability of AR in challenging misconceptions of pressure-related concepts
concerning the undetectable nature of air movement (Kariotoglou & Psillos, 1993) and students’ inabilities to
recognize complex relationships between scientific variables (Basca & Grotzer, 2001). From the interviews, we
can see clear evidence of students’ reasoning accurately about the inverse relationship between fluid speed and
pressure in Bernoulli’s principle (Hewitt, 2004)—they recognized that slower moving air has higher pressure and
faster moving air has lower pressure. Students were also able to articulate how the two kinds of air worked
together to make the ball float, which attests to growth in understanding of the complex relationship of variables
(Basca & Grotzer, 2001). Furthermore, some students were able to identify the non-obvious influence of the
surrounding air pressure, which they would not have understood without the AR.
Our findings in this most recent study in our ARIEL series investigating the use of AR devices in museums (e.g.,
Yoon et al., 2012a) convince us that AR can be used to support learning in informal environments through
specific scaffolds. We have shown in this study that AR not only supports learning of science content but can
also support learning of very challenging science content during brief periods of exploration. We acknowledge
that our results, though positive, revealed only modest conceptual gains, with participants’ improving on average
less than a level on the knowledge survey and reaching just above partial understanding in the interviews.
Looking forward, we are considering additional ways to scaffold the experience to induce greater learning while
at the same time preserving the informal experience. Designing for content understanding will inevitably require
increased and multiple scaffolds (Reiser & Tabak, 2014), some of which are already common practices within

164
museums, such as grouping exhibits into clusters based upon conceptually related content. In our future work,
we will investigate various aspects of exhibit design, in addition to what we have learned about AR devices in
museums, to understand how content learning is best facilitated.
Acknowledgements
This research was funded by NSF DRL Grant #0741659.
References
Asai, K., Sugimoto, Y., & Billinghurst, M. (2010). Exhibition of lunar surface navigation system facilitating collaboration
between children and parents in science museum. In Proceedings of the 9th ACM SIGGRAPH Conference on Virtual Reality
Continuum and Its Application in Industry (pp. 119-124). New York, NY: ACM.
Basca, B., & Grotzer, T. (2001, April). Focusing on the nature of causality in a unit on pressure: How does it affect student
understanding? Paper presented at the Annual Meeting of the American Educational Research Association, Seattle,
Washington.
Bell, P., Lewenstein, B., Shouse, A. W., & Feder, M. A. (Eds.). (2009). Learning science in informal environments: People,
places, and pursuits. Washington, DC: National Academies Press.
Bransford, J. D., Brown, A. L., & Cocking, R. R. (1999). How people learn: Brain, mind, experience, and school.
Washington, DC: National Academy Press.
Bulunuz, M., & Jarrett, O. S. (2009). Undergraduate and masters students’ understanding about properties of air and the
forms of reasoning used to explain air phenomena. Asia-Pacific Forum on Science Learning & Teaching, 10(2), 1-20.
Bulunuz, M., Jarrett, O. S., & Bulunuz, N. (2009). Middle school students’ conceptions on physical properties of air. Journal
of Turkish Science Education, 6(1), 37-49.
Chang, K. E., Chang, C. T., Hou, H. T., Sung, Y. T., Chao, H. L., & Lee, C. M. (2014). Development and behavioral pattern
analysis of a mobile guide system with augmented reality for painting appreciation instruction in an art museum. Computers
& Education, 71, 185-197.
Cheng, K. H., & Tsai, C. C. (2013). Affordances of augmented reality in science learning: Suggestions for future
research. Journal of Science Education and Technology, 22(4), 449-462.
Chi, M. T. (2005). Commonsense conceptions of emergent processes: Why some misconceptions are robust. Journal of the
Learning Sciences, 14(2), 161-199.
Clark, A., Dunser, A., & Grasset, R. (2011). Interactive augmented reality coloring book. In Proceedings of the 10th IEEE
International Symposium on Mixed and Augmented Reality. doi:10.1109/3DUI.2012.6184168
Damala, A., Cubaud, P., Bationo, A., Houlier, P., & Marchal, I. (2008). Bridging the gap between the digital and the physical:
Design and evaluation of a mobile augmented reality guide for the museum visit. In Proceedings of the 3rd ACM International
Conference on Digital and Interactive Media in Entertainment and Arts (pp. 120-128). New York, NY: ACM Press.
de Berg, K. C. (1995). Student understanding of the volume, mass, and pressure of air sealed within a sealed syringe in
different states of compression. Journal of Research in Science Teaching, 32(8), 871-884.
Dunleavy, M. (2014). Design principles for augmented reality learning. TechTrends, 58(1), 28-34.
Dunleavy, M., Dede, C., & Mitchell, R. (2009). Affordances and limitations of immersive participatory augmented reality
simulations for teaching and learning. Journal of Science Education and Technology, 18, 7-22.
Dunleavy, M., & Dede, C. (2014). Augmented reality teaching and learning. In J. Spector, M. Merrill, J. Elen, & M. Bishop
(Eds.), The handbook of research for educational communications and technology (pp. 735-745). New York, NY: Springer.
Faulkner, B. E., & Ytreberg, F. M. (2011). Understanding Bernoulli’s principle through simulations. American Journal of
Physics, 79(2), 214-216.
Glough, E. E., & Driver, R. (1985). What do children understand about pressure in fluids? Research in Science &
Technological Education, 3(2), 133-144.
Grotzer, T, & Tutwiler, M. S. (2014). Simplifying causal complexity: How interactions between modes of causal induction
and information availability lead to heuristic-driven reasoning. Mind, Brain and Education, 8(3), 97–114.
doi:10.1111/mbe.12054

165
Hall, T., & Bannon, L. (2006). Designing ubiquitous computing to enhance children’s learning in museums. Journal of
Computer Assisted Learning, 22(4), 231-243.
Hewitt, P. (2004). Bernoulli’s principle: Understanding Bernoulli’s principle as it applies to aerodynamic lift. National
Science Teachers Association, 71(7), 51-55.
Honey, M., & Hilton, M. (2011). Learning science through simulations and games. Washington, DC: National Academies
Press.
Ibáñez, M. B., Di Serio, Á., Villarán, D., & Delgado Kloos, C. (2014). Experimenting with electromagnetism using
augmented reality: Impact on flow student experience and educational effectiveness. Computers & Education, 71, 1-13
Jonassen, D. H., Campbell, J. P., & Davidson, M. E. (1994). Learning with media: Restructuring the debate. Educational
Technology Research and Development 42(2), 31-39.
Kamarainen, A. M., Metcalf, S., Grotzer, T., Browne, A., Mazzuca, D., Tutwiler, M. S., & Dede, C. (2013). EcoMOBILE:
Integrating augmented reality and probeware with environmental education field trips. Computers & Education, 68, 545-556.
Kariotoglou, P., & Psillos, D. (1993). Pupils’ pressure models and their implications for instruction. Research in Science &
Technological Education, 11(1), 95-108.
Kerawalla, L., Luckin, R., Seljeflot, S., & Woolard, A. (2006). “Making it real”: Exploring the potential of augmented reality
for teaching primary school science. Virtual Reality, 10, 163-174.
Kim, M. C., & Hannafin, M. J. (2011). Scaffolding problem solving in technology-enhanced learning environments (TELEs):
Bridging research and theory with practice. Computers & Education, 56, 403-417.
Kirkley, S. E., & Kirkley, J. R. (2004). Creating next generation blended learning environments using mixed reality, video
games and simulations. TechTrends, 49(3), 42-89.
Klopfer, E., & Squire, K. (2008). Environmental detectives - the development of an augmented reality platform for
environmental simulations. Education Technology Research and Development, 56, 203-228.
Koleva, B., Egglestone, S., Schnädelbach, H., Glover, K., Greenhalgh, C., Rodden, T., & Dade-Robertson, M. (2009).
Supporting the creation of hybrid museum experiences. In Proceedings of the 27th International Conference on Human
Factors in Computing Systems 2009 (pp. 1973-1982). New York, NY: ACM Press.
Lin, T. J., Duh, H. B. L., Li, N., Wang, H. Y., & Tsai, C. C. (2013). An Investigation of learners’ collaborative knowledge
construction performances and behavior patterns in an augmented reality simulation system. Computers & Education, 68,
314-321.
Martín-Gutiérrez, J., Saorín, J. L., Contero, M., Alcañiz, M., Pérez-López, D. C., & Ortega, M. (2010). Design and validation
of an augmented book for spatial abilities development in engineering students. Computers & Graphics, 34(1), 77-91.
New Media Consortium. (2014). NMC Horizon Report: 2014 K-12 Edition. Austin, Texas: The New Media Consortium.
Reiser, B., & Tabak, I. (2014). Scaffolding. In K. Sawyer (Ed.), The Cambridge Handbook of the Learning Sciences (pp. 44–
62). New York, NY: Cambridge University Press. doi:10.1017/CBO9781139519526.005
Rosenbaum, E., Klopfer, E., & Perry, J. (2007). On location learning: Authentic applied science with networked augmented
realities. Journal of Science Education and Technology, 16(1), 31-45.
Séré, M. G. (1982). A Study of some frameworks used by pupils aged 11-13 years in the interpretation of air pressure.
European Journal of Science Education, 4(3), 299-309.
Sommerauer, P., & Müller, O. (2014). Augmented reality in informal learning environments: A Field experiment in a
mathematics exhibition. Computers & Education, 79, 59-68.
Squire, K., & Jan, M. (2007). Mad city mystery: Developing scientific argumentation skills with a place-based augmented
reality game on handheld computers. Journal of Science Education and Technology, 16(1), 5-29.
Squire, K., & Klopfer, E. (2007). Augmented reality simulations on handheld computers. Journal of the Learning Sciences,
16(3), 371-413.
Stepans, J. (2003). Targeting students’ science misconceptions: Physical science concepts using the conceptual change
model. Clearwater, FL: Showboard, Inc.
Szymanski, M., Aoki, P., Grinter, R., Hurst, A., Thornton, J., & Woodruff, A. (2008). Sotto voce: Facilitating social learning
in a historic house. Computer Supported Cooperative Work (CSCW), 17(1), 5-34.
Wang, J. S. (2014). The Impact of multiple dynamic visualizations on family children’s learning in a science museum
(Doctoral dissertation). Available from ProQuest Dissertation and Theses database. (Paper AAI3622675)
Wojciechowski, R., Walczak, K., White, M., & Cellary, W. (2004). Building virtual and augmented reality museum
exhibitions. In Proceedings of the 9th international conference on 3D web technology (pp. 135–144). New York, NY: ACM.

166
Wu, H. K., Lee, S. W. Y., Chang, H. Y., & Liang, J. C. (2013). Current status, opportunities and challenges of augmented
reality in education. Computers & Education, 62, 41-49.
Yoon, S. A., & Wang, J. (2014). Making the invisible visible in science museums through augmented reality devices.
TechTrends, 58(1), 49-55.
Yoon, S., Elinich, K., Wang, J., Van Schooneveld, J., & Anderson, E. (2013). Scaffolding informational learning in science
museum: How much is too much? Science Education, 97(6), 848-877.
Yoon, S., Elinich, K., Wang, J., Steinmeier, C., & Tucker, S. (2012a). Using augmented reality and knowledge-building
scaffolds to improve learning in a science museum. International Journal of Computer Supported Collaborative Learning,
7(4), 519-541.
Yoon, S., Elinich, K., Wang, J., Steinmeier, C., & Van Schooneveld, J. (2012b). Learning impacts of a digital augmentation in
a science museum. Visitor Studies, 15(2), 157-170.
Zhou, F., Duh, H., & Billinghurst, M. (2008). Trends in augmented reality tracking, interaction and display: A Review of ten
years in ISMAR. In Proceedings of the 7th IEEE/ACM International Symposium on Mixed and Augmented Reality (pp. 193-
202). Cambridge, MA: IEEE.

167
Appendix
Knowledge Test
1
What is the relationship between the speed and pressure of moving air?
a. The faster the speed is, the higher the pressure.
b. The slower the speed is, the lower the pressure.
c. The faster the speed is, the lower the pressure.
d. Speed and pressure are not related.
2
A) When the air blower begins to blow, what will happen to the ball?
a. It will blow away and fall to the ground.
b. It will stay floating in the stream of fast-moving air.
c. It will float for a few seconds and then blow away.
B) Which air puts MORE pressure on the ball?
a. The fast-moving air from the blower (solid line).
b. The slow-moving air around the ball (dotted line).
c. Both put the same amount of pressure (solid & dotted lines).
AIR
BLOWER
X
Key:
Solid line = fast-moving air
Dotted line = fast-moving air
(X) Circle = plastic ball

168
3
Look at the picture.
The electric hair dryer is blowing a stream of fast-moving air upwards.
A small, lightweight, plastic ball is floating in the stream of fast-moving air. It will continue to float there until
the hair dryer stops blowing.
Why do you think the plastic ball floats in the stream of fast-moving air?
4
Which of the following real-world scenarios illustrates the concept you just learned? (Circle all that apply).
a. running a race c. climbing a mountain
b. paper lifting off the desk due to a gust of wind d. ejecting out of a plane