Effect of Environment Size on Spatial Perception in Virtual Reality

Morten Trier Bach, Poul Anker Jensen, Andrei Iuliu Lucaci, Damian Grzegorz Pupczyk

and Claus Brøndgaard Madsen

Computer Graphics Group, Department of Architecture and Media Technology, Aalborg University, Aalborg, Denmark

Keywords:

Virtual Reality, Virtual Environments, Spatial Perception, Blind Walking.

Abstract:

A trend of distance underestimations in Virtual Reality (VR) is well documented, but the reason still remains

unclear. Therefore, this paper investigates the effect of differently sized Virtual Environments (VEs) on ego-

centric distance perception in VR as a potential inﬂuence. Verbal assessment, blind walking, and our own

proposed method: walk and assess, were compared in an experiment, and blind walking was found to be the

most accurate. A virtual replica of a real-life location was created as a transitional environment, while small

(15m

), medium (35m

) and large (95m

) rooms were created to investigate the effect of VE size on spatial

perception in VR. To establish the differences in estimations between the real world and VR when using blind

walking, a study was conducted with the virtual replica and its real life counterpart at distances between 1 and

10 meters. Following this, the three distinct room sizes were used in an experiment to investigate the effect

of the size of rooms on spatial perception in VR. The ﬁndings showed consistent underestimates of distances,

and a trend for underestimation to grow as the distance grows was observed. Similarly, underestimates grew

with the size of the environment.

1 INTRODUCTION

Virtual Reality (VR) as a technology has been shown

to be inﬂuential within the world of architecture,

where it is used to represent and simulate architec-

tural spaces Loyola (2018). Unfortunately, quite a

few researchers have found that distances in Virtual

Environments (VEs) are generally underestimated by

10 to 26 percent (Loyola (2018), Witmer and Sad-

owski Jr. (1998), Ahmed et al. (2010), Renner et al.

(2013)), which threatens the validity of the platform

as a medium for representation of spaces, and makes

it a subject worth exploring.

In an experiment done by Lucaci et al. (2022),

ﬁndings indicated general overestimation of dis-

tances. They hypothesized that it might have been

due to the relatively small size of the environment

(12.6m

) compared to previous research which used

larger environments ranging from 13m

to 233m

(Simpson et al. (2018), Moscoso et al. (2021), Zhao

et al. (2019)).

Lucaci et al. (2022) used Verbal Assessment to

measure perceived distances, however, Thompson

et al. (2004) have criticised the method for being sub-

jected to bias and noise.

https://orcid.org/0000-0003-0762-3713

With this in mind, the aim of this study was to

identify a reliable method of quantifying perceived

distances, and use that to investigate the effect of dif-

ferent sizes of VEs on spatial perception in VR.

2 SPATIAL PERCEPTION IN

VIRTUAL REALITY

Spatial perception is a combination of physiological

and cognitive processes which utilize a variety of per-

ceived depth cues to turn a two-dimensional image

into a three-dimensional representation of the world

(Renner et al., 2013).

Perceived distances can be either ego-centric or

exo-centric. Ego-centric distances refer to distances

between an observer and a point, while exo-centric

distance is the distance between two external objects.

(Maruhn et al. (2019), Renner et al. (2013))

Underestimation of ego-centric distances in Vir-

tual Reality (VR) was found to be 19% by Loyola

(2018) and 15% by Witmer and Sadowski Jr. (1998),

while real-world distances were found to be underes-

timated by 8% by Witmer and Sadowski Jr. (1998),

and 4% by Messing and Durgin (2005). These un-

derestimations are particularly evident with egocen-

Bach, M., Jensen, P., Lucaci, A., Pupczyk, D. and Madsen, C.

Effect of Environment Size on Spatial Perception in Virtual Reality.

DOI: 10.5220/0011376400003323

In Proceedings of the 6th International Conference on Computer-Human Interaction Research and Applications (CHIRA 2022), pages 97-104

ISBN: 978-989-758-609-5; ISSN: 2184-3244

Figure 1: The three differently sized rooms used in the Room Size Experiment in Section 6.

tric distances above 1 meter. On the other hand,

some studies have indicated that exo-centric distances

in VR may be overestimated in VR. (Maruhn et al.,

2019)

However, the focus of this paper is only on ego-

centric distances to emphasize the room size effect. It

is also applicable to more methods of distance estima-

tion than exo-centric distances (see Section 4).

2.1 Depth Cues

To properly understand spatial perception, a basic un-

derstanding of the depth cues which allow us to per-

ceive depth is needed. Such an understanding can aid

in identifying what might lead to errors in distance

estimations.

The depth cues used to inform the three-

dimensional understanding of the world can be di-

vided into two categories; pictorial and non-pictorial

cues. Pictorial cues are cues that can be obtained

from a motionless image, while non-pictorial cues are

largely from motion in the scene or the human vi-

sual system (Renner et al., 2013). The pictorial cues

include occlusion, relative and familiar size, hori-

zon ratio, shading and texture gradients (Profﬁtt and

Caudek, 2012).

Non-pictorial cues include binocular disparity,

motion parallax and ocular motor (accomodation and

vergence) (Profﬁtt and Caudek, 2012). Maruhn et al.

(2019) suggest that visual cues might differ between

physical environments and VR in a multitude of ways,

such as by introducing the vergence-accomodation

conﬂict, which is a result of presenting different im-

ages to each eye using stereoscopic displays.

This phenomenon can be a source of visual dis-

comfort and dizziness.

When discussing distance perception of the world,

one should also consider the circular segmentation of

egocentric distances around a moving observer as pre-

sented by Cutting and Vishton (1995). The segments

and their associated depth cues are shown in Figure

2. The ﬁrst of these segments is referred to as Per-

sonal Space, and is the area closest to an individual

Figure 2: The interplay between segmentation of egocentric

distances and most important visual cues.

(within arm’s reach) delimited by Cutting and Vish-

ton (1995) to 2 meters. The second segment is called

Action Space which extends from the bounds of the

personal space out to about 30 meters - one can move

relatively quickly, speak, teach and toss objects with

relative ease in this segment. The ﬁnal segment is

called the Vista Space which extends beyond 30 me-

ters, at these distances the beneﬁts of binocular vi-

sion and movement of the observer signiﬁcantly di-

minishes. (Cutting and Vishton, 1995)

3 ROOM SIZES IN VIRTUAL

REALITY

When acting in an environment, the size and layout of

it profoundly impacts ones perceived personal, action,

and vista spaces, as distances to objects and walls

might limit the area one can interact in and the avail-

able depth cues. To our knowledge, no prior research

has been done with a particular focus on the effect of

room sizes on spatial perception. Nevertheless, the

sizes of Virtual Environments (VEs) used in Virtual

Reality (VR) research can be incorporated to inform

the design of our VEs.

Sizes of rooms used in such studies vary from just

13.46m

in a study by Zhao et al. (2019) all the way to

233.26m

(21.44m x 10.88m) in a study by Moscoso

et al. (2021).

CHIRA 2022 - 6th International Conference on Computer-Human Interaction Research and Applications

While Moscoso et al. (2021) only used two differ-

ent rooms in their studies, Simpson et al. (2018) incor-

porated 7 square rooms with the sizes 16m

, 18.75m

21.34m

, 25m

, 28.40m

, 32.14m

, 36m

, and uni-

form heights of 2.74m to study the impact of visual

design elements on perception of spaces.

Zhao et al. (2019) used 4 square rooms of sizes

16m

, 21.80m

, 28.40m

, and 36m

to the inﬂuence

of what is seen through windows of a room on per-

ceived spaciousness of rooms. Zhao et al. (2019) also

presented a room of 13.46m

size as a reference to

spaciousness of 1, and a room of size 40.06m

as a

reference for spaciousness 10.

3.1 Categorization of Room Sizes

Based on the rooms used by Simpson et al. (2018),

and Zhao et al. (2019), a survey was designed and

distributed to 45 people. The survey consisted of

10 rectangular ﬂoor plans (9m

, 19m

, 29m

, 39m

49m

, 59m

, 63m

, 75m

, 86m

, and 97m

), which

were presented to participants in random order. Par-

ticipants were asked to rate whether they found each

room to be small, medium or large.

The weighted average of the results were com-

puted for each group; the boundary between small and

medium was ∼25m

and between medium and large

was ∼52m

From these results the intervals were deﬁned as: 0-

25 m

for small, 25-50 m

for medium, and above 50

for large. This categorization was used to inform

the sizes of rooms used in the Room Size Experiment

presented in Section 6.

4 METHODS FOR QUANTIFYING

PERCEIVED DISTANCES

To investigate and compare perceived distances in a

set of virtual environments, one must be able to quan-

tify them. Two of the most prominent methods for

quantifying perceived distances are:

• Verbal Assessment: (Loyola (2018), Peer and

Ponto (2017), Armbr

uster et al. (2008), Gagnon

et al. (2020), von Castell et al. (2018), Kelly et al.

(2017), Klein et al. (2009))

• Blind Walking: (Interrante et al. (2006), Kelly

et al. (2018), Li et al. (2015), Kelly et al. (2017),

Ahmed et al. (2010))

It was decided to evaluate Verbal Assessment, Blind

Walking, and Walk and Assess (a method where par-

ticipants walk a distance of their choice from a start-

ing point, and then estimate it verbally, devised by

us) with participants to identify the method that per-

formed most similarly between real life and VR with

the highest accuracy. The experiment was conducted

on 30 participants with perfect 20/20 or corrected vi-

sion, who would estimate three distinct distances for

each chosen method. Verbal Assessment resulted in

22% average error in VR and 13% in the Real World

(RW), Walk and Assess showed 25% average error in

VR and 18% in RW, and Blind Walking showed 12%

average error in both VR and RW. Due to the fact that

Blind Walking performed most similarly between VR

and RW, and resulted in the least average error in both

VR and RW, we chose this method to be used in the

succeeding experiment.

5 BLIND WALKING BASELINE

EXPERIMENT

To establish what difference might exist in estimates

between the Real World (RW) and Virtual Reality

(VR) when using Blind Walking, an experiment to

evaluate the method was designed. The purpose was

the establishment of a baseline difference between

VR and RW, as well as it being a pilot test of the

method for the ﬁnal investigation of room sizes effect

on spatial perception in VR.

5.1 Participants

40 participants were recruited with normal (20/20) or

corrected vision. All participants were university stu-

dents recruited by convenience, and had an average

Interpupilary Distance (IPD) of 65mm.

5.2 Apparatus

A Quest 2 Head Mounted Display (HMD) was used,

untethered to allow free movement, connected to a

laptop with Oculus Airlink. The laptop had a Nvidia

Rtx 3060 graphics card and ran Windows 11.

In order to minimise the blurriness that partici-

pants might experience from an uncalibrated HMD,

each participant’s Interpupillary Distance (IPD) set-

tings were measured and adjusted to one of the three

physical settings on the Quest 2 (58mm, 63mm, and

68mm).

The Field of View (FOV) of the Quest 2 was also

measured, and found to be 94°(IPD 58mm), 93°(IPD

63mm) and 95°(IPD 68mm) vertically, with the hori-

zontal FOVs being 93°(IPD 58mm), 94°(IPD 63mm)

and 86°(IPD 68mm).

Effect of Environment Size on Spatial Perception in Virtual Reality

5.3 Procedure

The participants would enter a large empty room

(130m

) assigned as a testing area (as shown on Fig-

ure 3) and complete a consent form.

A facilitator (facilitator 2 in Figure 3) shortly

briefed each participant of the procedure of the ex-

periment, and would keep the participants informed

of what would happen next during the experiment. A

Latin square was utilized to counterbalance the order

of the conditions and distances between participants.

In the real world condition participants were ﬁrst

shown a target point, after which they were asked to

blindly walk to the target point. The participants were

asked to close their eyes and wear a sleeping mask

to prevent them from seeing the target point while

walking. After assessing a distance by walking to the

point, the distance was measured by a facilitator with

a laser range ﬁnder, and the participants were asked to

return to their starting position which was marked on

the ﬂoor with a cross. The process was repeated until

all distances were assessed.

In the VR condition, participants entered a vir-

tual replica of the empty room and went through the

same process as described above. Instead of a sleep-

ing mask, the screen in the HMD was faded to black

after participants were informed to close their eyes.

The participants’ walk was assessed by the software

instead of a laser range ﬁnder. The changing of tar-

gets, fading to black, and noting of measurements was

done by facilitator 1 as seen in Figure 3.

To ensure participants’ safety during their Blind

Walking, facilitator 3 (in Figure 3) would prevent

them from colliding with walls should they have

walked too far.

Figure 3: Visualization of the test setup for the Blind Walk-

ing baseline experiment.

5.4 Results

To compare the distance assessments across condi-

tions and distances, the amounts of error in the as-

sessments were expressed in percentage. The abso-

lute values of results were analyzed in the context of

accuracy, deﬁned as ”how much the assessments devi-

ate from the actual distances” (Loyola, 2018), while

the standard deviation of relative values were used to

represent the precision of the assessments, deﬁned as

how similarly participants within a certain group as-

sessed the distances (Loyola, 2018).

Figure 4: Participants’ precision in VR at given distances

reveals a tendency to underestimate as distances gets larger.

Figure 5: Participants’ precision in RW at given distances

shows a similar trend of underestimation as in the VR con-

dition in Figure 4.

The mean relative error for the VR condition was

-9.21%, and the mean relative error for the RW con-

dition was -6.84% - this shows that on average, par-

ticipants underestimated distances in both conditions,

with underestimations being greater by 2.37% in VR

than RW. The standard deviation of the relative errors

in VR was 7.42%, while the standard deviation for the

real world condition was 8.9%. As the VR condition

had a slightly lower standard deviation (1.51% dif-

ference), it shows that the VR condition had slightly

higher precision than in the RW condition. This was

shown to not be a statistically signiﬁcant difference

(p-value = 0.22 > 0.05).

CHIRA 2022 - 6th International Conference on Computer-Human Interaction Research and Applications

100

Plots showing the relative errors at each distance

for the VR condition can be seen on Figure 4, and the

relative errors at each distance for the RW condition

can be seen on Figure 5. The plots show a clear trend

towards greater degrees of underestimation as the dis-

tance increases, which is both observed in VR and

RW. Also noteworthy is that the underestimation gets

slightly smaller in the ﬁnal estimate (10.35m) poten-

tially due to it being close to the physical end of the

room used for testing.

The mean absolute error in the VR environment

was found to be 12.26%, and 10.76% in the RW envi-

ronment, suggesting that the estimates in the RW were

more accurate than in VR, and the difference between

the two conditions were found to be statistically sig-

niﬁcant (p-value = 0.04 < 0.05).

This study of Blind Walking as a method, estab-

lishes that the degrees of underestimation between

VR and RW are fairly similar (∼ 2% difference), with

less underestimation in RW than VR. The RW under-

estimation were in line with the prior research pre-

sented in Section 2, while the degree of underestima-

tion in VR was also lower than what was found by

prior research. As the method performed similarly to

prior work in the ﬁeld, with little difference between

VR and RW, it was chosen to be utilized in the ﬁnal

experiment regarding the effect of room sizes on spa-

tial perception in VR (described in Section 6).

6 ROOM SIZE EXPERIMENT

As a baseline for the relationship between Virtual Re-

ality (VR) and the Real World (RW) was already es-

tablished in Section 5, the rest of the experiments

were conducted purely in VR.

To investigate the inﬂuence of room size on spa-

tial perception three different rooms were designed: a

small room (∼15m

), a medium room (∼35m

) and a

large room (∼95m

). The sizes of the rooms were

based on the research described in Section 3.1 and

were all chosen to have the same 3:5 aspect ratio.

All the rooms give access to the personal and action

spaces, while the vista space is never reached, which

afforded the depth cues presented in Figure 2. The

rooms can be seen side by side in Figure 1.

6.1 Participants

36 participants were recruited with normal (20/20) or

corrected vision. Participants consisted of university

students recruited by convenience. The average IPD

of the participants was measured to be ∼65mm.

6.2 Apparatus

The apparatus used for the Room Size Experiment was

the same as for the Blind Walking Baseline Experi-

ment, as presented in Section 5.2.

6.3 Procedure

The participants would enter the same large empty

room as in the Blind Walking Baseline Experiment

with an area of 130m

, where they were informed of

the procedure of the experiment and ﬁlled out a con-

sent form.

Participants then entered a transitional environ-

ment in Virtual Reality (VR), which resembled the

large empty room they were present in. After a brief

delay, they would then be transported to the ﬁrst room

of the experiment (either the small, medium or large

one).

In each of the rooms of the experiment, partici-

pants would be presented with blue crosses on the

ground which served as targets, and once they said

that they felt ready to walk there without vision, they

were asked to close their eyes and the screen would

fade to black. This procedure would be repeated for

4 targets in the small room, 6 targets in the medium

room, and 8 targets in the large room as indicated on

Figure 6.

All targets and rooms were presented in random

orders (based on a Latin square) for each participant,

and since the targets 0.9m, 1.7m, 2.6m and 3.2m were

presented to participants in both the small, medium

and large rooms they would serve as the basis of later

comparisons. We refer to these distances as the uni-

versal distances.

Figure 6: Visualization of the transitional environment

setup with the three overlapping rooms and distances.

Effect of Environment Size on Spatial Perception in Virtual Reality

101

6.4 Results

The mean relative error for the universal distances

was found to be -4% in the small room, -5% in the

medium room, and -6% in the large room. This shows

a consistent tendency towards underestimation of dis-

tances, regardless of the room size, which seems to

grow slightly larger as the room size does. The rel-

ative error for universal distances was found to be

normally distributed, and no statistically signiﬁcant

difference was found between the small and medium

rooms (p = 0.49 > 0.05), however, a statistically sig-

niﬁcant difference was found between the small and

large room (p = 0.03 < 0.05). A statistically signiﬁ-

cant difference was also found between the error es-

timates for the medium (8%) and large room (9%) (p

= 0.04 < 0.05) when comparing all overlapping dis-

tances (universal distances, 4.8m and 5.8m).

The standard deviation for the universal distances

was found to be 12% for the small room, 13% for

the medium room and 11% for the large room. This

indicates that the precision is not linearly inﬂuenced

by the room size.

The relative error estimates for each distance in

each room size are presented in Table 1. The ta-

ble shows a general trend towards greater degrees of

underestimation as the distance increases. It is also

worth noting that the ﬁrst distance, 0.9 meters, was

consistently overestimated regardless of room size.

A statistically signiﬁcant difference was found be-

tween all distance estimates of consecutive distances

(p < 0.05), except for between 2.6m and 3.2m in

the small room, 1.7m and 2.6m in the medium room,

and 3.2m and 4.8m in the large room. In most cases

the statistical signiﬁcance shows growing underesti-

mation, with the exception of the cases 3.2m to 4.8m

in the medium room, and 5.8m and 7.1m in the large

room.

Table 1: Relative error of each distance and room size in

percentage. Universal distances are highlighted in green.

Light blue shows overlapping distances between medium

and large, and light purple are only for the large room.

Average Error in Estimates

Distance

(meters)

Small

Room

Medium

Room

Large

Room

0.9 +4% +3% +3%

1.7 -4% -4% -6%

2.6 -7% -7% -11%

3.2 -11% -12% -11%

4.8 -10% -11%

5.8 -16% -20%

7.1 -14%

10.8 -19%

The mean absolute error, representing the accu-

racy, was found to be 10% for the small room, 11%

for the medium room, and 11% for the large room.

The small room showed the lowest absolute error, but

no statistically signiﬁcant difference was found be-

tween any of the room sizes (small-large [p = 0.21

> 0.05], small-medium [p = 0.58 > 0.05], medium-

large [p = 0.31 > 0.05]). The absolute error data was

found to be non-normally distributed.

7 DISCUSSION

The mean relative errors for the universal distances

were found to be -4% in the small room, -5% in the

medium room, and -6% in the large room. The large

room was found to have statistically signiﬁcantly dif-

ferent estimates from the small and medium rooms.

However, this was not the case between the small and

medium rooms. This observation indicates that room

size affects participants’ ability to perceive distances,

with a tendency of larger rooms causing greater un-

derestimations.

The small and medium rooms showed no statisti-

cally signiﬁcant differences in participants’ error de-

gree, which suggests that the difference in size be-

tween the small and medium room has to be bigger

for the effect to have a statistically signiﬁcant inﬂu-

ence on the results.

The results of the experiments also show a ten-

dency of longer distances being underestimated more

than shorter distances. This is shown by most con-

secutive distance estimates being statistically signiﬁ-

cantly different from each other, while also maintain-

ing a trend of increasing underestimation.

In the Blind Walking Baseline Experiment as a

method for quantifying perceived distances, the av-

erage error was roughly -9% for Virtual Reality (VR)

and roughly -7% for Real World (RW). These degrees

of underestimation are largely consistent with previ-

ous studies investigating spatial perception in VR, us-

ing Blind Walking as a method. Ahmed et al. (2010)

found underestimations to be 10% and Witmer and

Sadowski Jr. (1998) reported 15% underestimation in

VR. The 7% underestimation found for the RW con-

dition is also in line with previous research, especially

insofar as being lower than the VR underestimations.

Loyola (2018) found 10% underestimation, Witmer

and Sadowski Jr. (1998) found 8% underestimation in

RW.

Based on the results from the Blind Walking Base-

line Experiment, it can also be theorized that the pres-

ence of a physical wall close to the furthest distance

affected the estimations. As can be seen in Figures 4

CHIRA 2022 - 6th International Conference on Computer-Human Interaction Research and Applications

102

and 5, in both conditions the accuracy decreased the

longer the distance - until 10.35, which was the dis-

tance closest to the wall. A reason for this might have

been some participants putting out their hands in front

of them to try to feel the wall, which was a ﬂaw of the

experiment.

Similarly, in the Room Size Experiment we ob-

served that estimation points close to the side walls

had less underestimation, with 4.8m being the closest

to a wall in the medium room and 7.1m in the large

room as seen in Figure 6 than the furthest point, and

the point before them. This was not seen in the small

room, likely because the points were similarly spaced

from the walls.

A noteworthy deviation from prior research is the

advancement of VR technology in the recent years

allowing the display of images in a higher resolu-

tion and at higher refresh rates. Another difference

is that most previous studies used tethered HMDs for

their experiments, while this study used an untethered

HMD. This might be a reason for the lower degree

of underestimations found in this study, since being

untethered affords more freedom of movement.

7.1 Validity and Reliability Concerns

The implications of this study should be evaluated

with caution due to some threats to the validity and re-

liability of the results. One of them being the fact that

the experiments were carried out on a convenience

sample of university students, opposed to a random

sample. This could affect how adequately they rep-

resent the general population. Nevertheless, as men-

tioned in Section 6.4, the relative error data was nor-

mally distributed. Additionally, since most of the par-

ticipants who took part in the experiments were stu-

dents pursuing technology and design related degrees,

it is also likely that some of them were experienced

with VR. This could affect the results of the exper-

iments, as it could translate into better performance

due to being more used to acting in a virtual space.

Another potential threat affecting the results could

be the sound cues caused by the movement of the fa-

cilitator, who ensured that participant’s would not col-

lide with walls, as it might have affected their spatial

awareness.

Additionally, the distances used in all experiments

were consistent across conditions, with the same

groups of participants testing in all conditions per ex-

periment. This could entail some carryover effects

that could affect the results, such as practice and vi-

sual fatigue. To account for these effects, the exper-

iments were designed with the use of Latin squares,

randomizing the order of distances. To further pre-

vent visual fatigue, the experiments were designed to

take up a short amount of time (on average 10 minutes

per participant).

8 CONCLUSION

The aim of this study was two-fold: to identify a re-

liable method of quantifying perceived distances, as

well as to use the identiﬁed method to investigate the

effect of different sizes of Virtual Environments (VEs)

on spatial perception in Virtual Reality (VR).

To investigate the reliability of the methods, they

were compared between VR and the Real World

(RW). Blind Walking was identiﬁed as a promising

method, as it resulted in the lowest degree of under-

estimation in both VR (12%) and RW (12%) and it

also performed the most similarly between RW and

VR compared to verbal assessment and walk and as-

sess. The method was further investigated in a large

(130m

) empty room to establish a baseline for the

difference between error estimation in VR and RW,

where underestimation of distances were found to be

∼9% in VR and ∼7% in RW.

The effect of different VE sizes on spatial percep-

tion was then investigated using VR and three differ-

ently sized rooms: a small room (15m

), a medium

room (35m

) and a large room (95m

). Underesti-

mations were found to be 4% in the small room, 5%

in the medium room, and 6% in the large room for

the universal distances (distances between 0.9m and

3.2m) in the Room Size Experiment. This indicates

that distances become more underestimated the larger

the area of the VE. It is worth noting that no statis-

tically signiﬁcant difference in estimates was found

between the small and medium rooms, however, the

large room had a statistically signiﬁcantly greater de-

gree of underestimation than the small and medium

rooms.

In general, the underestimations from the Blind

Walking Baseline Experiment and Room Size Experi-

ment are in line with previous literature as presented

in Section 2. Additionally a tendency for underesti-

mation to grow with the distance was found.

More importantly, the Room Size Experiment indi-

cated that room size could inﬂuence spatial perception

in VR, however, more research is needed to validate

this and ﬁnd the exact threshold where room size be-

gins to have a clear effect.

In conclusion, this study contributes to a better un-

derstanding of spatial perception in VR, how it is af-

fected by the size of the environment, as well as the

performance of Blind Walking as a method of quanti-

fying perceived distances.

Effect of Environment Size on Spatial Perception in Virtual Reality

103

REFERENCES

Ahmed, F., Cohen, J. D., Binder, K. S., and Fennema, C. L.

(2010). Inﬂuence of tactile feedback and presence

on egocentric distance perception in virtual environ-

ments. In 2010 IEEE Virtual Reality Conference (VR),

pages 195–202. IEEE.

Armbr

uster, C., Wolter, M., Kuhlen, T., Spijkers, W., and

Fimm, B. (2008). Depth perception in virtual reality:

Distance estimations in peri- and extrapersonal space.

CyberPsychology & Behavior, 11(1):9–15. PMID:

18275307.

Cutting, J. E. and Vishton, P. M. (1995). Chapter 3 - perceiv-

ing layout and knowing distances: The integration,

relative potency, and contextual use of different infor-

mation about depth*. In Epstein, W. and Rogers, S.,

editors, Perception of Space and Motion, Handbook of

Perception and Cognition, pages 69–117. Academic

Press, San Diego.

Gagnon, H. C., Buck, L., Smith, T. N., Narasimham,

G., Stefanucci, J., Creem-Regehr, S. H., and Boden-

heimer, B. (2020). Far distance estimation in mixed

reality. In ACM Symposium on Applied Perception

2020, SAP ’20, New York, NY, USA. Association for

Computing Machinery.

Interrante, V., Ries, B., and Anderson, L. (2006). Distance

perception in immersive virtual environments, revis-

ited. In Proceedings of the IEEE Conference on Vir-

tual Reality, VR ’06, page 3–10, USA. IEEE Com-

puter Society.

Kelly, J. W., Cherep, L. A., Klesel, B., Siegel, Z. D., and

George, S. (2018). Comparison of two methods for

improving distance perception in virtual reality. ACM

Trans. Appl. Percept., 15(2).

Kelly, J. W., Cherep, L. A., and Siegel, Z. D. (2017). Per-

ceived space in the htc vive. ACM Trans. Appl. Per-

cept., 15(1).

Klein, E., Swan, J. E., Schmidt, G. S., Livingston, M. A.,

and Staadt, O. G. (2009). Measurement protocols for

medium-ﬁeld distance perception in large-screen im-

mersive displays. In 2009 IEEE Virtual Reality Con-

ference, pages 107–113.

Li, B., Zhang, R., Nordman, A., and Kuhl, S. A. (2015).

The effects of miniﬁcation and display ﬁeld of view

on distance judgments in real and hmd-based environ-

ments. In Proceedings of the ACM SIGGRAPH Sym-

posium on Applied Perception, SAP ’15, page 55–58,

New York, NY, USA. Association for Computing Ma-

chinery.

Loyola, M. (2018). The inﬂuence of the availability of vi-

sual cues on the accurate perception of spatial dimen-

sions in architectural virtual environments. Virtual Re-

ality, 22:235–243.

Lucaci, A., Bach, M., Jensen, P., and Madsen, C. (2022).

Inﬂuence of texture ﬁdelity on spatial perception in

virtual reality. Proceedings of the 17th Interna-

tional Joint Conference on Computer Vision, Imag-

ing and Computer Graphics Theory and Applications

- GRAPP, 17:244–251.

Maruhn, P., Schneider, S., and Bengler, K. (2019). Measur-

ing egocentric distance perception in virtual reality:

Inﬂuence of methodologies, locomotion and transla-

tion gains. PLOS ONE, 14(10):1–24.

Messing, R. and Durgin, F. H. (2005). Distance percep-

tion and the visual horizon in head-mounted displays.

ACM Trans. Appl. Percept., 2(3):234–250.

Moscoso, C., Chamilothori, K., Wienold, J., Andersen, M.,

and Matusiak, B. (2021). Window size effects on

subjective impressions of daylit spaces: Indoor stud-

ies at high latitudes using virtual reality. LEUKOS,

17(3):242–264.

Peer, A. and Ponto, K. (2017). Evaluating perceived dis-

tance measures in room-scale spaces using consumer-

grade head mounted displays. In 2017 IEEE Sympo-

sium on 3D User Interfaces (3DUI), pages 83–86.

Profﬁtt, D. R. and Caudek, C. (2012). Depth perception and

the perception of events. In Handbook of Psychology,

Experimental Psychology, pages 212–235. John Wiley

& Sons, Incorporated.

Renner, R. S., Velichkovsky, B. M., and Helmert, J. R.

(2013). The perception of egocentric distances in vir-

tual environments - a review. ACM Comput. Surv.,

46(2).

Simpson, G., Sinnis-Bourozikas, A., Zhao, M., Aseeri, S.,

and Interrante, V. (2018). A virtual reality investiga-

tion of the impact of wallpaper pattern scale on qual-

itative spaciousness judgments and action-based mea-

sures of room size perception. In Bourdot, P., Cobb,

S., Interrante, V., kato, H., and Stricker, D., editors,

Virtual Reality and Augmented Reality, pages 161–

176, Cham. Springer International Publishing.

Thompson, W. B., Willemsen, P., Gooch, A. A., Creem-

Regehr, S. H., Loomis, J. M., and Beall, A. C. (2004).

Does the quality of the computer graphics matter

when judging distances in visually immersive envi-

ronments.

von Castell, C., Hecht, H., and Oberfeld, D. (2018). Bright

paint makes interior-space surfaces appear farther

away. PLOS ONE, 13(9):1–15.

Witmer, B. G. and Sadowski Jr., B. G. J. (1998). Nonvi-

sually guided locomotion to a previously viewed tar-

get in real and virtual environments. Human Factors,

40(3):478–488.

Zhao, M., Sinnis-Bourozikas, A., and Interrante, V. (2019).

Exploring the use of immersive virtual reality to as-

sess the impact of outdoor views on the perceived size

and spaciousness of architectural interiors. In Bour-

dot, P., Interrante, V., Nedel, L., Magnenat-Thalmann,

N., and Zachmann, G., editors, Virtual Reality and

Augmented Reality, pages 312–319, Cham. Springer

International Publishing.

CHIRA 2022 - 6th International Conference on Computer-Human Interaction Research and Applications

104