Analogy between Student Perception of Educational Space Dimensions and Size Perspective in 3D Virtual Worlds versus Physical World

Abstract

One of the prominent practices currently associated with 3D virtual worlds, such as Second Life, is their increased utilization as 3D virtual learning environments (3D VLEs). This study is part of a research in progress dedicated to evaluate different engineering design aspects of these emergent VLEs, and define the impact of their design features on delivering online education. The aim of this paper is to investigate and analogize between users’ perception of space in virtual worlds compared to its corresponding perception in the physical world in terms of area size, dimensions and overall 3D visual perspective. This is achieved by recording the visual estimations of different student categories, within diverse 3D virtual sites, in response to survey questions depicting space size and capacity for holding students and hosting e-learning sessions. Furthermore, the differences in student responses are analyzed and elucidated in order to formulate a hypothesis about how similar or dissimilar users perceive spaces in 3D virtual worlds in comparison with the physical world.

Full text

Page 1

Noha Saleeb & Georgios Dafoulas
International Journal of Engineering (IJE), Volume (4), Issue (3) 210
Analogy between Student Perception of Educational Space
Dimensions and Size Perspective in 3D Virtual Worlds versus
Physical World


Noha Saleeb
School of Engineering and Information Sciences
Middlesex University
London, UK

Georgios Dafoulas
School of Engineering and Information Sciences
Middlesex University
London, UK
N.Saleeb@mdx.ac.uk
G.Dafoulas@mdx.ac.uk

Abstract

One of the prominent practices currently associated with 3D virtual worlds, such
as Second Life, is their increased utilization as 3D virtual learning environments
(3D VLEs). This study is part of a research in progress dedicated to evaluate
different engineering design aspects of these emergent VLEs, and define the
impact of their design features on delivering online education. The aim of this
paper is to investigate and analogize between users’ perception of space in
virtual worlds compared to its corresponding perception in the physical world in
terms of area size, dimensions and overall 3D visual perspective. This is
achieved by recording the visual estimations of different student categories,
within diverse 3D virtual sites, in response to survey questions depicting space
size and capacity for holding students and hosting e-learning sessions.
Furthermore, the differences in student responses are analyzed and elucidated in
order to formulate a hypothesis about how similar or dissimilar users perceive
spaces in 3D virtual worlds in comparison with the physical world.

Keywords: visual perception in 3D virtual worlds, virtual learning environments, educational facilities in
Second Life, class capacity in e-learning spaces.


1. INTRODUCTION
Since the onset of 3D virtual worlds, whether used for gaming purposes or as learning
environments, 3D designers and builders have strived to create virtual constructions within them
that have proved to be both innovative and imaginative but also comfortably familiar for the user
[1]. This flourishing in 3D virtual design has been the result of the vast disparity between the
physical world and virtual worlds in terms of diminished constraints to free design [2], for as
previously asserted by Bourdakis and Charitos [3], the nature of space in virtual environments
(VEs) is fundamentally different from the nature of real space and thus subsequently the
architecture of VEs requires new theory and practice. Examples of these fundamental differences
include the non-presence of gravity, material and budget restraints, which have given rise to
many known and novel building styles in VEs such as Photo-realistic (identical replica of existing

Page 2

Noha Saleeb & Georgios Dafoulas
International Journal of Engineering (IJE), Volume (4), Issue (3) 211
in reality), artistically-realistic (similar to existing in reality), functionally-realistic (has no equivalent
in reality but is realistically designed), metaphorically-realistic (entails realistic functions), hybrid
(mixture of realistic and imaginative design), fantasy (imaginative design defying reality), and
abstract (ambiguous design) [4].

In 3D virtual places, designers are increasingly faced with higher-degree spatial organization than
in the physical world, comprising the cognitive relationship between content and space [1].
Cognition is a process proclaimed on the user’s sensory-motor and neurological systems. The
process of visual acquiring, assimilation and interpretation of environmental information is called
cognitive mapping to understand the relationship between the objects in a space [5]. Therefore,
since Downs and Stea [6] denote that “human spatial behavior is dependent on the individual’s
cognitive map of the spatial environment”, this indicates that a user’s perception of the virtual
space within a 3D VE can control his conduct within this virtual environment. This would
accordingly also imply that students’ perception of their learning spaces in 3D VEs would hence
affect their behavior inside them. It is thus the focus of this paper to investigate how students’
perception of 3D virtual e-learning spaces differs from their perception of physical learning spaces
in an attempt to explore whether this affects their overall learning process. Results of this
research can subsequently help educators and designers in VEs to enhance the architectural
design of virtual 3D learning spaces in VLEs to be more suitable for students’ e-learning within
them.
2. BACKGROUND
The effect of physical spaces on students’ learning in general has been amply asserted in
previous literature. Oladipupo and Oyelade [7] state that “there is more to students’ failure than
the students’ ability”. According to Kenchakkanavar and Joshi [8], incompatibility of classrooms
for teaching was one of the factors affecting student failures in their courses. Furthermore it has
been demonstrated that classes smaller than 900 sq. ft. in area are undesirable as they do not
allow for adequate movement between tables without bumping into students and their belongings;
crowded classrooms contribute to discipline problems [9]. However if a 900 sq. ft. class is built
inside a VE, will the students perceive it as the same size as in the physical world, or smaller or
larger, and thus will it be adequate for their needs? Moreover, narrow hallways that are too small
for student traffic between classes have been found to encourage fighting and hinder evacuation
in emergencies [10]. Again here while corridors of 2m width might be acceptable in the physical
world, would this width be perceived as sufficient in the virtual world? It is therefore imperative in
the case of 3D virtual learning environments to inquire into how a student identifies with the
surrounding spaces, perceives dimensions, shape, and perspective and how that is different from
perceiving the totality of spaces in the physical world. This realization is essential since if
differences prevail between the virtual and physical worlds in perception of space size, then this
necessitates a change in the engineering codes and guidelines used by educators, designers,
builders and architects to build inside 3D VLEs to counteract for these differences in perception.

Hence, in agreement with Lau and Maher [1], orienting users within efficiently designed spaces in
a virtual environment requires a “detailed study of environmental cognition”. Cognition and
visualization involve graphic rendering of data in such a way to take advantage of the human
ability to recognize patterns and see structures [11]. To understand how these cognitive principles
can be applied to the design of VEs, experiments with users, namely students in this study, are
required to capture students’ different perceptions of the 3D spaces they experience during their
e-learning sessions. To pursue this notion, it is necessary to initially differentiate between the
different types of user viewpoints available within 3D VLEs. While participants have the capability
to observe the environment from many perspectives [12], there are two basic types of perspective
viewpoints in 3D VLEs: i) virtual reality perspective and ii) virtual world perspective [13]:

Virtual reality can be defined as an environment created by the computer in which the user feels
immersed perceptually and psychologically in the digital environment [14]. The main difference
between virtual reality and virtual world viewpoint is the way the user experiences the virtual

Page 3

Noha Saleeb & Georgios Dafoulas
International Journal of Engineering (IJE), Volume (4), Issue (3) 212
environment. If the VE is experienced through the first person i.e. seeing the world through the
eyes of the avatar and surrounded by the environment, then this is virtual reality [15]. This
perspective or viewpoint can be achieved by manipulating camera controls in the VE (or wearing
head mounted display devices) and is the closest to “real life” physical perspective. In contrast,
virtual world view allows the user to see the VE in 3rd person by watching the avatar move at a
distance inside the VE [13]. These differences result in different spatial cognition by the users
[16].

The focus of this paper is to investigate the difference in student perception for 3D virtual space
size and dimensions, versus the “real-life” physical perception, using the “virtual world” (3rd
person) perspective explained beforehand.
3. RESEARCH RATIONALE AND METHODS
The “virtual world” perspective tested in this study is the default viewpoint utilized within 3D VLEs
in general and the more commonly used among students for navigation in 3D VLEs. In order to
capture the difference between students’ perception of space size between the virtual world and
the physical world, the subsequent research rationale and techniques were followed:

Several randomized samples of students from different categories (elaborated henceforth) were
asked to participate in short consecutive e-learning sessions inside 15 selected 3D virtual
learning spaces, inside which students were encouraged to navigate, using the “virtual world”
viewpoint, to assimilate the extent of the space size by being immersed inside each (explained
henceforth). At the end of the time spent inside each virtual site, the students were all asked to
record how many users they perceived this space could hold by choosing from a list of
predetermined ranges, also described consequently. Other closed ended questions were asked
of the students related to assessing more engineering and architectural design elements of the
space, but which are not the focus of this paper at hand. The numerical results offered by the
students concerning their perceptions were then averaged for each 3D virtual site used, and
these results were compared to the actual number of students that each site would actually hold if
built in the physical world with the exact same dimensions. This comparison was used to identify
whether space in virtual digitized worlds is recognized by users as being the same size as that in
reality or larger or smaller.

The study was conducted in Second Life as a representative of 3D Virtual Learning Environments
for its popularity among universities and educational institutions for delivering e-learning [17]. The
samples of consenting participants in this study were 84 students from the School of Engineering
and Information Sciences at Middlesex University. They were divided into 31 under graduate
students, 33 post graduate students, and 20 members of staff representing adult learners. The
participants were diverse in gender and cultural background. Results taken from all 3 categories
of students were analyzed comparatively and relevant conclusions were drawn accordingly.

The 15 selected 3D virtual learning spaces were chosen to represent a diverse number of
variations in space design characteristics in terms of:
 space shape (e.g. circular, rectangle, square)
 size (e.g. small, medium, large – criteria for size naming explained hereafter)
 dimension ratio (width:length:height e.g. 2:2:1)
 openness of space (i.e. whether space is confined by walls or not)
This variety in choice was essential in order to identify if there were any prominent architectural
design factors affecting student perception of spaces in 3D VLEs.
The ranges of answers that the students were asked to answer from included:
 space can withstand: 10 - 30 students
 space can withstand: 40 - 60 students
 space can withstand: 70 - 100 students or more

Page 4

Noha Saleeb & Georgios Dafoulas
International Journal of Engineering (IJE), Volume (4), Issue (3) 213
The above three answering criteria were determined based on real-life physical classroom
classifications where i) classrooms are considered “small” size learning spaces with capacity up
to 28 students. If 2.5 – 3m2 is required per student, then the average area of a classroom would
be 25-75m2. Educational spaces with similar area size in Second life were used as examples of
small learning spaces. ii) Seminar rooms are considered “medium” size learning spaces with
capacity of 45-60 students. If 2.2-2.6 m2 is required per student in a seminar room, then the
average area would be 100-150m2. Educational spaces with similar area size in Second life were
used as examples of medium learning spaces. iii) Lecture auditoriums are considered “large” size
learning spaces with capacity of up to 200 students. If 1.6-2 m2 is required per student in a lecture
hall, then the average area would be greater than 150m2. Educational spaces with similar area
size in Second life were used as examples of large learning spaces [18] [19]. Learning spaces
holding over 100 students were rare in Second Life since the current servers’ capabilities cannot
withstand more than this number of logged in users at the same time on the same site. Each
learning space used within this study was also classified as “open” if it did not contain
surrounding walls.

The equation used to calculate the average perceived number of users by students from each
category for each site was:
( (no. of “10-30” votes * 30) +
(no. of “40-60” votes * 60) +
(no. of “70-100” votes * 100) ) / Total number of participants

As an additional analysis, the standard deviation between the results of under graduate students,
post graduate students and adult learners was also calculated to find the discrepancy between
the values and how this may be related to different types of educational space shapes, sizes and
dimensions. The standard deviation measures the spread of the data around the mean value and
thus how widely dispersed they are from the maximum to the minimum value. The larger the
value of the standard deviation the more this implies that the individual data points are farther
from the average value. To calculate the standard deviation, the mean value is first calculated.
Next, the deviation of each data point from the average is calculated by subtracting its value from
the mean value. Each deviation is squared, and the individual squared deviations are averaged
together. The resulting value is known as the variance. Standard deviation is the square root of
the variance [20].
Diagrams illustrating the different findings were created accordingly, as demonstrated in the
subsequent sections.

4. RESULTS
Results in this paper were identified in three areas in accordance with the aims and focus of this
study:
 A comparison between the results obtained from under graduate, post graduate and adult
learners concerning their perception of the number of users that each 3D virtual learning
space in consideration can withstand.
 An analogy to compare between the overall students’ average perceived number of users
for each site (and thus what area size is implied for that site) versus the actual number of
users that could be withheld if this learning space was built with the exact same
dimensions in the physical world.
 Calculate the standard deviation between the results of the three student categories, for
each 3D virtual site, to find out factors affecting different perceptions by students.

4.1 Capacity of Users Perceived for each 3D Virtual Learning Space

Page 5

Noha Saleeb & Georgios Dafoulas
International Journal of Engineering (IJE), Volume (4), Issue (3) 214


FIGURE 1: Average number of users that can be withheld in each site as perceived by different categories
of students

The former Figure 1 illustrates the perceived number of users that each category of students
(under graduate, post graduate and adult learners) estimated can be withheld inside each of the
15 3D virtual learning spaces selected for this study.

As evident from the figure, it can be clearly noticed that average number of users perceived by
under graduate students for the different sites in general tends to be lower than results depicted
by the other two categories of students for all sites. Even more, while demonstrated results for
post graduate students are higher than those for under graduate students, they are still lower
than those offered by adult learners, who give the highest capacity of students for all sites. This
can provide a general trend where the older the age category, the larger the students’ perception
is of the size of the 3D virtual learning space and the capacity of users it can hold. The
implications of these findings are to be discussed in the conclusions.


4.2 Average Perceived Number of Users and Size of each Site compared to the Actual
Space Size
The average perceived numbers of users for each site demonstrated in Figure 2 denote the mean
values for all three combined categories of students for each 3D virtual site. It can be evidently
seen that the students’ overall estimation for the number of users withheld in each site (and thus
also approximation for the size of the space) is very similar to the actual number and size ranges
of each space. Small, medium and large spaces were correctly identified by students by correctly
estimating the number of users that should be within each space. This result holds true despite
the differences in the learning space shape, dimensions ratio, and openness of walls
differentiating the architectural design of the 3D virtual spaces from each other. The implications
of this result are to be discussed consequently.