Authenticity and interest in virtual reality: Findings from an experiment including educational virtual environments created with 3D modeling and photogrammetry

Abstract

Virtual Reality (VR) and photogrammetry are emerging technologies that facilitate and shape the ongoing digital transformation of education. VR offers new opportunities for creating immersive and interactive educational experiences. Photogrammetry enables new ways to create lifelike educational virtual environments and is becoming an alternative to manual 3D modeling with graphics software. The manner in which VR affects the authenticity of educational experiences has been addressed in previous educational and psychological research. Empirical papers have so far focused on the authenticity of educational VR environments created by 3D modeling. However, little is known about the authenticity of educational VR environments developed with photogrammetry. Given that VR provides rich multi-sensory experiences and interests can be stimulated by engaging contexts, educational VR environments also possess great potential to support interest development. What is still unknown regarding this topic are the beneficial characteristics of VR environments and the individual variables required to trigger and explain interest development. Consequently, we conducted an experiment following up on the mentioned authenticity and interest research questions in the context of higher education. A two-group between-subjects design was used and N = 64 educational science and psychology university students gathered information about a railroad bridge wearing a head-mounted display (HMD). The control group encountered an educational virtual environment created with 3D modeling. The intervention group was presented with the same educational virtual environment but the main object of the railroad bridge was generated by photogrammetry. Situational interest was measured in the pretest and the posttest; authenticity-related variables (i.e., presence and representation fidelity) were assessed in the posttest. Concerning authenticity, there were no significant group differences. Photogrammetry might thus not affect authenticity in educational contexts in which participants focus on gathering information. Regarding interest development, there were two main findings. First, interest in VR for learning increased from pretest to posttest, supporting that interest can be induced in VR. Second, a large share of posttest interest was explained by presence and pretest interest, highlighting the importance of these variables. CITATION Fink MC, Sosa D, Eisenlauer V and Ertl B (2023) Authenticity and interest in virtual reality: Findings from an experiment including educational virtual environments created with 3D modeling and photogrammetry. Front. Educ. 8:969966.

Full text

Frontiers in Education 01 frontiersin.org
Authenticity and interest in virtual
reality: Findings from an experiment
including educational virtual
environments created with 3D
modeling and photogrammetry
MaximilianC.Fink *, DianaSosa , VolkerEisenlauer and
BernhardErtl
Learning and Teaching with Media, Institute of Education, Universität der Bundeswehr München, Neubiberg,
Germany
Virtual Reality (VR) and photogrammetry are emerging technologies that facilitate and
shape the ongoing digital transformation of education. VR oers new opportunities
for creating immersive and interactive educational experiences. Photogrammetry
enables new ways to create lifelike educational virtual environments and is becoming
an alternative to manual 3D modeling with graphics software. The manner in which
VR aects the authenticity of educational experiences has been addressed in previous
educational and psychological research. Empirical papers have so far focused on
the authenticity of educational VR environments created by 3D modeling. However,
little is known about the authenticity of educational VR environments developed with
photogrammetry. Given that VR provides rich multi-sensory experiences and interests
can bestimulated by engaging contexts, educational VR environments also possess
great potential to support interest development. What is still unknown regarding this
topic are the beneficial characteristics of VR environments and the individual variables
required to trigger and explain interest development. Consequently, weconducted
an experiment following up on the mentioned authenticity and interest research
questions in the context of higher education. A two-group between-subjects
design was used and N = 64 educational science and psychology university students
gathered information about a railroad bridge wearing a head-mounted display
(HMD). The control group encountered an educational virtual environment created
with 3D modeling. The intervention group was presented with the same educational
virtual environment but the main object of the railroad bridge was generated by
photogrammetry. Situational interest was measured in the pretest and the posttest;
authenticity-related variables (i.e., presence and representation fidelity) were assessed
in the posttest. Concerning authenticity, there were no significant group dierences.
Photogrammetry might thus not aect authenticity in educational contexts in which
participants focus on gathering information. Regarding interest development, there
were two main findings. First, interest in VR for learning increased from pretest to
posttest, supporting that interest can be induced in VR. Second, a large share of
posttest interest was explained by presence and pretest interest, highlighting the
importance of these variables.
KEYWORDS
authenticity, interest, virtual reality, photogrammetry, 3D modeling, presence, educational
virtual environment, laser scanning
TYPE Original Research
PUBLISHED 27 January 2023
DOI 10.3389/feduc.2023.969966
OPEN ACCESS
EDITED BY
Gergana Vladova,
University of Potsdam,
Germany
REVIEWED BY
Fernando Moreu,
University of New Mexico,
UnitedStates
Malte Rolf Teichmann,
University of Potsdam,
Germany
*CORRESPONDENCE
Maximilian C. Fink
Maximilian.Fink@unibw.de
SPECIALTY SECTION
This article was submitted to
Digital Education,
a section of the journal
Frontiers in Education
RECEIVED 16 June 2022
ACCEPTED 03 January 2023
PUBLISHED 27 January 2023
CITATION
Fink MC, Sosa D, Eisenlauer V and Ertl B (2023)
Authenticity and interest in virtual reality:
Findings from an experiment including
educational virtual environments created with
3D modeling and photogrammetry.
Front. Educ. 8:969966.
doi: 10.3389/feduc.2023.969966
COPYRIGHT
© 2023 Fink, Sosa, Eisenlauer and Ertl. This is
an open-access article distributed under the
terms of the Creative Commons Attribution
License (CC BY). The use, distribution or
reproduction in other forums is permitted,
provided the original author(s) and the
copyright owner(s) are credited and that the
original publication in this journal is cited, in
accordance with accepted academic practice.
No use, distribution or reproduction is
permitted which does not comply with these
terms.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 02 frontiersin.org
1. Introduction
Virtual Reality (VR) and photogrammetry are new technologies that
will have a major impact on digital education in the coming years. In the
following, weprovide a brief introduction to these new technologies and
highlight shortly their importance for education.
Virtual Reality (VR) refers to computer simulations and specic
hardware that together convey a strong feeling of being physically
present within an interactive digital environment (Radianti etal.,
2020). Recently, head-mounted displays (HMDs) have become
increasingly popular as VR hardware. HMDs are goggles that enable
an immersive experience with sharp displays and track users’
movements in the physical world to project them into a digital space.
Some HMDs require a connection to a computer to operate, while
others can beused as a standalone device. Unlike other immersive VR
devices, such as CAVE systems – dedicated rooms consisting of
multiple screens and motion tracking systems – HMDs require less
space and do not have to be permanently installed (Jensen and
Konradsen, 2018). As this advantage is crucial for many educational
settings, such as bringing VR to regular classrooms, this paper will
focus on HMDs as VR devices.
Photogrammetry records real-world objects with light-wave based
and other sensors for survey and evaluation purposes (American Society
for Photogrammetry and Remote Sensing, 2022). As photogrammetry
can beused to create digital models of real-world objects and landscapes
(Historic England, 2017), it is becoming an alternative to 3D modeling
for creating educational virtual environments. In 3D modeling, a
designer creates objects manually using computer graphics soware in
a time-consuming process requiring extensive expertise (Erolin, 2019).
VR and photogrammetry are currently becoming more accessible
as the required soware and hardware become more aordable (Nebel
etal., 2020; Pellas etal., 2020). Likewise, both technologies are also
becoming easier to use: VR engines and asset stores frequently contain
templates that facilitate scene set-up and implement basic user
navigation. Photogrammetry soware oen includes detailed
documentation and automated workows, which foster the swi
creation of 3D objects. As a result of the improved accessibility and
higher convenience, an increased number of educational researchers
and practitioners can make use of VR and photogrammetry to deliver
or create content (Jensen and Konradsen, 2018; Nebel etal., 2020).
It is an open question whether innovative educational VR
environments created by photogrammetry are perceived as more
authentic than traditional educational VR environments created by 3D
modeling. On the one hand, photogrammetry may raise environmental
delity, that is, the “degree to which variables in the training environment
resemble those in the real world” (Waller et al., 1998, p. 130), by
providing photorealistic models. On the other hand, perceived
authenticity may also depend on other factors, such as the quality of the
interaction and the realism of the provided task (Witmer and
Singer, 1998).
Given that educational VR environments provide rich multi-sensory
experiences and interests can bestimulated by engaging contexts (Hidi
and Renninger, 2006), VR can also be used to support interest
development. Some empirical studies have been conducted on interest
development in educational VR environments (e.g., Makransky etal.,
2020; Petersen etal., 2020). Yet, much remains unknown about how
interest is triggered by educational VR environments and which
individual variables, such as pretest interest, and presence, contribute to
this process.
At the beginning of this section, wewill elaborate on creating
content for educational VR environments with 3D modeling and
photogrammetry. We will then address the mentioned research
questions on authenticity and interest development.
1.1. Creating content for educational virtual
reality environments with 3D modeling and
photogrammetry
In the past, the creation of educational VR content was almost
exclusively carried out using 3D modeling. In 3D modeling, designers
perform tasks such as creating meshes, applying textures, adjusting light
settings, and building animations using computer graphics soware
(such as Blender or Maya). is process is rather time-consuming and
requires extensive expertise (Erolin, 2019). Content for VR applications
is produced with this method usually by 3D designers who have
completed professional training or education in the eld of graphic
design. With the wider distribution and better availability of
photogrammetry soware and hardware, the creation of VR content can
now also be carried out with photogrammetry. Photogrammetric
methods can use various sensors including photo cameras, drone
cameras, laser-scanning, and geo-references (American Society for
Photogrammetry and Remote Sensing, 2022). Data from one of these or
multiple sources is then processed to a coherent 3D model relatively
automatically by algorithms comparing the correspondence between the
records (Historic England, 2017). Photogrammetry soware, such as
RealityCapture or AutoDesk, includes automated workows and requires
the user in easier projects to mainly clean and simplify the model until it
ts the project’s requirements (Trebuňa etal., 2020). Content for VR
applications can thus bemore easily created using this method by users
without specic professional training or prior education (Erolin, 2019).
Both types of content creation can beused to produce educational
VR environments including single objects, buildings, virtual humans,
and even full terrains. However, educational VR environments are
currently mainly created with 3D modeling. Researchers and designers
acquire suitable3D objects created by others from asset stores or create
these objects with a modeling soware such as blender. Soon,
photogrammetry will likely beused increasingly to develop models for
educational VR environments. Asset stores will then include more
objects created with photogrammetry, and researchers and designers
will become capable of producing photogrammetric 3D models
themselves. As mentioned, there are several dierences between
developing models for educational VR environments with 3D modeling
and photogrammetry. Table 1 lists and contrasts these dierences
between both content creation methods that concern the creation
process, as well as the structure, texture, lighting, and level of detail of
the created 3D model.
Despite its great potential, research on education in VR using the
new technology of photogrammetry is still lacking (Nebel etal., 2020).
us, this paper sets out to compare educational environments created
with 3D modeling and photogrammetry. In doing this, the paper focuses
on photogrammetry with a hybrid modeling approach that combines
records from photos and laser-scanning (Historic England, 2017). Laser-
scanning is a method in which a lidar generates multiple point clouds
(i.e., x, y, z coordinates and other characteristics such as color) of a
structure (Historic England, 2018). Following the mentioned hybrid
modeling approach, the photogrammetry soware then imports the
photographs and laser scans and generates a coherent 3D model

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 03 frontiersin.org
from them. e hybrid modeling approach was selected because it is
particularly suitable for capturing large objects, and our study’s
educational environment included a large railroad bridge as the main
object and center of attention.
1.2. Increasing the authenticity of
experiences through virtual reality
One goal that VR has had since its inception is undoubtedly to
increase the authenticity of experiences. A classical example of one of
the rst VR systems illustrates this goal. e Sensorama created by VR
pioneer Morton Heilig in 1961 was a VR movie theater that stimulated
multiple senses not normally considered when watching movies
(McLellan, 2008). Users were surrounded by displays and experienced
smell and motion in realistic situations such as during helicopter ights
(McLellan, 2008). is example shows that VR technology uses both
soware and hardware together to enhance authenticity. On the soware
side, authenticity is pursued in the realms of education by providing
interactive, realistic simulations of didactically-selected systems,
situations, and content (Witmer and Singer, 1998; Chernikova etal.,
2020). On the hardware side, devices such as HMDs have immersive
displays and intuitive user input that may evoke a sense of physicality
and authenticity (Velev and Zlateva, 2017). Next, dierent types of
variables used and examined in educational authenticity research will
bedistinguished from each other.
Two important authenticity-related variables include presence and
representation delity (Dalgarno and Lee, 2010). Presence is the feeling
of being within a virtual environment yet being physically located in
another place (Witmer and Singer, 1998). is concept contains,
according to a factor analysis by Schubert etal. (2001), the aspects of
involvement, realness, and spatial presence. Involvement refers to
focusing attentively on the environment, realness is a rating to what
extent the environment is as realistic and believable as the real world,
and spatial presence denotes feeling located physically in an environment
(Schubert etal., 2001). Representation delity can bedened as the level
of realism that the educational 3D environment with its included objects
possesses and is aected by factors such as image and lighting quality
(Dalgarno etal., 2002). Even though it has been argued that presence is
a more subjective concept and representation delity is more objective,
it has been theorized that both aspects aect each other (Lee etal., 2010;
Makransky and Petersen, 2021).
ere are three main reasons for developing educational VR
environments with high authenticity. First, providing authentic
educational virtual environments can promote important aective
outcomes. In this respect, positive eects of authentic educational virtual
environments have been reported for motivation (Makransky and
Petersen, 2019), curiosity (Schutte, 2020), and interest (Makransky etal.,
2020; Petersen etal., 2020). Second, authenticity-related variables are
considered to beimportant mediators of learning (Lee etal., 2010;
Makransky and Petersen, 2021). In this respect, empirical studies have
shown that highly authentic educational virtual environments can evoke
deeper cognitive processing but can also lead to negative side eects
such as cognitive overload (Makransky etal., 2019; Škola etal., 2020).
ird, situated learning approaches contend that educational virtual
environments with high functional and physical authenticity increase
the contextualization of what is experienced and facilitate knowledge
acquisition (Renkl etal., 1996). In fact, a meta-analysis by Wu etal.
(2020) reported that more declarative knowledge was acquired in
educational VR environments on average than in regular lecture-based
training. However, this meta-analysis also showed relatively comparable
knowledge levels were acquired in HMD-based educational VR
environments with higher authenticity and regular desktop-based VR
environments with lower authenticity (Wu etal., 2020).
e literature provides several insights on whether VR created by
3D modeling or photogrammetry would bemore authentic. According
to Witmer and Singer (1998), experienced presence depends on
multiple, mostly subjective judgments. ese judgments include,
amongst other things, the scene realism (i.e., the feeling that the stimuli
are related to each other), the perceived level of control (i.e., the level of
interactivity), and the meaningfulness (i.e., the personal value).
However, Witmer and Singer (1998) also acknowledge that external
factors such as image and HMD resolution aect these subjective
judgments. Skulmowski etal. (2021) take a more technical stance and
contend that the level of detail provided by the mesh, texture, and
lighting of a 3D model aect its realism rating. In line with these
theoretical reasons, authenticity-related variables should behigher in
VR created by photogrammetry than in VR created by 3D modeling.
1.3. Interest development through virtual
reality
Another goal that VR can pursue in educational contexts lies in
supporting interest in new topics and content. In the following
paragraphs, wewill dene interest and discuss its relevance for education
before wedescribe major tenets of interest development theories.
Interest is an aective-cognitive construct focusing on specic
objects or topics (Krapp, 2002) that evokes repeated engagement and
increased persistence (Hidi and Renninger, 2006; Harackiewicz etal.,
2016). is construct is essential for life-long learning (Krapp, 2002). For
instance, associations of interest with attention (Bolkan and Grin, 2018),
TABLE1 Dierences between creating educational virtual environments with 3D modeling and photogrammetry.
3D modeling Photogrammetry
Creation process Created by designers through modeling in 3D graphics soware Created semi-automatically with photogrammetry soware
Structure of the 3D model Manually created object structure which is frequently not based
on correct dimensions and proportions
Real object structure with correct dimensions and proportions
Texture of the 3D model Pattern overlay selected for the manually-created object
structure. Frequently, stock textures are used
Real surface texture of the object. Frequently, ne details are included
Lighting of the 3D model Based on the decisions of the designer Based on the lighting during the recording of the object
Level of detail of the 3D model Depending on the eort and skill of the designer creating the
model
Depending on the scan resolution and technique as well as selected
photogrammetry algorithms

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 04 frontiersin.org
learning strategies (Schiefele, 1991; McWhaw and Abrami, 2001), and
academic achievement (Schiefele etal., 1992; Kpolovie etal., 2014) have
been shown.
According to two popular interest development theories (Krapp,
2002; Hidi and Renninger, 2006), this construct evolves from a
situational to an individual interest. Situational interest is a temporary
state that can either fade away or last longer in the further course of
interest development (Hidi, 1990). Individual interest, however, is
more stable, lasts longer, and is rmly rooted in a subject’s personality
(Renninger et al., 2002). e mentioned development process
includes phases in which interest is rst induced and then sustained
over a longer period to transform qualitatively (Krapp, 2002; Hidi
and Renninger, 2006). Value and feeling judgments, individual
variables (e.g., cognitive prerequisites), and external variables (e.g.,
availability of engaging materials) play an important role and interact
in the formation and sustainment of interest (Krapp, 2002; Hidi and
Renninger, 2006). As this paper focuses on the early phase of interest
development, wewill now discuss the triggering of situational interest
and variables associated with interest development in this phase.
Educational virtual environments can bean important trigger of
situational interest (Hidi, 1990). In particular, educational environments
that generate positive value and feeling judgments are thought to induce
situational interest (Krapp, 2002; Hidi and Renninger, 2006). Multiple
studies have supported this point, showing that meaningful educational
environments foster the development of situational interest and related
constructs in high school and college students (Cordova and Lepper,
1996; Verhagen etal., 2012; Siklander etal., 2017). Recently, these ndings
have also been substantiated for educational VR environments.
Makransky etal. (2020) discovered in two experiments that taking part in
laboratory VR simulations with HMDs increased interest in science topics.
In terms of variables associated with early interest development, the
two popular theories mentioned (Krapp, 2002; Hidi and Renninger,
2006) argue that pre-existing individual interest and triggered situational
interest are linked with the development of later situational interest in
new content. Findings from studies on interest trajectories using shorter
and longer digital simulations and inquiry environments supported this
claim (Tapola etal., 2013; Chen etal., 2016; Rodríguez-Aecht etal.,
2018). As digital simulations and inquiry environments share many
commonalities with educational VR environments (e.g., high
interactivity and authenticity), webelieve this nding should also hold
true for educational VR environments. Furthermore, the variable
presence could also bea predictor of early interest development. is
could bethe case because presence is assumed to bebenecial for many
regulatory and aective aspects of self-directed learning (Krapp, 2002).
In line with this reasoning, an experiment in which psychology students
took part in educational VR simulations with high and low immersion
wearing HMDs showed that the development of situational interest was
aected by varying the degree of presence (Petersen etal., 2022).
1.4. Research questions and hypotheses
Against this background, wepose three research questions:
RQ1: To what extent do the authenticity-related variables involvement,
realness, spatial presence, and representation delity dier in educational
VR environments created by 3D modeling and photogrammetry?
We hypothesize that involvement (H1.1), realness (H1.2), spatial
presence (H1.3), and representation delity (H1.4) are higher in the VR
created by photogrammetry than in the VR created by 3D modeling.
RQ2: To what extent does interest develop in educational VR
environments created with 3D modeling resp. photogrammetry?
We assume that interest in VR for learning (H2.1) and interest in
bridges (H2.2) increase across both educational VR environments.
RQ3: To what extent can posttest situational interest bepredicted by
pretest situational interest, group (3D modeling resp. photogrammetry),
presence, and representation delity?
We expect that posttest interest in VR for learning (H3.1) and
posttest interest in bridges (H3.2) is predicted by pretest interest, group,
presence, and representation delity.
2. Method
2.1. Participants
N = 64 educational sciences and psychology university students of a
German university took part in the study. e sample consisted of 17.2%
(n = 11) participants below the age of 21, 59.4% between 21 and 24
(n = 38), 20.3% (n = 13) above 24, and 3.1% (n = 2) participants with a
missing age. 50% (n = 32) of the participants were freshmen, 39.1%
(n = 25) sophomores, 6.2% (n = 4) juniors, and 1.6% (n = 1) seniors; 3.1%
(n = 2) did not report their study year. A majority of 81.2% (n = 52) of
the participants had none or very little VR experience, 14.1% (n = 9) had
little experience with VR, and 4.7% (n = 3) did not provide an answer.
2.2. Study design, procedure, and
randomization
e experiment employed a two-group between-subjects design
with a pretest and a posttest. e pretest included questions on
demographics and interest. Aerwards, participants took part in an
electronic tutorial which made them familiar with the educational VR
environment. e subsequent intervention phase focused on exploring
a railroad bridge in VR. e control group (n = 33) encountered an
educational virtual environment created with 3D modeling in this
phase. e intervention group (n = 31), however, was presented with the
same educational virtual environment in which the main object of the
railroad bridge was generated by photogrammetry. Both groups are
depicted in Figure1 and are closer described in the remainder of this
chapter. Participants were randomly assigned to one of the two described
groups by drawing a ticket. Consequently, the experimenter
administered the corresponding educational VR environment to them.
e posttest assessed the experienced presence and representation
delity and participants’ interest aer taking part.
2.2.1. Educational standpoint and VR scenario
From an educational standpoint, the study can beconsidered a
virtual eld trip. A virtual eld trip is “a journey taken without actually
making a trip to the site” (Woerner, 1999, p.5) that aims to reach
predened educational goals. e main educational goals the study
pursued were sparking interest in a construction engineering topic,
experiencing a concrete representation of a railroad bridge, and
conveying basic knowledge about bridge construction and
characteristics. Consistent with self-determination theory (Deci etal.,
1991), two instructional design choices were made. e educational
virtual environment emphasized free exploration to provide a sense of
autonomy. Moreover, the educational virtual environment only

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 05 frontiersin.org
conveyed basic knowledge to allow the participating education and
psychology students to experience a sense of competence. Finally, the
study draws on theories of multimedia education (Makransky and
Petersen, 2021; Vogt, 2021), which emphasize that delity and
instructional support (e.g., visualizations) can improve motivational and
aective states.
e VR scenario depicted in Figure 2 focused on exploring a
railroad bridge and gathering information about it. More specically,
participants received 7 min to teleport between eight dierent platforms
freely. Depending on the active teleportation platform, participants
could interact with one of two types of user interaction: video screens
and visualizations. e video screens played lms of PowerPoint slides
with audio recordings that focussed on the bridge’s features and
characteristics. e visualizations (such as blueprints) depicted the
bridge’s features and characteristics graphically. With respect to
content, it should behighlighted that some of the content was related
to engineering (e.g., the blueprints) while other content had to do with
the bridge itself (e.g., the environment). e scenario, video screens,
visualizations and the content were created in collaboration with the
civil engineering department and reviewed by them for correctness.
2.2.2. Tutorial, terrain, and contents in the VR
At the beginning of the experiment, participants took part in an
electronic tutorial. is tutorial was situated in a neutral virtual
environment and lasted about four minutes. As part of the tutorial,
participants familiarized themselves with teleporting across the
platforms and interacting with video screens and visualizations. is
way, the tutorial prepared the participants well for the educational VR
environment which took place aerwards.
Both educational VR environments used the same simple terrain.
is terrain consisted of a valley with a dried-out river, hills, and
conifers. e railroad bridge, which was varied in both groups, was
located in the middle of this terrain and crossed the dried-out river.
rough its large size and center position, the railroad bridge was the
educational environment’s main object and center of attention.
Next, we will discuss how both of the created educational VR
environments depicted in Figure 1 diered from each other. In the
control group, a regular 3D model of a railroad bridge was used. is 3D
model was created by hand using the soware blender (e Blender
Foundation, 2021). e proportions and surface texture of the 3D model
were loosely based on the real bridge. A neutral, grey texture that
FIGURE1
Overview of the 3D model contained in the control and intervention groups. The used 3D models had a relatively similar geometry and comparable
dimensions, even though they diered on several other objective characteristics (please see the box). In the photogrammetry group, details such as grati
were included and fine lines of the structure can beseen.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 06 frontiersin.org
matched to the level of detail of the surrounding environment was
applied to the railroad bridge. In the intervention group, a high-
resolution photogrammetric model of a real railroad bridge was included.
is model was recorded with the terrestrial laser scanner FARO FocusS
(Faro Technologies, 2022) and then post-processed with the soware
RealityCapture (Epic Games, 2021). us, the model’s proportions and
surface structure were very close to the original railroad bridge. e
texture was created from photos of the original bridge also taken by the
terrestrial laser scanner and included grati and realistic lighting. To
hold both described groups as comparable as possible, the used 3D
models had a relatively similar geometry and comparable dimensions.
e mentioned texture dierence between the two groups (i.e., the
neutral texture in the 3D modeling group and the high-resolution texture
in the photogrammetry group) was deliberately retained. Firstly, this
decision magnied a particular feature of the two experimental groups
that could have been detected in our comparison of authenticity (RQ1).
Adding a high-resolution texture and details to the 3D modeling group
would have decreased the expected dierences between the 3D modeling
group and the photogrammetry group wewere investigating. Secondly,
weopted for this decision because the described dierence in texture
detail could occur frequently when these two content creation methods
are used in educational settings. Educational scientists and practitioners
who create 3D models with graphics soware will mostly use simple
stock textures. is group of people typically does not have the skills and/
or time to create high-resolution textures for 3D models that contain ne
details and artwork themselves. It is unquestionable that educational
scientists and practitioners with large budgets can obtain 3D models with
ne details and artwork. e IT companies hired to do this either employ
their own 3D modelers and art designers or purchase high-resolution 3D
models from third parties. When educational scientists and practitioners
A
BC
FIGURE2
Scenario and user interaction included in the VR environment. Based on the scenario (A), participants started at the platform “South” and then teleported
freely to seven other teleportation platforms in the order of their choosing. Video screens (B) and visualizations (C) were oered as user interaction at the
teleportation platforms.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 07 frontiersin.org
use photogrammetry, matters are dierent. Educational scientists and
practitioners can use photogrammetry to create 3D models with high
resolution and detail without spending large sums of money and
hiring contractors.
2.2.3. Creation and implementation of the
educational VR environments
e described educational virtual environments and their terrain
were created with the game engine Unity (Unity Technologies, 2020).
Functions specic to VR, such as input and output via the HMDs, were
developed using the XR Interaction Toolkit (Unity Technologies, 2021).
Scripts for specic purposes and customizations of the educational
virtual environment (e.g., creating the control ow of the educational
VR environments), were written with the programming language C#.
e experiment was conducted in a small, dedicated VR laboratory.
In this lab, the participants had sucient space for walking around and
turning their heads. Further, participants were observed by an
experimenter who ensured their safety. While instrument data were
collected using paper-and-pencil questionnaires, the tutorial and the
educational environments ran on a gaming laptop. An HTC Vive (HTC,
2016) was used as a HMD.
2.3. Instruments
2.3.1. Interest
Interest was assessed with a 6-item scale (Rotgans and Schmidt,
2011). Two versions of this scale were created: One measuring interest
in VR for learning and one tapping into interest in bridges. For brevity
reasons, the rst scale is also referred to as VR interest, while the second
scale is called Bridge interest. Both scales were used in the pretest and
the posttest. e described instrument utilized a Likert scale ranging
from 1 (strongly disagree) to 5 (strongly agree). Lower values indicated
lower interest, whereas higher values represented higher interest.
2.3.2. Presence
Presence was measured with a 13-item questionnaire (Schubert
etal., 2001). is questionnaire comprised items on the involvement,
realness, and spatial presence of the educational environment. It was
administered directly aer taking part in the VR and used a 7-point
Likert scale reaching from −3 to 3 with varying anchors. Lower values
stood for lower presence, higher values for higher presence.
2.3.3. Representation fidelity
Data on representation delity was collected with a 3-item
instrument (Lee etal., 2010). is instrument focused mainly on the
experienced realism of the 3D scene. Anchors of the instrument ranged
from 1 (strongly disagree) to 5 (strongly agree).
3. Results
3.1. Descriptive statistics and manipulation
checks
e descriptive statistics for the experiment are reported in Table2.
Manipulation checks using two-sided t-tests were conducted to screen
for unwanted pretest group dierences. Pretest interest in VR for learning
and pretest interest in bridges did not dier signicantly between
the control and the intervention group, as intended [t(61) = 0.74,
p = 0.461 resp. t(62) = 1.11, p = 0.270].
3.2. Dierences in authenticity-related
variables in educational VR environments
created by photogrammetry and through 3D
modeling (RQ1)
To examine RQ1, one-sided independent samples t-tests of the
authenticity-related variables involvement, realness, spatial presence,
and representation delity split by the group were conducted. ese
analyses are visualized and reported in a boxplot in Figure3. All four
mentioned authenticity-related variables were not signicantly greater
in the intervention group than in the control group. us, the hypotheses
H1.1-H.1.4 were not supported.
3.3. Interest development in educational VR
environments (RQ2)
To investigate RQ2, one-sided dependent sample t-tests of the
interest development from the pretest to the posttest were calculated.
ese analyses are reported and visualized in Figure4. Interest in VR for
learning increased from the pretest to the posttest in the full sample.
erefore, H2.1 was substantiated. However, bridge interest did not
improve signicantly from the pretest to the posttest in the full sample.
erefore, H2.2 was not supported. Additionally, explorative analyses
were also conducted separately for the intervention group and the
control group (see Figure4). ese explorative analyses rely on small
subsamples and should not beoverinterpreted. However, the explorative
analyses provide an initial indication that interest increases did not dier
strongly according to the group (resp. content creation method).
3.4. Variables explaining situational posttest
interest (RQ3)
Before conducting regression analyses, we inspected
intercorrelations between the interest-related and authenticity-related
TABLE2 Means (and SDs) for authenticity-related and interest-related
variables in the control group, the intervention group, and combined for
both groups.
Variable Control Intervention Combined
Authenticity-related variables 0.49 (0.63) 0.40 (0.57) 0.45 (0.60)
Spatial presence 0.94 (0.73) 0.59 (0.78) 0.77 (0.77)
Involvement 0.47 (0.69) 0.40 (0.69) 0.43 (0.68)
Realness 0.06 (1.23) 0.22 (0.75) 0.14 (1.02)
Representation delity 4.10 (0.79) 4.30 (0.58) 4.20 (0.70)
Interest-related variables
Pretest VR interest 4.24 (0.39) 4.17 (0.39) 4.20 (0.39)
Posttest VR interest 4.37 (0.38) 4.25 (0.48) 4.32 (0.44)
Pretest bridge interestᵃ3.55 (0.51) 3.34 (0.50) 3.44 (0.51)
Posttest bridge interestᵃ3.62 (0.65) 3.45 (0.49) 3.54 (0.58)
ais measure was only assessed for a subsample of the study.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 08 frontiersin.org
variables (see Table3). Both posttest interest variables (interest in VR for
learning and interest in bridges) correlated signicantly with the
corresponding pretest interest, the full presence score, the authenticity-
related variables realness, and representation delity. Despite being
smaller and sometimes not reaching signicance, there were also
associations between both posttest interest variables and involvement
and spatial presence. Representation delity and the full presence score
displayed a medium positive correlation (r = 0.32) with each other.
Additionally, separate correlation analyses for the control group and the
intervention group are available in Supplementary Tables 1, 2.
3.4.1. Interest in VR for learning
A hierarchical regression for posttest interest in VR for learning as
outcome was conducted and is presented in Table4. Model V1 (R
2
= 0.21),
containing the pretest interest, was signicant and explained a substantial
amount of variance. Model V2 (R
2
= 0.22), additionally including the
group, did not explain signicantly more variance than Model V1 [F(1,
60) = 1.20, p = 0.277, ΔR
2
= 0.02]. In Model V3 (R
2
= 0.31), presence was
added as predictor which resulted in additional explained variance in
comparison to Model V2 [F(1, 59) = 7.54, p = 0.008, ΔR
2
= 0.09]. Model
V4 (R
2
= 0.36) included all mentioned predictors plus representation
delity. Model V4 explained signicantly more variance than Model V3
[F(1, 58) = 4.76, p = 0.033, ΔR2 = 0.05]. e overall model including pretest
interest, group, presence and representational delity was signicant.
However, group was not a signicant predictor of posttest interest in VR
for learning. Consequently, H3.1 was partially substantiated.
3.4.2. Bridge interest
A hierarchical regression for posttest bridge interest as a dependent
variable was carried out and is shown in Table5. Model B1 (R2 = 0.34),
including the pretest bridge interest, was signicant and explained a
substantial amount of posttest bridge interest. Model B2 (R
2
= 0.34),
which also contained the group, did not explain signicantly more
variance than Model B1 [F(1, 33) = 0.09, p = 0.761, ΔR
2
= 0.00]. Model
B3 (R
2
= 0.49) included, as an additional variable, the full presence score
and resulted in a raise of explained variance compared to Model B2
[F(1, 32) = 9.41, p = 0.004, ΔR
2
= 0.15]. Model B4 (R
2
= 0.52) consisted of
all mentioned variables plus representation delity and did not increase
explained variance compared to Model B3 [F(1, 31) = 1.48, p = 0.233,
ΔR
2
= 0.02]. Bridge interest was explained by pretest interest, and
presence. Yet, group and representational delity were not signicant
predictors. Based on these results, H3.2 was partially substantiated.
4. Discussion
4.1. Authenticity in virtual reality created by
3D modeling and photogrammetry (RQ1)
is paper examined to what extent authenticity-related variables
dier in educational VR environments created by 3D modeling and
photogrammetry. Our results indicate that the three presence-related
variables (involvement, spatial presence, and realness) and
FIGURE3
Group comparisons using one-sided independent t-test results of the authenticity-related variables. Involvement, realness, and spatial presence ranged from
−3 to 3. Representation fidelity ranged from 1 to 5. The box of the boxplot visualizes the interquartile range, including the median, Q2 (the 25th percentile),
and Q3(the 75th percentile). Whiskers of the boxplot show Q1 and Q4. The red circles represent the mean. Eect sizes are reported using Cohen’s d.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 09 frontiersin.org
representational delity do not dier between educational environments
created by using 3D modeling and photogrammetry. ere are several
possible explanations for our results. One explanation may be that
subjective judgments play a particularly important role in creating the
feeling of presence (Witmer and Singer, 1998). ese subjective
judgments might have been comparable in both groups and perhaps led
to the reported equal presence and representation delity level. Another
explanation for our ndings could lie in the measurement of the
authenticity-related variables. Our instruments did not distinguish
between the multiple, varying objects contained in the 3D scene but
measured perceived authenticity globally. Dierences between the
authenticity of the environment (i.e., the terrain) and the varied main
object of interest (i.e., the railroad bridge) could therefore not
be captured. It seems possible that especially in hybrid educational
environments involving objects created with 3D modeling and
photogrammetry, separate measurements of authenticity are necessary
to reveal relationships between object characteristics and authenticity-
related variables. A nal explanation of our ndings could bethat while
the authenticity of virtual environments might generally depend on the
detail, texture, and lighting of 3D models (Skulmowski etal., 2021), this
relationship may bereduced in hybrid educational virtual environments.
Participants likely focused so strongly on gathering information in our
educational virtual environment that they did not notice the objective
dierences in the detail, texture, and lighting created by the main object
(i.e., the railroad bridge) included in the scene. Consequently, further
research should more closely examine in which contexts and under what
conditions, authenticity-related variables are determined by the detail,
texture, and lighting of 3D objects.
4.2. Triggering of situational interest through
virtual reality (RQ2)
Another focus of this paper was to investigate the triggering of
situational interest through educational VR environments. In doing so,
wemeasured two types of interest: interest in VR for learning and interest
in bridges. Interest in VR for learning increased signicantly from the
pre- to the posttest in the full sample. is nding is in line with other
studies’ results that regular technology-enhanced educational
environments can induce situational interest and related constructs
(Cordova and Lepper, 1996; Verhagen etal., 2012; Siklander etal., 2017).
Moreover, this nding adds support to two studies that found that
interest can be triggered within educational VR environments
(Makransky etal., 2020). Interest in bridges rose slightly from the pre- to
the posttest in the full sample but did not reach signicance. e
discovered non-signicant interest growth for this type of interest was
smaller than for interest in VR for learning. One possible explanation for
this result could bethat it may have been more dicult to induce
interest in an engineering topic among the participating educational
FIGURE4
Boxplots of the interest scores and results of one-sided paired-samples
t-tests. The results for the main analyses are based on the full sample.
The explorative analyses are reported separately for the control and the
photogrammetry group. Interest scores ranged from 1 to 5. The box of
the boxplot visualizes the interquartile range, including the median, Q2
(the 25th percentile), and Q3 (the 75th percentile). Whiskers of the
boxplot show Q1 and Q4. The red circle depicts the mean. Eect sizes
are reported using Cohen’s d.
TABLE3 Intercorrelations between authenticity-related and interest-related variables combined for both educational virtual environments.
1. 2. 3. 4. 5. 6. 7. 8.
1. Presence full
2. Spatial presence 0.55***
3. Involvement 0.74*** 0.08
4. Realness 0.86*** 0.17 0.58***
5. Representation delity 0.32** 0.27*0.12 0.29*
6. Pretest VR interest 0.28*0.14 0.25*0.24 0.28*
7. Posttest VR interest 0.44*** 0.22 0.29*0.41*** 0.40** 0.46***
8. Pretest bridge interest 0.30 0.14 0.08 0.37*0.18 0.35*0.46**
9. Posttest bridge interest 0.55*** 0.38*0.32 0.47** 0.41*0.15 0.54*** 0.58***
Two-tailed Pearson correlations were calculated. *p < 0.05, **p < 0.01, ***p < 0.001.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 10 frontiersin.org
science and psychology students. is explanation is backed up by short
debrieng talks, in which multiple participants commented that they
had diculty getting excited about engineering topics.
Our study also provides new insights into inducing interest in
educational VR environments using 3D modeling and photogrammetry.
e interest growths encountered in our explorative analyses (see
Figure2) did not dier strongly and coherently between these two
content creation methods. A dive into the literature shows that this
nding is not surprising. Interest development is determined – apart
from individual predispositions – to a large extent by value and feeling
TABLE4 Regression analyses for posttest VR interest as outcome.
Predictor b ß ß 95% CI pEquation/Fit
Model V1 F(1,61) = 16.06, p< 0.001
Intercept 2.20 <0.001 R2= 0.21
Pretest VR interest 0.51 0.46 [0.23, 0.68] <0.001 Adj. R2= 0.20
Model V2 F(2,60) = 8.66, p< 0.001
Intercept 2.30 <0.001 R2= 0.22
Pretest VR interest 0.49 0.44 [0.22, 0.67] <0.001 Adj. R2= 0.20
Group −0.11 −0.13 [−0.35, 0.10] 0.277
Model V3 F(3,59) = 8.92, p< 0.001
Intercept 2.60 <0.001 R2= 0.31
Pretest VR interest 0.40 0.36 [0.13, 0.58] 0.002 Adj. R2= 0.28
Group −0.09 −0.10 [−0.32, 0.11] 0.339
Presence full 0.22 0.31 [0.08, 0.54] 0.008
Model V4 F(4,58) = 8.30, p< 0.001
Intercept 2.24 < 0.001 R2= 0.36
Pretest VR interest 0.34 0.30 [0.008, 0.53] 0.009 Adj. R2= 0.32
Group −0.13 −0.15 [−0.36, 0.07] 0.169
Presence full 0.17 0.24 [0.02, 0.47] 0.036
Representation delity 0.15 0.25 [0.02, 0.48] 0.033
b represents unstandardized regression weights. ß represents standardized regression weights. CI, condence interval.
TABLE5 Regression analyses for posttest bridge interest as outcome.
Predictor b ß ß 95% CI pEquation/Fit
Model B1 F(1,34) = 17.57, p< 0.001
Intercept 1.24 0.033 R2= 0.34
Pretest bridge interest 0.67 0.58 [0.30, 0.87] <0.001 Adj. R2= 0.32
Model B2 F(2,33) = 8.60, p< 0.001
Intercept 1.30 0.037 R2= 0.34
Pretest bridge interest 0.66 0.57 [0.28, 0.87] <0.001 Adj. R2= 0.30
Group −0.05 −0.04 [−0.34, 0.25] 0.761
Model B3 F(3,32) = 10.33, p< 0.001
Intercept 1.51 0.008 R2= 0.49
Pretest bridge interest 0.53 0.47 [0.19, 0.74] <0.001 Adj. R2= 0.44
Group 0.03 0.02 [−0.24, 0.29] 0.863
Presence full 0.41 0.41 [0.14, 0.68] 0.004
Model B4 F(4,31) = 8.23, p< 0.001
Intercept 0.92 0.211 R2= 0.52
Pretest bridge interest 0.52 0.45 [0.18, 0.72] 0.002 Adj. R2= 0.45
Group −0.02 −0.01 [−0.28, 0.26] 0.921
Presence full 0.32 0.32 [0.01, 0.63] 0.044
Representation delity 0.17 0.18 [−0.12, 0.48] 0.233
b represents unstandardized regression weights. ß represents standardized regression weights. CI, condence interval.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 11 frontiersin.org
judgments toward the content (Krapp, 2002; Hidi and Renninger,
2006). Probably, value and feeling judgments toward the content were
relatively similar in both groups as both educational virtual
environments possessed many shared features (e.g., the same content
on the video screens and in the visualizations) except for the physical
appearance of the main 3D object (i.e., the railroad bridge). It is an
open question, to what extent other characteristics of educational
virtual environments interact with interest development. In this regard,
the match between the functions educational virtual environments
provide and the situations they represent (Chernikova etal., 2020)
could be an important characteristic. Functionally more realistic
educational virtual environments could be associated with higher
interest development, particularly because they could increase value
and feeling judgments toward the content. Photogrammetry and VR
could beused in tandem with realistic tasks and interaction possibilities
to develop such educational virtual environments oering high physical
and functional realism.
4.3. Variables related to developing
situational interest (RQ3)
is paper also sought to gain insights into the variables related to
developing situational interest in the context of VR. As the majority of
results were similar for the outcomes of interest in VR for learning and
interest in bridges, wewill discuss these results together.
Our study showed that pretest situational interest was a predictor of
posttest situational interest. is nding is consistent with theoretical
reasoning put forward by interest development theories (Krapp, 2002;
Hidi and Renninger, 2006) and aligns well with empirical ndings on
interest trajectories conducted for other technology-enhanced
educational methods such as digital simulations and inquiry
environments (Tapola etal., 2013; Chen etal., 2016; Rodríguez-Aecht
et al., 2018). e content creation method of the educational VR
environment (3D modeling resp. photogrammetry) was, however, not a
signicant predictor of posttest situational interest. One explanation for
this result could be– as pointed out in Section 4.2 which addressed a
related topic – that both methods of content creation lead to comparable
value and feeling judgments (Krapp, 2002; Hidi and Renninger, 2006).
is could be particularly the case because both educational
environments oered participants the same tasks and information
despite their visual dierences.
Authenticity-related variables like presence and their association
with interest development were also examined. Presence explained
substantial variance in posttest situational interest. is nding
corresponds to an experiment by Petersen etal. (2022) on education in
VR which found that varying degrees of immersion aected interest
development. Our experiment goes beyond this experiment with two
points. First, it used presence as a broader operationalization of
authenticity than immersion. Second, our experiment showed that
presence explained posttest interest over and above the variance
accounted for by preexisting situational interest. To conclude, Petersen
etal.’s (2022) and our experiment identify presence as a vital variable to
consider when analyzing interest development in educational
environments. is link between presence and interest development in
VR can perhaps bebest explained by the positive regulatory and aective
aspects that may go along with the experience of presence (Krapp, 2002).
More specically, an increased feeling of presence can beaccompanied
by improved focus and positive aect states like ow. In our study, such
positive regulatory and aective aspects may have led to improved
interest development when experiencing higher levels of presence.
Our ndings on associations with representation delity dier for
the two interest outcomes. For interest in VR for learning, representation
delity was a signicant predictor. Moreover, the regression model
including representation delity explained more variance than the
model only including pretest interest, the group, and presence. is
result indicates that representation delity can beused as an additional
predictor of interest development in some contexts. For interest in
bridges, representation delity was, however, not a signicant predictor.
One explanation of this result could bethat representation delity does
not aect interest development directly but is- in line with Makransky
and Petersen (2021)- mainly an antecedent of presence. In support of
this point, wediscovered a positive correlation between representation
delity and the full presence score (r = 0.32).
4.4. Limitations
One limitation of our study is the restricted generalizability of the
reported ndings on the content creation methods. To critically discuss
this point, wewould like to revisit our experimental manipulation
briey. In both experimental groups, the same terrain and environment
created with 3D modeling was used. Between both experimental groups,
the content creation method was varied and the railroad bridge as the
main object of interest and center of attention was either created by 3D
modeling or photogrammetry. is decision was primarily made to
reach a high degree of experimental standardization across the groups.
Webelieve it resulted in experimentally comparable bridge models
which possessed few manifest dierences in the mesh, texture, and
lighting, but substantial objective dierences in these characteristics (see
Figure2). In contrast, it could beargued that particular features like the
grati contained in the photogrammetry group were seductive details,
as dened as “interesting but irrelevant adjuncts” (Harp and Mayer,
1998, p.414, p.414). is point is relevant as seductive details could
have potentially bolstered perceived authentiticy and interest
development in the photogrammetry group. e fact that authenticity
levels and interest increases in the control group and the
photogrammetry group were relatively comparable suggests that
seductive details did not play a crucial role in our study. Based on these
considerations, webelieve that our results can begeneralized to new
types of hybrid educational virtual environments that consist primarily
of 3D modeled terrain and environments and also include one or a few
objects created with photogrammetry. ese hybrid educational
environments are relatively comparable to the photogrammetry group
studied in this paper.
Another limitation of our study is the relatively small sample of
N = 64 participants. Developing this innovative study was intricate and
required considerable man-hours in programming, graphics design, and
creating photogrammetric models on top of regular research tasks.
Likewise, conducting the study was time-consuming as all participants
had to betested in one-on-one settings in a VR laboratory that had to
be set up each time. Due to the higher eort of developing and
conducting the study, it cannot beexpected that the study reaches the
same number of participants as questionnaire studies. Nevertheless, a
larger sample size would have been desirable for examining the
described interest and authenticity research questions. is point is
highlighted by the eect sizes our study reported and should
beconsidered by future studies on authenticity and interest in VR.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 12 frontiersin.org
Finally, it should beemphasized that the reported ndings were
obtained for educational VR environments displayed with HMDs. For
this reason, generalizations to other VR technologies should bemade
only with caution. CAVE sytems, in which participants experience a VR
in a dedicated room consisting of multiple screens, typically oer a
higher resolution and/or wider viewing angles than HMDs. erefore,
CAVE systems could represent educational environments in more detail
than HMDs, possibly evoking authenticity and interest dierences
between 3D modeling and photogrammetry as content
creation methods.
5. Conclusion and outlook
We investigated authenticity and interest development in an
educational VR environment. 3D modeling and photogrammetry
were employed as content creation methods to develop two different
educational virtual environments and then examined further. With
respect to authenticity, wediscovered that varying the content
creation methods did not evoke a substantial difference. This
finding implies that creating educational environments with 3D
modeling may besufficient for many educational contexts in which
participants mainly focus on the task and have no need to note fine
details of contained 3D objects. Regarding interest development,
there were two main findings. First, wediscovered that interest in
VR for learning increased from the pretest to the posttest. This
finding aligns well with other studies’ results and corroborates that
educational virtual environments, including VR with HMDs, can
trigger situational interest (Cordova and Lepper, 1996; Verhagen
etal., 2012; Siklander etal., 2017; Makransky etal., 2020). Second,
weexamined the variables associated with interest development and
identified presence as an important predictor that explained
substantial amounts of variance. This finding is consistent with a
result by Petersen et al. (2022) and can also be explained
theoretically. Perhaps the positive regulatory and affective aspects
that accompany presence (Krapp, 2002) also link this variable with
higher interest development.
Aer carrying out the study, weare convinced that photogrammetry
and 3D modeling both have unique benets and drawbacks and are
suitable for dierent use-cases of content creation in education. 3D
modeling can beparticularly suitable in contexts in which the user’s
attention is not on the object itself but on learning. Researchers and
practitioners, may, thus develop or acquire 3D modeled objects for such
use-cases and devote time and eort to the content, tasks, and support
of the participants. Photogrammetry, however, can beparticularly suited
to creating educational virtual environments in which ne details of
objects have to bevisualized. is technology also allows researchers
and practitioners who have no experience with graphic design to easily
create highly-detailed objects.
To wrap this paper up, wewill now take a brief look at the future of
VR and photogrammetry in education. VR enables users to make multi-
sensory, highly-immersive experiences. Within the next few years, VR
hard and soware will likely make further progress and become more
accessible. VR hardware, such as HMDs, will become more aordable,
compact, and overall more practical and may thus enable a better
integration within educational contexts. VR soware, such as simulations
and meta-verse applications, will become more authentic, interactive,
and meaningful. Webelieve that VR will then become an important part
of education, that may fulll particular goals, such as providing
participants with realistic contexts for improved problem-solving and
transfer as well as sparking interest. Photogrammetry has been primarily
part of surveying, but it will likely soon become more accessible and will
beused in other elds. Since photogrammetry enables the automatic and
time-ecient creation of 3D models, it can become an alternative to
manual 3D modeling with graphics soware. When creating educational
virtual environments, photogrammetry can beused to create life-like
objects and terrains. Our study illustrates that researchers and
practitioners can already use the two new technologies, VR and
photogrammetry, to create authentic and interesting educational virtual
environments. is nding is noteworthy because the two technologies
do not appear to have been used together in educational contexts before,
even though they t well together and complement each other.
Data availability statement
e raw data supporting the conclusions of this article will be made
available by the authors, without undue reservation.
Ethics statement
e studies involving human participants were reviewed and
approved by Ethics committee of Universität der Bundeswehr
München. e patients/participants provided their written informed
consent to participate in this study.
Author contributions
MF, DS, VE, and BEdesigned the experiment together. MF and DS
programmed and created the educational environments and used
stimuli. MF and BEpre-processed and analyzed the data. MF created
the rst dra of the manuscript. VE wrote individual sections of the
abstract and theory. DS, VE, and BEprovided feedback as well as critical
revisions of the manuscript. All authors read and approved the nal
version of the manuscript for submission.
Funding
is research paper is funded by dtec.bw – Digitalization and
Technology Research Center of the Bundeswehr [project RISK.twin].
We acknowledge nancial support by Universität der
Bundeswehr München.
Acknowledgments
We thank Johannes Wimmer for his advice and support in creating
the concept of the educational environment. Moreover, weare grateful
to Fabian Seitz for reviewing and correcting the materials of the
educational virtual environment. Credit also goes to Maximilian Gründl
who served as a consultant in creating the photogrammetric model of
the railroad bridge. MF thanks Larissa Kalteeiter for her advice
and support.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 13 frontiersin.org
Conflict of interest
e authors declare that the research was conducted in the absence
of any commercial or nancial relationships that could beconstrued as
a potential conict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their aliated organizations,
or those of the publisher, the editors and the reviewers. Any
product that may be evaluated in this article, or claim that may be
made by its manufacturer, is not guaranteed or endorsed by
the publisher.
Supplementary material
e Supplementary material for this article can befound online at:
https://www.frontiersin.org/articles/10.3389/feduc.2023.969966/full#
supplementary-material
References
American Society for Photogrammetry and Remote Sensing (2022). “What is ASPRS?”
http://www.asprs.org/a/society/about.html.
Bolkan, S., and Grin, D. J. (2018). Catch and hold: instructional interventions and their
dierential impact on student interest, attention, and autonomous motivation. Commun.
Educ. 67, 269–286. doi: 10.1080/03634523.2018.1465193
Chen, J. A., Tutwiler, M. S., Metcalf, S. J., Kamarainen, A., Grotzer, T., and Dede, C. (2016).
A multi-user virtual environment to support students' self-ecacy and interest in science.
A latent growth model analysis. Learn. Instr. 41, 11–22. doi: 10.1016/j.learninstruc.2015.09.007
Chernikova, O., Heitzmann, N., Stadler, M., Holzberger, D., Seidel, T., and Fischer, F.
(2020). Simulation-based learning in higher education: a meta-analysis. Rev. Educ. Res. 90,
499–541. doi: 10.3102/0034654320933544
Cordova, D. I., and Lepper, M. R. (1996). Intrinsic motivation and the process of
learning: benecial eects of contextualization, personalization, and choice. J. Educ.
Psychol. 88, 715–730. doi: 10.1037/0022-0663.88.4.715
Dalgarno, B., Hedberg, J., and Harper, B. (2002). “e contribution of 3D environments
to conceptual understanding” in Winds of change in the sea of learning: Proceedings of the
19th annual conference of the australasian society for computers in learning in tertiary
education. ed. O. J. McKerrow (Auckland: UNITEC, Institute of Technology), 149–158.
Dalgarno, B., and Lee, M. J. W. (2010). What are the learning aordances of 3-D virtual
environments? Br. J. Educ. Technol. 41, 10–32. doi: 10.1111/j.1467-8535.2009.01038.x
Deci, E. L., Vallerand, R. J., Pelletier, L. G., and Ryan, R. M. (1991). Motivation and
education: the self-determination perspective. Educ. Psychol. 26, 325–346. doi:
10.1080/00461520.1991.9653137
Epic Games (2021). RealityCapture. Available at: https://www.capturingreality.com/
Erolin, C. (2019). “Interactive 3D digital models for anatomy and medical education,” in
Biomedical visualization: Advances in experimental medicine and biology. ed. P. M. Rea
(Cham: Springer), 1–16. doi: 10.1007/978-3-030-14227-8_1
Faro Technologies (2022). Available at: https://www.faro.com/
Harackiewicz, J. M., Smith, J. L., and Priniski, S. J. (2016). Interest matters: the
importance of promoting interest in education. Policy Insights Behav. Brain Sci. 3, 220–227.
doi: 10.1177/2372732216655542
Harp, S. F., and Mayer, R. E. (1998). How seductive details do their damage: a theory of
cognitive interest in science learning. J. Educ. Psychol. 90, 414–434. doi:
10.1037/0022-0663.90.3.414
Hidi, S. (1990). Interest and its contribution as a mental resource for learning. Rev. Educ.
Res. 60, 549–571. doi: 10.3102/00346543060004549
Hidi, S., and Renninger, K. A. (2006). e four-phase model of interest development.
Educ. Psychol. 41, 111–127. doi: 10.1207/s15326985ep4102_4
Historic England (2017). Photogrammetric Applications for Cultural Heritage: Guidance
for Good Practice. Swindon: Historic England.
Historic England (2018). 3D Laser Scanning for Heritage: Advice and Guidance on the Use
of Laser Scanning in Archaeology and Architecture. Swindon: Historic England.
HTC (2016). VIVE. Available at: https://www.vive.com/us/support/vive/
Jensen, L., and Konradsen, F. (2018). A review of the use of virtual reality head-mounted
displays in education and training. Educ. Inf. Technol. 23, 1515–1529. doi: 10.1007/
s10639-017-9676-0
Kpolovie, P. J., Joe, A. I., and Okoto, T. (2014). Academic achievement prediction. Role
of interest in learning and attitude towards school. Int. J. Humanit. Soc. Sci. Educat. 1,
73–100.
Krapp, A. (2002). Structural and dynamic aspects of interest development. eoretical
considerations from an ontogenetic perspective. Learn. Instr. 12, 383–409. doi: 10.1016/
S0959-4752(01)00011-1
Lee, E.-L., Wong, K. W., and Fung, C. C. (2010). How does desktop virtual reality
enhance learning outcomes? A structural equation modeling approach. Comput. Educ. 55,
1424–1442. doi: 10.1016/j.compedu.2010.06.006
Makransky, G., and Petersen, G. B. (2019). Investigating the process of learning with
desktop virtual reality: a structural equation modelling approach. Comput. Educ. 134,
15–30. doi: 10.1016/j.compedu.2019.02.002
Makransky, G., and Petersen, G. B. (2021). e Cognitive Aective Model of Immersive
Learning (CAMIL): a theoretical research-based model of learning in immersive virtual
reality. Educ. Psychol. Rev. 33, 937–958. doi: 10.1007/s10648-020-09586-2
Makransky, G., Petersen, G. B., and Klingenberg, S. (2020). Can an immersive virtual
reality simulation increase students’ interest and career aspirations in science? Br. J. Educ.
Tec hn ol . 51, 2079–2097. doi: 10.1111/bjet.12954
Makransky, G., Terkildsen, T. S., and Mayer, R. E. (2019). Adding immersive virtual
reality to a science lab simulation causes more presence but less learning. Learn. Instr. 60,
225–236. doi: 10.1016/j.learninstruc.2017.12.007
McLellan, H. (2008). “Virtual realities,” in Handbook of Research for Educational
Communications and Technology: A Project of the Association for Educational
Communications and Technology. ed. D. H. Jonassen (New York: Taylor & Francis),
461–498.
McWhaw, K., and Abrami, P. C. (2001). Student goal orientation and interest: eects on
students' use of self-regulated learning strategies. Contemp. Educ. Psychol. 26, 311–329. doi:
10.1006/ceps.2000.1054
Nebel, S., Beege, M., Schneider, S., and Rey, G. D. (2020). A review of photogrammetry
and photorealistic 3D models in education from a psychological perspective. Front. Educ.
5:144. doi: 10.3389/feduc.2020.00144
Pellas, N., Dengel, A., and Christopoulos, A. (2020). A scoping review of immersive
virtual reality in STEM education. IEEE Trans. Learn. Technol. 13, 748–761. doi: 10.1109/
TLT.2020.3019405
Petersen, G. B., Klingenberg, S., Mayer, R. E., and Makransky, G. (2020). e virtual eld
trip: investigating how to optimize immersive virtual learning in climate change education.
Br. J. Educ. Technol. 51, 2099–2115. doi: 10.1111/bjet.12991
Petersen, G. B., Petkakis, G., and Makransky, G. (2022). A study of how immersion and
interactivity drive VR learning. Comput. Educ. 179:104429. doi: 10.1016/j.compedu.2021.104429
Radianti, J., Majchrzak, T. A., Fromm, J., and Wohlgenannt, I. (2020). A systematic review
of immersive virtual reality applications for higher education: design elements, lessons
learned, and research agenda. Comput. Educ. 147:103778. doi: 10.1016/j.compedu.2019.103778
Renkl, A., Mandl, H., and Gruber, H. (1996). Inert knowledge: analyses and remedies.
Educ. Psychol. 31, 115–121. doi: 10.1207/s15326985ep3102_3
Renninger, K., Ewen, L., and Lasher, A. (2002). Individual interest as context in
expository text and mathematical word problems. Learn. Instr. 12, 467–490. doi: 10.1016/
S0959-4752(01)00012-3
Rodríguez-Aecht, G., Jaakkola, T., Pongsakdi, N., Hannula-Sormunen, M.,
Brezovszky, B., and Lehtinen, E. (2018). e development of situational interest during
a digital mathematics game. J. Comput. Assist. Learn. 34, 259–268. doi: 10.1111/jcal.12239
Rotgans, J. I., and Schmidt, H. G. (2011). e role of teachers in facilitating situational interest
in an active-learning classroom. Teach. Teach. Educ. 27, 37–42. doi: 10.1016/j.tate.2010.06.025
Schiefele, U. (1991). Interest, learning, and motivation. Educ. Psychol. 26, 299–323. doi:
10.1080/00461520.1991.9653136
Schiefele, U., Krapp, A., and Winteler, A. (1992). “Interest as a predictor of academic
achievement: a meta-analysis of research” in e Role of Interest in Learning and
Development. eds. K. A. Renninger, S. Hidi and A. Krapp (Hillsdale: Erlbaum), 183–212.
Schubert, T., Friedmann, F., and Regenbrecht, H. (2001). e experience of presence.
Factor analytic insights. Presence. 10, 266–281. doi: 10.1162/105474601300343603
Schutte, N. S. (2020). e impact of virtual reality on curiosity and other positive
characteristics. Int. J. Hum. Comput. Interact. 36, 661–668. doi: 10.1080/10447318.2019.1676520
Siklander, P., Kangas, M., Ruhalahti, S., and Korva, S. (2017). “Exploring triggers for
arousing interest in the online learning,” in INTED2017 Proceedings: 11th International
Technology, Education and Development Conference, ed. L. Gómez Chova, A. López
Martínez and I. Candel Torres Valencia (IATED), 9081–9089.

Fink et al. 10.3389/feduc.2023.969966
Frontiers in Education 14 frontiersin.org
Škola, F., Rizvić, S., Cozza, M., Barbieri, L., Bruno, F., Skarlatos, D., et al. (2020). Virtual
reality with 360-video storytelling in cultural heritage: study of presence, engagement, and
immersion. Sensors 20:5851. doi: 10.3390/s20205851
Skulmowski, A., Nebel, S., Remmele, M., and Rey, G. D. (2021). Is a preference for
realism really naive aer all? A cognitive model of learning with realistic visualizations.
Educ. Psychol. Rev. 34, 649–675. doi: 10.1007/s10648-021-09638-1
Tapola, A., Veermans, M., and Niemivirta, M. (2013). Predictors and outcomes of situational
interest during a science learning task. Instr. Sci. 41, 1047–1064. doi: 10.1007/s11251-013-9273-6
e Blender Foundation (2021). Blender. Available at: https://www.blender.org/
Trebuňa, P., Mizerák, M., Trojan, J., and Rosocha, L. (2020). 3D scanning as a modern
technology for creating 3D models. Acta Technol. 6, 21–24. doi: 10.22306/atec.v6i1.74
Unity Technologies (2020). Unity engine. Available at: https://unity.com/
Unity Technologies (2021). XR Interaction Toolkit. Available at: https://docs.unity3d.
com/Packages/com.unity.xr.interaction.toolkit@0.9/manual/index.html
Velev, D., and Zlateva, P. (2017). Virtual reality challenges in education and training. Int.
J. Learn. Teach. 3, 33–37. doi: 10.18178/ijlt.3.1.33-37
Verhagen, T., Feldberg, F., van den Hoo, B., Meents, S., and Merikivi, J. (2012).
Understanding users’ motivations to engage in virtual worlds: a multipurpose model
and empirical testing. Comput. Hum. Behav. 28, 484–495. doi: 10.1016/j.chb.2011.10.020
Vogt, A. G. (2021). Fostering deep learning in immersive virtual reality: Interplay of
learner’s characteristics with internal and external support. dissertation. Universität Ulm.
Available at: https://oparu.uni-ulm.de/xmlui/bitstream/handle/123456789/44276/
Dissertation_Vogt.pdf?sequence=1&isAllowed=y
Waller, D., Hunt, E., and Knapp, D. (1998). e transfer of spatial knowledge in virtual
environment training. Presence 7, 129–143. doi: 10.1162/105474698565631
Witmer, B. G., and Singer, M. J. (1998). Measuring presence in virtual environments: a
presence questionnaire. Presence 7, 225–240. doi: 10.1162/105474698565686
Woerner, J. J. (1999). “Virtual eld trips in the earth science classroom,” in Proceeding
of the annual conference of the Association for the Education of Teachers in Science, ed.
P. A. Rubba, J. A. Rye and P. F. Keig (Greenville, NC), 1232–1244.
Wu, B., Yu, X., and Gu, X. (2020). Eectiveness of immersive virtual reality using head-
mounted displays on learning performance: a meta-analysis. Br. J. Educ. Technol. 51,
1991–2005. doi: 10.1111/bjet.13023