Affordances of Augmented Reality in Science Learning:
Suggestions for Future Research
Kun-Hung Cheng
•
Chin-Chung Tsai
Published online: 3 August 2012
Ó Springer Science+Business Media, LLC 2012
Abstract Augmented reality (AR) is currently considered
as having potential for pedagogical applications. However,
in science education, research regarding AR-aided learning
is in its infancy. To understand how AR could help science
learning, this review paper firstly has identified two major
approaches of utilizing AR technology in science educa-
tion, which are named as image-based AR and location-
based AR. These approaches may result in different
affordances for science learning. It is then found that stu-
dents’ spatial ability, practical skills, and conceptual
understanding are often afforded by image-based AR and
location-based AR usually supports inquiry-based scientific
activities. After examining what has been done in science
learning with AR supports, several suggestions for future
research are proposed. For example, more research is
required to explore learning experience (e.g., motivation or
cognitive load) and learner characteristics (e.g., spatial
ability or perceived presence) involved in AR. Mixed
methods of investigating learning process (e.g., a content
analysis and a sequential analysis) and in-depth examina-
tion of user experience beyond usability (e.g., affective
variables of esthetic pleasure or emotional fulfillment)
should be considered. Combining image-based and loca-
tion-based AR technology may bring new possibility for
supporting science learning. Theories including mental
models, spatial cognition, situated cognition, and social
constructivist learning are suggested for the profitable uses
of future AR research in science education.
Keywords Augmented reality Science education
Spatial ability Practical skills
Conceptual understanding Inquiry-based learning
Introduction
In the past two decades, the applications of augmented
reality (AR) have been increasingly receiving attention.
Since the 1990s, several special issues on AR have been
published by journals such as Communications of the ACM
(1993), Presence: Teleoperators and Virtual Environments
(1997), Computers and Graphics (1999), and International
Journal of Human–Computer Interaction (2003). More-
over, according to the 2011 Horizon Report, AR, with its
layering of information over 3D space, creates new expe-
riences of the world. With these new prospects of infor-
mation access, the prevalent employment of AR has been
in marketing, social engagement, or entertainment (John-
son et al. 2011). In addition to these consumer uses, the
2011 Horizon Report also suggested that AR should be
adopted in the next 2–3 years to provide new opportunities
for teaching, learning, research, or creative inquiry. By
examining article publications on Google Scholar, Martin
et al. (2011) reported that AR is in its initial stage
according to its publication impact, and they have proposed
that it will probably have significant influences on educa-
tion in the future.
K.-H. Cheng (&)
Digital Content Production Center, National Chiao Tung
University, #1001, University Rd., Hsinchu 300, Taiwan
e-mail: kuhu@mail.nctu.edu.tw
K.-H. Cheng
Graduate Institute of Applied Science and Technology, National
Taiwan University of Science and Technology, Taipei, Taiwan
C.-C. Tsai
Graduate Institute of Digital Learning and Education, National
Taiwan University of Science and Technology, #43, Sec.4,
Keelung Rd., Taipei 106, Taiwan
e-mail: cctsai@mail.ntust.edu.tw
123
J Sci Educ Technol (2013) 22:449–462
DOI 10.1007/s10956-012-9405-9
As far as the development of educational technologies is
concerned, investigating how technology assists students’
learning is an important issue. Also, in science education,
researchers have continued to devote their efforts to
exploring technology-aided learning. In Linn’s (2003)
review, the technology of providing customizable envi-
ronments (e.g., a function of allowing users to graphically
organize their concept maps of scientific arguments) and
the development of visualization tools to enhance scientific
spatial understanding (e.g., earth structures or molecular
geometry) were highlighted as trends in science learning.
Similarly, a recent review also reported that computer
simulations which can visualize invisible phenomena and
provide opportunities of manipulating experimental vari-
ables have positive learning effects (Rutten et al. 2011). In
addition to its advantages in science education, Rutten
et al. (2011) indicated a potential issue regarding how
users’ immersion in computer simulation environments
contributes to learning effects. The notion of users’ per-
ceived immersion brings us to a consideration of the im-
mersive experiences afforded by AR technology. With
capability of infusing digital information throughout the
real world, AR technology could engage learners in an
immersive context along with authentic experiences to
make scientific investigations, collect data outside class-
room, interact with an avatar, or communicate face-to-face
with peers (e.g., Dunleavy et al. 2009). Therefore, it may
suggest that researchers should pay attention to how AR
technology could further help science education.
Since AR is considered as having potential for pedagog-
ical applications (Johnson et al. 2011), several studies have
probed its effects on science education regarding the issues
of conceptual change (Shelton and Stevens 2004), laboratory
work (Andu
´
jar et al. 2011), inquiry-based learning (Squire
and Klopfer 2007), scientific argumentation (Squire and Jan
2007), ecological preservation (Koong Lin et al. 2011), and
spatial ability (Martı
´
n-Gutie
´
rrez et al. 2010). The results of
these studies mostly showed learners’ positive attitudes
toward AR (e.g., satisfaction or perceived usefulness), and to
some extent, indicated improvement in student outcomes.
Due to the fact that educational research regarding AR-
aided learning is in its infancy (Martin et al. 2011), this
paper argues that it is worth understanding what role AR
technology may play in science learning and to further
identify future directions for AR-related study. Therefore,
the background and characteristics of the current AR
applications are firstly discussed in this paper. Because the
technology utilized in AR applications has been developed
for a period time and continues evolving, it is necessary
that the contemporary and emerging features of AR tech-
nology should be identified. With an overall understanding
of the features of this technology, the affordances of AR in
science learning are then explored.
Moreover, to pedagogically examine AR-related studies
on science education from a broader perspective, the
dimensions (i.e., learning concepts, technical features,
learner characteristics, interaction experience, learning
experience, learning process, and learning outcomes) of the
virtual reality (VR)-based learning model proposed by
Salzman et al. (1995) are adopted in this paper. Based on
these pedagogical dimensions, this paper finally proposes
suggestions for future research regarding AR-aided science
learning. In summary, based on a review of the literature,
the purposes of this paper are as follows:
1. To identify the current features of AR technology in
science education.
2. To understand the affordances of AR in science
learning.
3. To examine the focal point of the current research on
AR-related science learning.
4. To make suggestions for future science research
regarding AR-related learning.
Background of Augmented Reality
In the 1960s, the idea of virtual reality (VR) was initially
proposed by computer graphics pioneer Ivan Sutherland to
construct a synthetic environment through visualization
using a head-mounted device (Sutherland 1968). With the
growth of VR, in the 1990s, the term
augmented reality
(AR) was originated by scientists at the aircraft manufac-
turer Boeing, who were developing an AR system that
blended virtual graphics onto a real environment display to
help aircraft electricians with cable assembly (Caudell and
Mizell 1992). At about the same time during the early
1990s, several introductory applications of AR were pub-
lished, such as a surgical training program (Bajura et al.
1992) and a laser printer maintenance demonstration (Fe-
iner et al. 1993). Since AR evolved from and shared partial
commonalities with VR because of their computer-gener-
ated elements, in 1994, Milgram and Kishino presented the
concept of a virtuality continuum, as shown in Fig. 1,
defining environments consisting solely of physical objects
(e.g., a video display of a real-world scene) on the left, and
environments consisting solely of virtual objects (e.g., a
computer graphic simulation) on the right. Described as the
possibilities of the mixture of real-world and virtual-world
objects within a single display, mixed reality exists at any
point on this continuum and encompasses both augmented
reality and augmented virtuality (Fig. 1).
The definition of AR commonly adopted by relevant
studies is what Azuma (1997) described as a variation of
VR. While VR entirely immerses a user in a synthetic
environment, AR allows a user to see a real world with
450 J Sci Educ Technol (2013) 22:449–462
123
virtual elements overlapped upon it in real time. To avoid
limiting AR to specific technologies or required devices
such as head-mounted displays (HMD), Azuma (1997)
identified three characteristics of AR: (1) combines real
and virtual, (2) interactive in real time, and (3) registered in
3D. Recently, instead of emphasizing the 3D characteristic,
Klopfer (2008) proposed a spectrum describing AR with
lightly to heavily virtual information provided to users.
While light AR refers to experiencing a lot of physical
reality along with limited virtual information access, heavy
AR represents massive virtual information input in an
augmented environment. Several studies have also devel-
oped AR systems without 3D virtual information register-
ing in physical environments (e.g., Dunleavy et al. 2009;
Squire and Jan 2007; Squire and Klopfer 2007). In this
paper, it is hence considered that a current characteristic of
AR may be that it is not necessarily presented in 3D virtual
objects (or information).
The initial stage of AR development seemed to rely on
HMD-related devices for implementing research. The
HMD devices which are used to combine real-world and
computer-generated information are often categorized into
two forms: optical see-through and video see-through
(Azuma 1997). While the optical system superimposes
virtual images on a user’s view of the real world, video
systems blend computer graphics with camera images that
approximate what a user would normally see. For the
reason that the see-through HMD is deemed as a high-end,
expensive, or obtrusive device, such as an additional
backpack with computer apparatus, AR hardware charac-
terized by simplicity and portability may have greater
opportunities of widespread use. That is why mobile AR
applications have recently become popular on Google
search. Also, some studies regarding AR-related learning
(e.g., Ha et al. 2011) or AR games (e.g., Broll et al. 2008)
have suggested a future direction for developing a mobile
AR platform.
Features of Augmented Reality
While a variety of AR-related equipment is utilized, there
is a need to understand the current features of AR. For
example, the enabling technologies of computing hardware
(e.g., wearable PC, tablet-PC, or smart phone), software
architectures (e.g., Wireless and 3G networking), and
tracking and registration (e.g., GPS) for mobile AR have
been summarized in a previous literature review (Papagi-
annakis et al. 2008). In order to simply categorize the
present state-of-the-art developments in AR, the two types
of AR application reported in Pence’s (2011) study, namely
(1) marker-based and (2) markerless AR, could be a clas-
sification of AR for general acceptance. However, as an
infant AR technology of natural image recognition being
developed beyond artificial marker identification, it is
necessary that the definition of marker-based AR should be
enriched. Therefore, this paper attempts to re-coin the two
types of AR as (1) image-based and (2) location-based AR,
further offering a broader feature of AR applications.
Image-Based AR
Basically, marker-based AR requires specific labels to
register the position of 3D objects on the real-world image.
As presented in Fig. 2, an AR book with basic equipment
such as a webcam and marker labels is one of the typical
marker-based applications and has been employed in sev-
eral studies (e.g., Koong Lin et al. 2011; Martı
´
n-Gutie
´
rrez
et al. 2010). A marker label in typical marker-based
Fig. 1 Virtuality continuum defining the possibilities of the mixture of real-world and virtual-world objects (modified from Milgram and
Kishino’s 1994 study)
Fig. 2 The concept of an AR book modified from Koong Lin et al.’s
2011 study (p. 184)
J Sci Educ Technol (2013) 22:449–462 451
123
applications is commonly presented as an iconic coded
image (refer to Fig. 3). By detecting a marker label on a
book through a webcam capture, a virtual element is then
generated by the AR software. This virtual element would
be shown upon the book and recovered on the computer
screen that can be manipulated by tilting or rotating the
book. Moreover, setting with a projector in a traditional
classroom, students can operate the card with an AR
marker to control the 3D objects on a projected screen
(Nu
´
n
˜
ez et al. 2008) or a whiteboard (Kerawalla et al.
2006). Also, it is a trend that marker-based AR applications
with mobile devices are being developed, such as the An-
dAR project initiated by Google in 2011. With mobile
devices, the applications of AR would not be restricted on
the front of desktop computers.
Beyond the artificial labels identified in marker-based
AR, natural image recognition has been integrated into AR
technology. For example, in Ajanki et al.’s (2011) study,
the augmented information regarding an individual syn-
opsis could be overlaid on the display by recognizing the
image of a human face. Recently, several handheld appli-
cations (e.g., the junaio app) have been developed to show
commercial promotion information by recognizing graph-
ics. For example, when a restaurant poster designed with
actual beverage graphics is detected by a mobile camera, a
virtual object (e.g., a 3D beverage model) is then popped
on the mobile screen for the business purposes. In the
present paper, this detection process refers to natural
graphics recognition. Instead of marker identification,
graphic recognition has also been gradually applied in AR
because the graphics naturally fit human visual experi-
ences. To summarize, this paper concludes that artificial
markers and natural graphics recognition could be deemed
as a type of image-based AR feature.
Location-Based AR
In contrast to image-based AR, markerless AR uses posi-
tion data launched from mobile devices, such as a wireless
network or global positioning system (GPS), to identify a
location, and then superimposes computer-generated
information (as illustrated in Fig. 4). Several studies have
demonstrated location-aware AR educational games with
mobile devices. For example, McCall et al. (2011) devel-
oped a handheld AR game to immerse users in exploring
the history of a city via the assistance of position orien-
teering and augmented elements (e.g., characters, objects,
and buildings) on the scene. Popular mobile apps such as
Layer and Wikitude are also designed for discovering
augmented information around users (e.g., restaurant
information or scenic spots) by detecting their position.
Obviously, these AR applications share similarity in their
technical features, in that location-based augmented
information is shown on the users’ mobile screens in real
time, and can thus be generalized as a type of location-
based AR.
A Comparison of Image-Based and Location-Based AR
To clearly identify the similarities and differences between
image-based and location-based AR, Fig. 5 illustrates a
comparison. While the recognition of artificial labels or
natural graphics is the main feature of image-based AR,
GPS or a wireless network is used as the recognition
technique to register users’ positions and to offer them real-
time information in a location-based AR environment.
After the process of recognition, both features of AR
technology will add augmented assets (e.g., text, audio,
video, 3D model) to the physical elements on the users’
display. The use of these two categories might enhance
understanding of the features of AR applications, regard-
less of what hardware and software is used. Before dis-
cussing the affordances of AR in science learning based on
the two categories, a case of AR applications in science
education is presented in the following.
Fig. 3 Example of a marker label in image-based AR
Fig. 4 The concept of location-based AR modified from the design
of the Layer app
452 J Sci Educ Technol (2013) 22:449–462
123
A Case of AR Applications in Science Education
To clearly address the mechanism of AR technology and
how it supports science education, a case of image-based
AR learning activity by recent research is presented in this
section. In Martı
´
n-Gutie
´
rrez et al.’s (2010) study, an
image-based AR book system was established for engi-
neering graphics learning. The hardware settings consist of
a desktop computer, a webcam, and an AR book with
several marker labels on the pages. By detecting an iconic
marker on the AR book through the webcam, a 3D virtual
geometry object is shown upon the AR book and mounted
on the computer screen. The learning task is required the
participants to identify surfaces and vertexes on both
orthographic and axonometric views of the 3D object and
further practice sketch exercises for evaluating their spatial
ability. With the affordances of AR, the participants are
allowed to freely tilt or rotate the AR book to manipulate
the 3D object when they need to inspect it from different
perspectives. This operation of AR empowers learners to
interact with the 3D geometry objects without wearing
headsets and immediately exercise on the paper book to
reflect spatial concepts.
Affordances of AR in Science Learning
AR technology has been widely utilized due to its possi-
bilities in a variety of fields such as manufacturing, urban
design, museum exhibitions, or clinical psychology. In the
educational domain, researchers are continually endeavor-
ing to develop AR in learning. To initially understand how
AR could help learning in science education, relevant
studies were firstly searched for through the Web of
Knowledge and Scopus database using keywords such as
augmented reality and science learning or science educa-
tion. Basically, studies with either empirical data or topics
related to science (e.g., astronomy, chemistry, biology, or
engineering) were selected. It should be noted, however,
that to depict the current AR technology in science edu-
cation, several studies defined as using AR applications
which were actually developed using VR-based systems
were excluded in this paper. Hence, 12 articles regarding
AR-related work were chosen for analysis. As shown in
Table 1, the technical features, focus topics, participants,
and affordances in science learning of these studies are
summarized. Moreover, in order to fully describe what has
been investigated in AR-related science learning from a
pedagogical perspective, the VR-based learning model
(Salzman et al. 1995) was used as a basis for examining the
affordances of AR in science learning. That is, the selected
articles are discussed according to the dimensions in the
model (i.e., technical features, science concepts, learner
characteristics, interaction experience, learning experience,
learning process, and learning outcomes).
Technical Features and Science Concepts
Through the selected articles on the application of AR,
their supports in science learning are addressed by exam-
ining the associations between technical features (i.e.,
image-based AR and location-based AR) and science
concepts.
Image-Based AR for Spatial Ability, Practical Skills,
and Conceptual Understanding
Several image-based AR applications have been designed
for science learning. For instance, Martı
´
n-Gutie
´
rrez et al.
(2010) designed an AR book with marker identification,
namely AR-Dehaes, which utilizes iconic markers and
offers 3D virtual objects displayed on the screen to help
students to handle and visualize engineering graphics and
further enhance their spatial ability. For inorganic chem-
istry education, the image-based AR setup in a multimedia
classroom could support students in developing spatial
intuition regarding the 3D arrangement of crystalline
structures (Nu
´
n
˜
ez et al. 2008). Another image-based AR
applied with teachers’ instructional guidance for geosci-
ences in a classroom, which was employed by Kerawalla
et al. (2006), required children to hold an AR tile to
manipulate the spatial relationships between the 3D objects
of the Earth, Sun, and Moon.
Similarly, in the field of astronomy, Shelton and Stevens
(2004) also used an image-based AR system (with HMD)
to provide a way to understand the spatial concept of
Earth–Sun relationships, as well as to make a conceptual
Fig. 5 A comparison of image-based and location-based AR
J Sci Educ Technol (2013) 22:449–462 453
123
