CYBERPSYCHOLOGY & BEHAVIOR

Volume 12, Number 2, 2009© Mary Ann Liebert, Inc.DOI: 10.1089/cpb.2008.0141

Rapid Communication

Behavioral Indications of Object-Presence in Haptic Virtual Environments

Miriam Reiner, Ph.D. and David Hecht, Ph.D.

Abstract

The success of many virtual environment (VE) applications relies on their ability to induce in their users a senseof presence. In immersive VE, presence is a sense of being and acting inside a virtual place, whereas in a non-immersive haptic VE, it is a sense of being able to touch and manipulate a virtual object. This latter sense ofobject-presence is typically measured by questionnaires, and the current study aimed to find objective behav-ioral indications for it. Participants moved a stylus along the blades of a virtual razor and along identical vir-tual lines with haptic feedback but without the context of a razor. Our measurements show that participants’movements were slower and they exerted less force on the stylus in the razor condition than in the simple linescondition. This behavioral pattern suggests that some degree of object-presence illusion was formed, whichcaused participants to act more cautiously in order to avoid any harm from the virtual razor.

183

Introduction

VIRTUAL ENVIRONMENT (VE) TECHNOLOGIES are advancingthe fields of medicine, engineering, education, design,

entertainment, aviation, and military training.1–3 The effec-tiveness and success of many VE applications relies heavilyon their capability to induce a sense of presence.4 For in-stance, as part of a treatment program for anxiety, a psy-chotherapist may invite a client into a VE scenario depictinghis fearful element (e.g., spiders, heights, elevators, flying)that triggers anxiety. The programmable and controlled VEallows a gradual exposure that can ease the process of reac-tivating the fear structure, desensitizing and modifying it.5–9

This gradual exposure therapy can be successful only if theclient actually has a sense of being present in the fearful sit-uation (e.g., of spiders crawling on his hands, of being in-side an elevator or airplane), thus enabling him to graduallyreexperience the phobia while deconstructing it.10–13

The sense of presence is not similar in all VE systems. Infully surrounding and immersive VE systems that capturethe entire perceptual field, through a head-mounted displayor 360-degree presentation like the CAVE, the user developsa feeling that she is actually inside a place that does not ex-ist in the real world. Thus, the sense of presence refers to hersense of being present in a (virtual) place. Whereas in non-immersive VE systems (e.g., projection tables, projection-augmented models, and haptic VEs like the Reachin®

PHANTOM desktop) users do not develop a feeling of be-ing transformed into a different place. Nevertheless, the com-bined visuohaptic sensations often generate an illusion thatthe visual scene is not merely an image of some objects: be-cause the haptic sensations are real, the user feels she is ac-tually touching and manipulating an object that does not ex-ist in the real world. Hence, the sense of presence, in thelatter VE systems, is the presence of a (virtual) object. In theexample of the gradual exposure therapy, an effective VEtechnology for agoraphobia can create in the client the illu-sion of being on a busy sidewalk surrounded by manystrangers: presence in a place. And the ideal VE technologyfor the fear of snakes will induce in the user the sense thathe is actually touching a snake: presence of an object. Thisdistinction is elaborated further elsewhere14–15 and illus-trated in Figure 1.

While the sense of being present in a (virtual) place wasdocumented not only through subjective reports but also bybehavioral and physiological measurements,16–20 the senseof object-presence is still measured by subjective reports orby the Object-Presence Questionnaire.15 The present studyaimed to develop an experiment in which objective andquantifiable behavioral measures would test this subjectivesense of object-presence. For that purpose, participants werepresented with a virtual razor, and their task was tosmoothly move a stylus along its blades in a predefined tra-jectory. The same task was performed also in a control con-

The Touch Laboratory, Department of Education in Technology and Science, Technion–Israel Institute of Technology, Haifa, Israel.

dition with identical visuohaptic feedback but without thecontext of the razor; thus the image was not perceived asblades but as just lines (see Fig. 2).

We hypothesized that when participants are presented vi-sually with a razor that is combined with haptic feedback(i.e., it can be touched), some degree of object-presence illu-sion will be generated, and participants will behave accord-ingly. Therefore, we aimed at comparing the magnitude ofthe forces that participants would exert on the stylus whilemoving it along what was perceived as “just lines” with theforces applied on it within the context of a razor so the lineswere perceived as blades. Specifically, we expected less forceto be exerted on the stylus if participants had a sense thatthey were touching blades. For the same reason, we also com-pared the movement velocities in both conditions, assumingthat if participants felt some sense of object-presence, theywould be more “cautious” and move the stylus along the ra-zor blades more slowly in the blades condition than in thelines condition to avoid any harm from the blades.

Methods

Participants

Fourteen students (7 male, 7 female; mean age, 22.6) withnormal or corrected-to-normal vision and without anyknown tactile dysfunction participated in the paid experi-ment. They were unaware of the purpose of the experimentexcept that it tested eye–hand coordination in different con-ditions. The experiment was carried out under the guidelinesof Technion’s ethics committee.

Apparatus and stimuli

We used the Reachin system with the PHANTOM (desk-top) device, a collocated system in which touch and visionoccur at the same spatial location. Participants saw a visualdisplay of 12 lines (each 5.7 cm long, 2 mm wide) tilted di-agonally. These lines were programmed to generate a cor-responding haptic sense when being touched. The partici-pants’ dominant hand that held the stylus was under the flathorizontal mirror and unseen, but the stylus was representedon the screen as a stick figure with a small ball at its tip (the

contact point), so every movement of the stylus was simul-taneously represented on the screen. The task was to posi-tion the stylus on the lines, to touch and feel them haptically,and then to move the stylus along the edges of four specificlines (the second, third, sixth, and eighth) from left to right.Time restrictions were not imposed, and participants werefree to move at their own pace. However, they were in-structed to perform these movements precisely (i.e., to moveonly along the lines’ edges) and smoothly (i.e., to performeach movement in one act, avoiding hesitations/stops whilemoving). Each participant performed these four movementtasks along the four designated lines twice: once, in the linescondition where the display contained only the lines andonce in the blades condition where the same lines were em-bedded in the context of a razor and appeared to be razorblades (see Fig. 2). To keep the lines condition free of anymeaningful associations, the order of the presentations waskept strict and constant: all participants were first presentedwith the lines and then with the razor blades. Also, prior tothe test, participants were asked, “What are these lines?” andnone of them thought that they are razor blades.

Results

Paired t test analyses revealed significant differences be-tween the lines and the blades in both the speed of move-ments and the force exerted on the stylus. Participants per-formed the similar task of moving the stylus along the linessignificantly faster, on average, than their movement veloc-ity along the blades (3.7 and 5.5 seconds respectively): t(13) ��3.39, p � 0.002. The total force participants exerted on thestylus when moving it along the lines was significantlyhigher than when they moved it along the blades (peakingat 4.2 and 2.6 Newtons respectively): t(13) � 3.21, p � 0.003.Further analysis of the separate forces exerted in the X,Y,Zdirections showed that participants exerted higher forces onthe lines, as compared to the blades, in all three directions;nevertheless, the difference was statistically significant onlyin the Z direction (t(13) � �3.01, p � 0.006). See Figure 3.

A within-participants analysis showed that the differencein the movement velocities between the lines and the bladeswas present in 12 of the 14 participants and that the differ-

REINER AND HECHT184

“Real World” “Real World”

Vir

tual

Env

iro

nm

ent

Vir

tual

Env

iro

nm

ent

UserUser

FIG. 1. Presence in a place (left): user experiences beingfully inside the environment. Presence of an object (right):user experiences the virtual object as real and can interactwith it but is not fully immersed in the virtual environment.(Illustration adapted from Stevens et al.15)

FIG. 2. Participants “followed the lines” with a stylus overan image based on the blades of a razor but without the ra-zor context (left) and over the same image set in the razorcontext (right).

ence in the force exerted on the stylus occurred in 11 of the14, indicating that both differences were characteristic ofmost of the individual participant’s performances.

Discussion

Our measurements show that participants moved the sty-lus along the virtual razor blades significantly slower thanwhen identical visuohaptic feedback was given in a neutralsetup without the razor context. Furthermore, participantsapplied less force on the virtual razor blades than on thevirtual lines. As can be seen in Figure 3, the main differenceoccurred in the force exerted in the Z direction. The reasonfor this is that the diagonally tilted position of both the linesand blades on the screen required participants to exert theforce mainly in the Z direction and only slightly in the Xand Y directions. The behavioral pattern observed here sug-gests that the visuohaptic presentation of the razor gener-ated some degree of object-presence illusion that caused par-ticipants to adjust their behavior accordingly. Consequently,the velocity of their hand movements and the forces theyexerted on the stylus were affected by some degree of care-fulness that was not present when they did the same taskon the lines.

Although participants always started with the lines con-dition and followed with the blades condition, it is very un-likely that the differences in the force and speed betweenthese two conditions reflect a learning effect. First, repetitionof similar tasks and stimuli tend to gradually decrease re-sponse times, a phenomenon known as repetition priming ef-fect. Typically, prior experiences influence the processing ofsimilar subsequent events, which is expressed in faster andsometimes more accurate performance in the succeedingevents.21–22 Thus, the opposite pattern of increased movementtime of our participants in the blades condition, comparedto the former lines condition, cannot be explained in termsof learning and skill acquisition. Furthermore, it strengthens

our suggestion that despite some learning that may have oc-curred in participants’ motor skills related to the given taskwhich should have shortened the total movement time, themovements along the razor blades were slower, possibly asa result of experiencing some sense of touching a razor (andnot just an image of a razor) and consequently proceedingmore slowly and carefully because the blades can harm theuser. Second, in regard to the forces applied on the stylus, ifthe less force applied in the blades condition reflects a learn-ing effect, one may expect to find a gradual learning curvealso within the four trials in the lines condition (i.e., the forcesshould have decreased with each trial) and a similar patternwithin the four trials of the blades condition until a learningplateau was reached. However, there were no significant dif-ferences in the average force between the four trials withineach condition, whereas a significant difference was foundbetween the two conditions.

This behavioral demonstration of a sense of object-presencecorresponds with the general theory of embodied cognitionand emotion: the idea that cognition and emotion arise frombodily interactions with the world, and as such, cognition andemotion are tightly linked to the physiological and sensori-motor operations of the organism.23–24 Furthermore, the cur-rent study is in accord with the approach that the sense of pres-ence is grounded in action—the ability to act and perform inthe virtual environment or on the virtual object—and that pres-ence will be increased when interaction techniques are em-ployed for the user to be engaged in body movements.25,4

Engineers and designers of haptic VEs testing the effec-tiveness of their systems and applications in generating theillusion of object-presence may find the objective measuresof force and task-completion time very useful for comparingusers’ behavior in the real world with their behavior in VEs.The specific method used here with the razor blades may beuseful for testing the effectiveness of a future haptic tool thatwill simulate the tactile and force-feedback sensations of a“finger placed in a thimble” within a VE.

BEHAVIORAL INDICATIONS OF OBJECT-PRESENCE 185

Movements along lines/blades

Time (seconds)

Forc

e (N

ewto

ns)

4.5

4

3.5

3

2.5

2

1.5

1

0.5

00 1 2 3

Fx

Fy

Fz

Fz

Fy

Fx

Total Force

Total Force

Lines Blades

Blades

Lines

4 5 6

FIG. 3. Velocity of movement and force exerted on the stylus along the lines and blades.

Acknowledgments

This research was funded by the EU project IM-MERSENCE—Multimodal immersion into interactive vir-tual environments. We thank Mr. Gad Halevy for his pro-gramming help and Mrs. Tatiana Gelfeld for her help in thedata analysis.

Disclosure Statement

The authors do not have a conflict of interest.

References

1. Stanney KM, Mourant RR, Kennedy RS. Human factors is-sues in virtual environments: a review of the literature. Pres-ence: Teleoperators & Virtual Environments 1998; 7:327–51.

2. Holden M. Virtual environments for motor rehabilitation:review. CyberPsychology & Behavior 2005; 8:187–211.

3. Satava RM. Historical review of surgical simulation—a per-sonal perspective. World Journal of Surgery 2008; 32:141–8.

4. Sanchez-Vives MV, Slater M. From presence to conscious-ness through virtual reality. Nature Reviews Neuroscience2005; 6:332–9.

5. Rothbaum BO, Hodges L, Smith S, et al. A controlled studyof virtual reality exposure therapy for the fear of flying.Journal of Consulting & Clinical Psychology 2000; 68:1020–6.

6. Emmelkamp PM, Krijn M, Holsbosch AM, et al. Virtual re-ality treatment versus exposure in vivo: a comparative eval-uation in acrophobia. Behavior Research & Therapy 2002;40:509–16.

7. Garcia-Palacios A, Hoffman H, Carlin A, et al. Virtual real-ity in the treatment of spider phobia: a controlled study. Be-havior Research & Therapy 2002; 40:983–93.

8. Rothbaum BO, Hodges L, Anderson PL, et al. Twelve-monthfollow-up of virtual reality and standard exposure therapiesfor the fear of flying. Journal of Consulting & Clinical Psy-chology 2002; 70:428–32.

9. Hoffman HG, Garcia-Palacios A, Carlin A, et al. Interfacesthat heal: coupling real and virtual objects to treat spiderphobia. International Journal of Human Computer Interac-tion 2003; 16:283–300.

10. Glantz K, Durlach NI, Barnett RC, et al. Virtual reality (VR)for psychotherapy: from the physical to the social environ-ment. Psychotherapy 1996; 33:464–73.

11. Rizzo A, Buckwalter JG. Special issue: virtual reality appli-cations in neuropsychology: guest editor’s introduction.Presence: Teleoperators & Virtual Environments 2001; 10:iii-v.

12. Hoffman HG, Richards T, Coda B, et al.The illusion of pres-ence in immersive virtual reality during an MRI brain scan.CyberPsychology & Behavior 2003; 6:127–31.

13. Krijn M, Emmelkamp PMG, Olafsson RP, et al. Virtual re-ality exposure therapy of anxiety disorders: a review. Clin-ical Psychology Review 2004; 24:259–81.

14. Stevens B, Jerrams-Smith J. The sense of object-presence withprojection-augmented models. Lecture Notes in ComputerScience 2001; 2058:73–5.

15. Stevens B, Jerrams-Smith J, Heathcote D, et al. Putting thevirtual into reality: assessing object presence with projec-tion-augmented models. Presence: Teleoperators & VirtualEnvironments 2002; 11:79–92.

16. Dillon C, Keogh E, Freeman J, et al. (2000) Aroused and im-mersed: the psychophysiology of presence. Proceedings of the6th Annual International Workshop on Presence. Eindhoven,The Netherlands.

17. Wiederhold BK, Jang D, Kaneda M, et al. (2001) An investi-gation into physiological responses in virtual environments:an objective measurement of presence. In Riva G, GalimbertiC, eds. Towards cyberpsychology: mind, cognition, and society inthe Internet age. Amsterdam: IOS Press, pp. 176–83.

18. Meehan M, Insko B, Whitton MC, et al. (2002) Physiologicalmeasures of presence in stressful virtual environments. Pro-ceedings of the 29th Annual Conference on ComputerGraphics and Interactive Techniques (SIGGRAPH). San An-tonio, Texas, pp. 645–52.

19. Slater M, Brogni A, Steed A. (2003) Physiological responses tobreaks in presence: a pilot study. Proceedings of the 6th AnnualInternational Workshop on Presence. Aalborg, Denmark.

20. Bailenson JN, Blascovich J, Beall AC, et al. Interpersonal dis-tance in immersive virtual environments. Personality & So-cial Psychology Bulletin 2003; 29:1–15.

21. Tulving E, Schacter DL. Priming and human memory sys-tems. Science 1990; 247:301–6.

22. Hauptmann B, Karni A. From primed to learn: the satura-tion of repetition priming and the induction of long-termmemory. Cognitive Brain Research 2002; 13:313–22.

23. Wilson M. Six views of embodied cognition. PsychonomicBulletin & Review 2002; 9:625–36.

24. Niedenthal PM. Embodying emotion. Science 2007; 316:1002–5.

25. Slater M, Steed A, McCarthy J, et al. The influence of bodymovement on subjective presence in virtual environments.Human Factors 1998; 40:469–77.

Address reprint requests to:Dr. David Hecht

The Touch Laboratory, Gutwirth BuildingTechnion–Israel Institute of Technology

Haifa, 32000Israel

E-mail: davidh@tx.technion.ac.il

REINER AND HECHT186