Education and Information Technologies 5:4 (2000): 277±289

# 2000 Kluwer Academic Publishers, Manufactured in The Netherlands

Perfect presence: What does this mean for the design of

virtual learning environments?
DENISE WHITELOCK*
Institute of Educational Technology, The Open University, Walton Hall, Milton Keynes, MK7 6AA. * Correspond-
ing author. E-mail: d.m.whitelock@open.ac.uk

DANIELA ROMANO
Computer Based Learning Unit, University of Leeds. E-mail: d.m.romano@cbl.leeds.ac.uk

ANNE JELFS
Behavioural Studies, University College Northampton, Boughton Green Road, Northampton NN2 7AL. E-mail:
anne.jelfs@northampton.ac.uk

PAUL BRNA
Computer Based Learning Unit, University of Leeds. E-mail: p.brna@cbl.leeds.ac.uk

One of the advantages of building a virtual reality system is that it allows students to enter new worlds which in
these instances include trips to the bottom of the Atlantic Ocean, a ®eld visit to an Oak Wood and a close
encounter in a 3D maze. In all these environments the factors affecting a sense of `being there' or presence was
investigated. Enhanced audio feedback increased a subjective sense of presence but did not increase students'
conceptual learning scores. We have also found that a sense of social presence enhanced the notion of `being
there' together with measures of collaboration. However `being there' can take its toll on students and our ®ndings
suggest it imposes a cognitive overload. Where students have a choice, they try and reduce this overload by asking
for conceptual tools to assist them in their learning tasks. The studies reported in this paper provide some
benchmark data about these issues which deserve further investigation if we are to design effective virtual
environments for conceptual learning.

Keywords: virtual reality; presence; conceptual learning; collaborative virtual environments; conceptual tools.

Introduction

Virtual Reality systems offer users exciting opportunities to enter new worlds. They no
longer have to be passive spectators but can experience and manipulate these virtual worlds
in a number of novel ways. For software designers, these worlds can mean constructing
anything from a games environment which can be placed on a mountain slope or located
on an imaginary island inhabited by some rather strange creatures, to the implementation
of a more serious training system. Indeed Virtual Reality systems that have been designed
to represent `real' environments, such as the British Aerospace virtual cockpit (Kalawsky,
1991a; Kalawsky, 1991b), have a proven successful track record for training pilots.
Multimedia specialists have also been called upon to design more educational worlds;
278 WHITELOCK ET AL

which transport students to some rather exotic ®eld trips, such as to the bottom of the
ocean ¯oor, exploring the North Atlantic Ridge (Whitelock, 1999).
One of the driving forces behind the building of these types of Virtual Reality
environments was to create a strong sense of presence or of `being there' for the user. It
has been argued that the tasks then become more authentic and better training outcomes
can be realised. However another bonus for students working with these Virtual Reality
systems is that they can provide a more stimulating and motivating teaching environment
where learning can be both challenging and, at the same time, fun.
We have all been involved in the design and evaluation of a number of Virtual Reality
systems. (Whitelock, 1999; Brna, 1999; Romano & Brna, 2000; Jelfs & Whitelock, 1999).
The Open University group has concentrated on producing Virtual Reality environments
for Undergraduate Science students, while the Leeds team have more recently focused on
producing software to train professional ®re-®ghters. One of our main objectives has been
to create a comprehensive sense of presence in the worlds we have created, but also to give
the users access to another world, a conceptual or `theory world'. This Paper discusses the
®ndings from both our projects with respect to creating a sense of presence of being there
and how this affects the user, together with how this factor impinges upon the user's
conceptual understanding of the contents of these Virtual Reality systems.

Creating a sense of presence

The notion of `presence' is considered to be an important conceptual component of any

Virtual Environment whether it is immersive or desktop. So what is presence? Presence is
where we are immersed in a very high bandwidth stream of sensory input, organised by our
perceiving systems, and out of this `bath' of sensation emerges our sense of being in and of
the world.
Presence does not just refer to one's surrounding as they exist in the physical world, but
to the perception of those surrounding (Steuer, 1992). Whereas Steuer refers to tele-
presence as the extent to which one feels present in the mediated environment, rather than
the immediate physical environment. This means that the dependent variables of virtual
reality must all be measures of individual experience. Applying this course of action
should provide an obvious means of applying knowledge about perceptual processes and
individual differences in determining the nature of virtual reality (Steuer, 1992). This point
has important rami®cations for interface design success. A well designed piece of software
makes the interface transparent. For Virtual Reality systems the more an individual is
aware of the interface then the harder it will be to achieve a high level of telepresence. To
lessen the awareness of the interface, there needs to be increased level of presence.

The role of presence and conceptual understanding

In the case of the ¯ight simulator, there is as true a sense of being in the real thing as is
possible. The instrument and commands used in ¯ight simulations are a perfect replica of
those applied in the real world. But in other educational settings it might not be possible or
PERFECT PRESENCE 279

economical to use a perfect replica of the environment. This raises a set of questions. One
of these is how can one maximise a sense of being there for students working with desktop
virtual environments? Secondly, does this sense of being there or presence promote
conceptual understanding? Since in education we are primarily concerned with teaching
students to look beyond their senses to more scienti®c predictive models of the world. We
also need to motivate them to enter this `theory world' by building teaching environments
that are fun to use and where time seems to pass quickly. Understanding how to create this
balance is one of the goals of educational media production. Together with providing
`learning as fun' environments since Virtual Reality systems offer students trips to wild
and exciting areas of the world, for example, the ocean ¯oor.
One way which apparently evokes a greater level of presence, notwithstanding the
quality of the graphics, is through the use of sound. Sound gives feedback to the user and
offers greater levels of reality. Laurel (1993) refers to the use of sound to evoke emotional
responses, whereas Rheingold (1991) refers to sound as valuable for feedback, particularly
3D acoustics. As he says: `there's nothing like the sound of footsteps behind you to help
convince you that you are in a dark alley late at night in a bad part of town ± sounds have
the ability to raise the hairs on the back of our neck' p. 151. One of the important
components of presence stressed by a number of researchers appears to be good sound
feedback. However, will an emphasis on sound be important for conceptual learning?
Learning and `sense making' is such a complicated process that one could not argue that
an increased sense of presence per se in a Virtual Reality system world in itself offers a
conceptual advantage. That is why the present authors, Whitelock and Brna together with
Holland, have proposed an abstract framework of these properties from which to evaluate
Virtual Reality systems (Whitelock et al., 1996).
In initial work, Whitelock, Brna and Holland developed a `cube' of person-centred
properties ± representational ®delity, immediacy of control (a measure of the percep-
tion=action loop) and presence. It is in part because Whitelock et al's approach is
person centred, that allows it to be generalised to real world simulations as well as to
two dimensional desktop ones.
Brie¯y, representational ®delity is subdivided into: technical ®delity; representational
familiarity; and representational reality. Technical ®delity is the degree to which the
technology delivers realistic renderings, colours, textures, motion etc. Representational
familiarity is the extent to which the environment that is simulated is familiar to the user.
An unfamiliar world might be a simulation of the `surface' of Jupiter. Representational
reality is the extent to which the world is possible.
Immediacy of control is affected by the medium through which control is channelled.
The use of hand motions close to those used in the real world to achieve a corresponding
`real' effect illustrates an aspect of near perfect immediacy ± provided that the system is not
subject to excessive latency. At the other end of the spectrum, communicating instructions
through a command line interface is an example of low immediacy. Intermediate positions
are possible depending, for example, on how much of the hand's ¯exibility is supported for
control purposes.
One way of extending Whitelock's Cube is to take into account the social space in which
people are `immersed'. This in turn requires the development of a way of describing the
interactions with other participants (Brna, 1999).
280 WHITELOCK ET AL

Social factors need to be included in a hypercube if we wish to use the resulting

framework both to guide the design of Collaborative Virtual Learning Environments
(CVLEs) and to examine the use of a CVLE. The term social factors is taken to refer to the
interactions with a group of fellow participants in the environment which includes both
communicative acts (speech, writing, pointing etc.) and the social relationships that hold
(classmate, student-teacher, collaborator, team leader etc.).
The social factors described by Brna parallel the axes of Whitelock's Cube. These axes
are named social ®delity, immediacy of discourse and social presence.
Social ®delity parallels the notion of representational ®delity. It too can be subdivided
into: social familiarity; social reality and embodiment, a set of issues relating to the
manner in which the self and others in the environment are represented.
Immediacy of discourse is affected by the social medium (voice, gestures, touch etc.)
through which human communication is channelled. It is closely related to the effective-
ness of the dialogue between participants, and to the degree to which the dialogue between
participants is interlocked.
Social presence is taken to be a measure of `social richness' along with a sense of being
a `social actor within a medium' (Lombard & Ditton, 1997). Social richness is related to
the number and degree of saturation of human ± human communication channels.
The hypercube covers issues connected with both the physical environment and the
social environment but one element has been neglected: the affective=cognitive=sensori-
motor aspects of the individual participant. This is not taken further here.
Whitelock's Cube provides a way of examining VLEs with a view to testing a number of
hypotheses about interactions between the three different fundamental properties and the
suitability of a VLE for conceptual learning. The extension to include social factors
provides a way of examining the effectiveness of CVLEs for conceptual learning.
The paper continues with a discussion of the empirical work undertaken at the two
centres of the Open University and Leeds, where evaluations of two different desktop
virtual reality environments have taken place. These investigations assess the role of
physical presence and social presence for the advancement of conceptual learning with
virtual reality systems.

Investigating Undergraduate Science Field Work

At the Open University we have concentrated on producing a number of desktop virtual

reality environments to assist undergraduate science students in their ®rst year of study.
One of the environments that was produced was known as the North Atlantic Ridge. It
took students to the bottom of the ocean ¯oor while the piece of software known as the
Oak Wood allowed students to participate in a ®eld trip to an ancient woodland site. Both
these programs were developed by the BBC to give the students as much a feeling of
`being there' as possible. It was thought that this condition should enhance the students
motivation to continue working with the conceptual notions embodied in these virtual
reality environments.
PERFECT PRESENCE 281

Software: The North Atlantic Ridge

The North Atlantic Ridge was developed in MTropolis. It allowed the students to explore
the terrain of the North Atlantic Ridge. It did this by taking them in a submarine down to
the Ridge itself at the bottom of the ocean. Students could move around the Ridge in the
submarine to explore the terrain for geological structures and biological life in seven major
locations along the Ridge. This virtual environment introduced the student to an unfamiliar
terrain and hence on the scale for representational ®delity would have a low rating.
Students experienced high immediacy of control with a low to medium value for presence
(as this was a desktop system). The submarine could move in all four major directions of
the compass, North, South, East and West. Speed of travel varied in how quickly one
moved the mouse across the screen and there were two views of the proceedings. One was
the view from the submarine which ®lled most of the screen, while the other comprised of
a plan view of the submarine itself and where it was located presently in the Ridge. The
latter vista was designed to aid the navigation around the Ridge ± a terrain which was
indeed interesting and compelling to understand, but very unfamiliar.
The students were literally steering the submarine and could choose what to investigate
and, when they found allocation, where to stop. This meant they could look more closely at
the geology or the ¯ora and fauna by viewing movies which were accompanied by a short
written text. These video clips were taken by the Alvin Dive Team and illustrate the probes
used to gain samples of geological material and shots of the ¯ora and fauna that thrive on
the different mounds.

Investigating the effect of audio on a sense of presence using the North Atlantic Ridge
software

The preliminary investigations undertaken with the North Atlantic Ridge suggested that
more audio feedback would enhance a sense of presence with this software. As we are still
building virtual reality systems, it became important to understand what parameters
actually enhance a sense of presence and whether in turn how changes to so called
presence parameters effect how students learn with these systems. Therefore, the following
investigations were undertaken to test whether audio feedback enhanced the students
feeling of presence in a virtual learning environment. Two experimental conditions were
arranged. One where there was enhanced audio feedback and the other where there was
limited audio feedback. The study was designed ®rstly to see if the enhanced audio
condition did increase the participants' sense of presence. Secondly, we wanted to see how
the enhanced audio condition affected learning outcomes.

Participants and task The empirical study involved Year 12 students, aged between 16
and 17 y, (n ˆ 20). Prior to the experimental session which took place in the school,
students were asked to complete a brief questionnaire concerning their previous computing
and game playing experience. The students were then paired so that prior experience was
controlled as a variable. The students were grouped under two different conditions, one
282 WHITELOCK ET AL

group had audio feedback and the other had no audio feedback. At the outset, students
were asked to complete a pre-test on their cognitive understanding of the geology of the
ocean ¯oor, then they were given a series of tasks to complete and a post-test was
administered. The tasks involved using the virtual environment of the North Atlantic
Ridge, where students explored the environment via a submersible. Students were
requested to move around the environment to become acquainted with the location and
the movements of the submersible. After a brief session of `getting to know' the
environment, students were presented with a printed list of locations and biological life
which they had to look for and comment on. Their conversations were recorded for later
transcription.

Findings We found that enhanced audio feedback increased the students' sense of
presence from their subjective ratings of presence in the sound and on sound conditions
(see Table 1 below). This enhanced sense of presence contributed to notions of engage-
ment and time passing quickly when using the environment. It also contributed to a better
on task performance.
The enhanced audio condition did not affect the students' perception of control since
this was rated as 5.8 and 5.9 respectively. However what is interesting, is a change in
representational ®delity which was noted in the two conditions. It is apparent that without
sound the visual representation appeared more realistic when in fact the model terrain was
not an exact replica but an academic representation of the ocean ¯oor of the North Atlantic
Ridge. The visual representation appeared more real and exact when it was a silent one, but
the variable of Presence was enhanced by sound. It is questionable as to whether the value
of representational ®delity needs to be very high for conceptual learning. But what is
apparent from this study is that the visual input is dominant when audio feedback is low.
Other ®ndings from our work suggest that a high level of Presence provided a
motivating environment in which students can explore conceptual notions. It is certainly
a primary aim for educational technologists to provide stimulating and motivating
materials for students, especially for those who work alone at home, as is the case with
the undergraduates at the Open University. However, it is still important to check whether
students are learning with these programs. Therefore students' on task performance
measures were recorded.

How well did students understand the concepts introduced by the software? A pre- and
post-test was administered to ascertain what the students had learned. The mean score on a

Table 1. Mean ratings of Presence, Control and Representational Fidelity

by students in the sound and no sound conditions

Sound No Sound

Presence 6.3 4.9

Control 5.8 5.9
Representational Fidelity 6.4 7.5
PERFECT PRESENCE 283

®ve point scale for the pre-test was 3.3 for the enhanced audio condition and 2.9 for the
normal audio condition. These students were not familiar with the ¯ora and fauna of the
North Atlantic Ridge and were approximately all at the same starting point with respect to
prior knowledge of this domain. There was however more of an improvement in the post-
test scores for the normal audio group, improvement of 8 points than for the enhanced
audio condition with an improvement of 4 points. This was a signi®cant difference (F ˆ 4,
p < 0.05; one way ANOVA of cognitive change score versus audio condition). It is a
surprising ®nding when one considers how students worked through the tasks when using
the software. These ®ndings are described in the section below.

Increased presence and its effect on understanding The students reported less dif®culty
in ®nding the different active sites when an audio track was present. In fact more students
stated it was neither easy nor dif®cult to locate various regions when an audio track was
played when they used the program. What is more surprising is that students experienced
more dif®culty understanding the conceptual content in the different areas when an audio
track is present.
One inference could be that `being there' is very motivating but could well take up too
much of the users attention and produce a cognitive overload when it comes to under-
standing conceptual notions. Hence we need to obtain more data about the balance of these
salient variables that facilitate conceptual learning.

Developing conceptual tools that assist understanding yet support an increased sense of
presence

In our development of another virtual reality environment which introduced students to the
ecology of an Oak Wood the students took part in a ®eld trip to an actual Oak Wood. This
location had been ®lmed for over a year to obtain the appropriate footage in order to create
this particular virtual reality system. The program presents the user with a large number of
species to investigate and, from their inter-relationships, concepts such as ecosystems, food
chains and energy transfer levels can be appreciated. The ®rst part of the program
introduces the students to a number of species in the wood and provides them with a ®eld
guide which was used as a resource for further study. The second part asks them to create a
mini food web with respect to the life cycle of the sparrow hawk.
The task was equivalent to that in The North Atlantic Ridge where ®rstly they had to
®nd their way around the environment and identify species; for example, identify
geological features in The North Atlantic ridge, here ®nd the various species in an Oak
Wood and then to carry out a conceptual task. With the Oak Wood the task was to
construct a mini food web, and in so doing appreciate that in the grand order of things, it
takes the leaves from one whole oak tree to successfully pass through the food chain, to
raise a sparrow hawk chick.
With the Oak Wood evaluation (Whitelock 1999) some surprising ®ndings were
uncovered. This program used Quicktime VR to allow the students to search an authentic
Oak Wood and hence there was a high degree of presence recorded by the students.
284 WHITELOCK ET AL

However, the students did not want this virtual reality system to be like the real thing. They
became frustrated when trying to ®nd the specimens in the Oak Wood, such as an adder
and the deer. These two species had been carefully camou¯aged as they would be in real
life, but the program produced an impatience among the subjects that would not have been
experienced in the ®eld.
Users did have a zoom in and zoom out facility in the Oak Wood and this was an
environment with a high representational ®delity; i.e. it required students to search out the
animals in a way which is similar to that experienced in a ®eld trip. However, our subjects
did not want to search in the same way as they would in real life. In fact they wanted to use
`the magic of the computer' to assist their discovery of specimens. In a sense they wanted
some conceptual tools which would not have been available to them in the real life
situation and in fact in the ®nal version of the software we did provide a tool that would
highlight specimens that could not be easily found.

Summary

Increased audio feedback gave students an increased sense of presence. This in turn
assisted them in navigating around the North Atlantic Ridge and also provided a
motivating learning experience. However it surprisingly did not enhance the conceptual
learning for these students. Some clues to this ®nding can be found in the second study
which used the Oak Wood as the virtual reality environment. Here the high degree of
presence was not what the students wanted at all. They requested more conceptual tools
and aids to help them ®nd things in a much quicker way than they would do on a ®eld trip
in real life. Hence we have two interesting sets of ®ndings that suggest going for a high
degree of presence per se in an educational virtual reality system should be looked upon
with caution.
The second half of this paper continues to the discussion of the role of presence but with
respect to collaborative virtual reality systems. Will the picture that emerges here throw
some more light on the advantages of `being there'?

Investigating the features required by a Training Program for Professional

Fire-®ghters

Most virtual environments give a representation with a level of physical ®delity that
matches the requirements for the world's purposes. This desire for representing exactly the
objects of the world, or their accepted abstractions=metaphors, can lead to the neglect of
the role of re¯ection in the training experience.
The standard way of providing team training in ®re includes physical training dealing
directly with a real ®re, as well as `pen and paper' exercises where the world in not
represented with physical, perceptual ®delity in order to concentrate more on the cognitive
issues in command and control situations. Therefore designing a Virtual Environment for
training ®re-®ghters we have to consider both aspects of training.
PERFECT PRESENCE 285

A Virtual Environment for training ®re-®ghters should be a true 3D reproduction of the

high-risk dynamic environment to provide a high level of cognitive load with opportunities
for highly situated decision making. If this is associated with a strong sense of presence it
will lead to knowledge recallable in the real world as a consequence of an engrossing
experience acquired in the `virtual world'. But we believe that the virtual environment as
well as providing a close physical resemblance, immediate feedback, and a strong sense of
presence; should also support various ways of re¯ecting on the relationship between the
elements of the dynamically changing situation and the learner's goal which has been
de®ned as situation awareness (Bass, 1998).
Since we are in a team-training context, we should support a sense of presence through
both individual perception and group awareness, owing to the need for a shared
environment in which team members can re¯ect together and collaborate in resolving
the problems that arise in the ®re ®ghting task and acting out their decisions.
The problem with a training system for ®re-®ghters is to engineer it in a way such that
situation awareness, presence and the cognitive issues of a command and control situation
are all achieved. Our research issue is therefore to identify the features needed to optimise
such a training process.
As part of our work on naturalistic decision making in high risk dynamic situations, we
studied how collaborative decision-making takes place in a shared dynamic desktop VE,
and how a sense of both physical and social presence in¯uences this collaborative process.
Here, we are concerned with issues connected with the use of desktop VR to engage the
learner, and whether collaboration in a CVLE improves the sense of presence or degrades it.
A well-known multi-participant desktop virtual game was used to observe to what extent
the participants feel `present' in an environment with visual and auditory information. The
players have to ®nd their way out of a 3D maze while surviving the attack of other humans
and animals represented in the 3D world.

Participants

There were twelve participants in the experiment, seven female and ®ve males, varying in
age from the mid twenties to the mid thirties. They were all familiar with the use of
multimedia desktop personal computers and some with the game itself. Preliminary
training on the basic features of the game was given to the subject that had never
played before.

Materials

Two multimedia personal computers with 1500 monitors were used with the game installed
on each computer. A video camera was placed in a ®xed position with an overview of the
subjects and the screen on which the game is played to record activity for later
examination. The game represented a world with constraints similar to reality, where the
player has to breathe if swimming, dies if shot by the hostile creatures and so on. Each
286 WHITELOCK ET AL

player has a limited number of `lives'. The only visual perception that each participant has
of himself is the representation of his gun on the screen, while he can see the overall body
of the other players and of the creatures in the environment. A combination of mouse and
arrow keys was needed to move and interact with the game. Different game modalities
were utilised. The one used in the collaborative experiment had two participants ®ghting
together to ®nd the way out. The two players could not harm one another. The pairs of
players for each session of the experiment were selected based on their being used to
working together in real life.

Results

Burton, Brna and Treasure-Jones (1997) suggest that roles peoples adopt within a group
are strictly related to the effectiveness of their collaboration and to the learning gains.
Therefore the roles undertaken in the collaborative exercise were examined. We noticed
that the behaviours in the virtual environment resembled those of the real world and we
have categorised three types of behaviour. Each of these was shown by two different
couples, and are:

 The teacher-pupil role, where one player does most of the talking, guides the other
through the maze and looks after him as he kills all the hostile entities and takes
decisions considering the other player's needs. The other would follow the teacher as
happens in reality and perform what is requested with a better competence and
con®dence than it is likely he would have done when playing on his own. In this
case the teacher also reported that this way of playing was more stressful and required
more concentration than playing alone.
 The equal contribution players, that share all the resources and come to decisions by
reaching a consensus. Those also waited for the partner to join them during movements
around the virtual world, guiding them to the location describing what they could see. They
did not report more concentration than when playing alone, and found the game more
enjoyable. Also team performance was far superior to the single performances demon-
strated, showing both team members bene®ted from the form of the collaboration.
 Dominant behaviour where one of the two players, despite the knowledge of the other
player, ignored his suggestions most of the time and took decisions for both players. In
this case the second player followed the ®rst player, keeping his distance so that he could
enjoy playing the game on his own while monitoring the ®rst player's behaviour. The
dominant player's performances were not improved and the second player reported that
he would have done better playing on his own. This was considered to be a case of non
collaboration.

A post event questionnaire was completed by the participants. The validity and
reliability of it need to be established, but it suggests that 92% of the subjects in the
experiment felt they were in the place they were looking at on the screen always (58%) or
PERFECT PRESENCE 287

sometimes (33%). Also players with a greater experience reported a higher sense of
personal presence.
As for shared sense of presence 33% of the players felt that their partner was always in
the environment presented by the game, while 67% felt that he was sometimes in the
environment. Nobody reported that he never felt the partner in the environment on the
screen. On the other hand, 58% felt that they were both in the environment and that
collaboration was possible even if they did not see their partner on the screen. While 33%
felt that collaboration was enhanced when they could see their partner and sometimes they
both felt themselves to be in the environment.
Only one person (8%) reported that she never felt a sense of personal presence during
either training or the collaborative experiment, but she reported feeling a sense of shared
presence sometimes when her partner was on the screen and always when both together in
the environment.
Similarly one person (8%) reported that despite feeling sometimes that he or his partner
was in the virtual environment, he never felt both together. This was reported in one of the
couples where dominant behaviour was noticed.
This game had been chosen since it seemed to us that its popularity was due to a strong
sense of presence being felt by the players of the game. In fact, even subjects not generally
interested in video games once trained and con®dent, found it very dif®cult to stop playing.
The post event questionnaire after the training revealed that 92% of the subjects felt a
sense of personal presence at least sometime. Also it was noticed that most of the trainees
that felt present on the screen only sometimes, changed their answer to a complete sense of
presence after the collaborative experiment.
It is also interesting to report that in the post event questionnaire the mouse, which
allowed the player to look around, was described in terms of the players' head movements,
while the arrow keys, which allow backwards, forwards and sideways movements were
related to the legs of the player.

Did the Sense of Presence Improve the Training?

The game scenario is strongly analogous to team training in a high risk dynamic
environment. While it is always possible in VEs that there is signi®cantly less perceived
risk than in real life, it did appear that the game playing context selected was effective in
engaging the participants. In addition, where collaboration was effective a strong sense of
shared presence was reported and performances were improved. Also the behaviours
observed during the experiment reproduce those of real life. This result leads us to believe
that the construction of a shared dynamic VE with similar settings of the analysed game
could be used for training collaborative naturalistic decision making skills.
While this is a preliminary result that could bene®t from a more thorough analysis of the
data, it does support the notion that social presence can be effective in maintaining a strong
sense of physical presence. This is good news for desktop VR's potential to engage the
learner. However, the implication is that, as with other forms of ICT, we should have regard
288 WHITELOCK ET AL

for what we already know about small group learning, and take some care in forming
teams of students to work together in a CVLE.
The implications for concept learning are indirect. From an educational perspective on
CVLEs, the notion of affordance is a key issue. We can use them to bringing concepts
`into play' within a rendering of a physical world in a CVLE where we want `affordances
that trigger a need to satisfy educational goals which have a theoretical content, a desire
for understanding certain abstract concepts, an interest in theoretically signi®cant actions'
(Brna, 1999). Whitelock and Brna have coined the term `handles on the theory world' to
describe those affordances that provide an individual with the `entry point' into a
theoretical world. The work with teams of ®re-®ghters does have an interesting but
implicit conceptual structure (the nature of ®re, the way it spreads, the structure of
materials, concepts of heat and temperature, conduction, convection etc.). Later work
will need to add these `handles' to provide a conceptual learning experience for novice
®re-®ghters.

Conclusions

A feeling of enhanced presence was induced by an augmented sound system. This was
experienced by the students as a greater sense of `being there' at the bottom of the North
Atlantic Ridge, exploring the terrain in their own submersible. Enhanced sound made
navigation around the environment easier but, more importantly, the students thought the
time passed more quickly using this version of the software. The subjects worked in pairs
and appeared to argue less in this environment. This was because they could complete the
tasks more easily and a joint strategy did not have to be constructed in order to solve a
problem. Altogether the enhanced sound version appeared to be a more motivating venue
for exploration.
On the other hand the more super®cial visual clues of the terrain did not give the
students enough information to complete the tasks. They had to explore the individual sites
in more detail to understand and construct in `mental maps' of what they were doing. This
could have contributed to their increased cognitive change scores.
These ®ndings suggest that sound does contribute to a greater sense of presence for the
user but it interferes with the cognitive processing tasks. Hence `being there' is very
motivating but could well take up too much of the user's attention and produce a cognitive
overload, when it comes to understanding conceptual notions. In fact this seems to be the
case with the Oak Wood software, where students asked the designers to decrease the
cognitive overload of `being there' and to provide them with tools to help them locate the
hidden specimens in the wood.
The studies undertaken in a collaborative working environment suggest that another
factor that enhances a sense of presence is that of social awareness. This social presence
enhances a feeling of team work and co-ordination is, of course, an important considera-
tion in the design of real life training systems. However more work in designing conceptual
tools to support learning in these environments has still to be done.
PERFECT PRESENCE 289

These studies provide some benchmark data about all the above issues which require
further investigation if we are to design effective virtual environments for conceptual
learning. Hence we are still left with probing more deeply the important question of
whether `being there' is enough for conceptual learning to be facilitated in virtual reality
environments.

References

Bass, E.J., (1998). Towards an Intelligent Tutoring System for Situation Awareness Training in Complex,
Dynamic Environments. Lecture Notes in Computer Science 1452, Intelligent Tutoring Systems, Proceedings
of the 4th International Conference, ITS 198, San Antonio, Texas, USA, August 1998, pp 26±35.
Brna, P. (1999). Collaborative virtual learning environments for concept learning. International Journal of
Continuing Engineering Education and Life-Long Learning, 9, Nos. 2=3=4, 315±327.
Burton, M., Brna, P. and Treasure-Jones, T. (1997). Splitting the Collaborative Atom: How to Support Learning
About Collaboration. In Arti®cial Intelligence in Education: Knowledge and Media in Learning Systems, du
Boulay, B. and Mizoguchi, R. (eds) Amsterdam: IOS. pp 135±142.
Jelfs, A. and Whitelock, D. (1999). Activity Theory and Virtual Learning Environments: A lever to understanding
the role of Presence. In Proceedings of the Sixth UK Virtual Reality Special Interest Group Conference. Edited
by T. Fernando. Salford University, September. pp. 1±10.
Kalawsky, R.S. (1991a). Pilot integration and the implications on the design of advanced cockpits. In Conference
on the Man-Machine Interface in Tactical Aircraft Design and Combat Automation, Agard, Stuttgart.
Kalawsky, R.S. (1991b). From visually coupled systems to virtual reality: an aerospace perspective. In
Proceedings Computer Graphics '91, Blenheim On-line.
Kalawsky, R. (1993). The Science of Virtual Reality and Virtual Environments. Reading, MA: Addison-Wesley.
Laurel, B. (1993). Computers as Theatre. Addison Wesley, New York.
Lombard, M. and Ditton, T. (1997). At the Heart of it all: The Concept of Telepresence. Journal of Computer
Mediated Communication, 3(2).
Rheingold, H. (1991). Virtual Reality. Simon & Schuster, London.
Romano, D.M. and Brna, P. (2000). ACTIVE World: Manipulating Time and Point of View to Promote a Sense of
Presence in a Collaborative Virtual Environment for Training in Emergency Situations. 3rd International
Workshop on Presence, Delft University of Technology, Delft.
Steuer, J. (1992). De®ning Virtual Reality: Dimensions Determining Telepresence. Journal of Communication,
42(4), 73±93.
Whitelock, D. (1999). Investigating the role of task structure and interface support in two virtual learning
environments. International Journal of Continuing Engineering Education and Life-Long Learning, 9, Nos.
2=3=4, 291±301.
Whitelock, D., Brna, P. and Holland, S. (1996). What is the Value of Virtual Reality for Conceptual Learning?
Towards a Theoretical Framework. In Brna, P., Paiva, A. and Self, J.A. (eds.) Proceedings of the European
Conference on Arti®cial Intelligence in Education, pp 136±141, Edicoes Colibri, Lisbon.