Postural instability-induced compensative movements in virtual reality

During virtual reality usage, two egocentric mental representations are constructed simultaneously. The first representation is rooted in the physical reality in which VR is set up, and the second originates from the mental construction of a computer-generated virtual environment. In both cases, participants configure their posture based on multimodal stimuli while responding to environmental cues. In most cases, the postural cues provided by the digital and real environment may be conflicting. In this study, 50 right-handed volunteers were enrolled. In a pre-test session, attentional focus-related personality bias (perspective-taking) was assessed, and afterward, postural movements and presence experiences were measured while the participants performed a spatial orientation task in VR. Participants were placed in an upright position with their right hands positioned in front of a physically real point on the laboratory wall. Afterward, participants were exposed to a VR environment in which they performed a room-tilting task. Participants with higher hand-related presence scores showed decreased compensatory hand drift in the VR environment. The rate of contralateral hand drift showed a reversed association with the intensity of the perspective-taking trait. VR-induced postural instability can be attenuated by the compensative hand drift that alleviates the conflicts between the two rival inner VR and outer VR environments that compete for attention and provide different reference cues.


Introduction
Postural stability and the related confident body representation are constructed by the integration of multisensory stimuli involving sensory modalities-tactile, proprioceptive, vestibular, auditory, and interoceptive cues-with the vision of the body and its spatial environment (Azañón et al. 2016). Semantic and contextual congruency is fundamental for efficient integration, and the manifested stimuli should be potentially related to a defined object where attention is allocated (Freiherr et al. 2013;Spence 2011). The integration of multisensory information by the body is an automatic, uncontrolled process (Kennett et al. 2002) that forms a body schema that serves as a permanent frame of reference to start motor acts depending on the person's level of arousal or mental state. This reference system is considered the background for the successful navigation and control of one's body movements and reaction to physical and virtual environmental challenges. It also plays a role in multimodal integration and temporal and spatial matching of the context, 1 3 creating a bridge between interoceptive and exteroceptive stimuli while coordinating hand, head, and trunk positions and movements (Gallagher 2006). However, the mismatch in the multimodal stimuli coming from the hand and head in contralateral or ipsilateral space frequently yields conflicting perceptual and motor responses (Stamenskovic et al. 2018) and yields postural instability. Compensative bodypart movements serve as the main agent in the static and dynamic control of one's posture in real and virtual space (Crawford et al. 2004;Guerraz and Bronstein 2008;Ha et al. 2014;Horak and Macpherson 1996). Postural sway and the coordinated movements of the head and chest are the basic conditions for a stable stance, successful defensive behavior, and goal-directed movements. When conflicting environmental cues cause postural instability, diverse patterns of arm and head movements are available to maintain or recover balance stability. Balancing movements and reflex reactions can be described in a Cartesian x, y coordinate system that can provide reliable data depicted on the horizontal and vertical lines about the movements of the hands and head (de Souza et al. 2019).
For individuals with higher anxiety levels, the aforementioned postural compensative behavior is associated with sensitive responses, such as vertigo, motion sickness, unstable gait, and postural difficulties. However, in extreme cases, the compensative repertoire is narrowed, and postural sway is reduced; moreover, spatial orientation disorders and physical support searching can manifest in adults and children (Blanchard and Blanchard 2008;Jacob et al. 2009;Jacobs and Nadel 1999;Kállai et al. 2007). The postural system is considered a multimodal perceptual and response surface. Personality traits and neurocognitive conditions influence both psychological uncertainty and postural functions in healthy participants (Stins et al. 2009). The present study aimed to determine whether a person who is participating in a virtual reality (VR) scenario can match visual and postural cues while performing tasks in VR and a physically real environment synchronously.

Presence
The sense of presence manifests when the mind and the body are consciously immersed in a mediated environment, where they attend the same place at the same time in a virtually constructed spatial context (Witmer and Singer 1998). While the mediated space is constructed by the users, their activity is goal-directed, and meaningful, and induces coherent changes in social and physical peripersonal space. Furthermore, users are aware that this relation between the computer-generated space and themselves could be separated (broken) at their will (Salter 2009). The definition of presence involves different personality traits, cognitive controlled affective relations, and capabilities used to create a meaningful reality, while the participants perceive that they are located in the present space (being there) and suppose that adequate action can be executed in this mediated environment (Wirth et al. 2007). Several personality trait predispositions were found to be associated with various components of VR presence (reviewed by Wallach et al. 2012). The intensity of the presence, among other things, depends on the user's attachment styles, the presence of psychopathological cases with weakened self-other boundaries (Bouchard et al. 2005;Kállai 2019;Riva 2009), and the emotional valence of the events (Riva et al. 2007;Thacher et al. 2005;Wallach et al. 2009). There is strong evidence that general immersive tendencies in a mediated environment (watching films, reading a well-written and immersive book, listening to classical music, etc.) predispose individuals to report a high-level presence in and after VR exposure (Laarni et al. 2005). Furthermore, higher VR-induced presence was found in participants with strong empathy, in response to certain scenes, and individuals with elevated reactions to suggestions (Nicovich et al. 2005;Wallach et al. 2010). Considering the abovementioned personality predispositions in the present study, a specific reference frame-relevant factor of empathy (capability of perspective-taking, PT) is preferred for revealing the interaction between VR-induced postural instability and mental representation of synchronically constructed VR and physically real space. All these traits are related to one's ability to control reality, involving both affective and cognitive components. Movements without uncertainty in VR depend on the appropriate adequate cognitive and mental capabilities to control action-perception relationships and a personality predisposition that allows one to be willing to step out from the immediate physical reality and step into a VR environment. This conscious movement between immediate physical and computer-generated virtual environments frequently leads to instability in the regulation of the spatial reference since the body and the related self cannot be present at two different times and in two different spaces. However, this duality is not impossible: the spatial representation of the experience of being someone located somewhere is different from the mental representation of doing something anywhere. Sanchez-Vives and Slater (2014) and Wirth et al. (2007) described a spatial presence theory in which the spatial presence of self-location and the spatial presence of possible actions are defined as different but integrated presence types.

The mental construction of the spatial environment in VR
A spatial constriction is based on two different mental representations (egocentric and allocentric). For an egocentric reference, the environment is scanned from the observer's point of view, where the environmental cues are defined by the observer's right or left half of the body or by the frontal plane or back-side of the observer. An allocentric reference focuses on the object and its spatial relations using a map-like representation of the environment (Klatzky1998; O' Keefe and Nadel 1978). In a closed space, such as a physically real laboratory or a virtual laboratory room, the egocentric Euclidian geometric structure is dominant. Consequently, the user's presence experience in VR can be originated from different cues of spatial location. This means that a person's location in the current physical space can be described in two egocentric frames of reference simultaneously. One frame is focused on the visuospatial stimuli inside (inner) the VR, and the other frame is attended to the actual physical real environment outside (outer) the VR. The mental representation of the inner and outer VR environment has not been isolated from each other, but the rate of presence experience and the associated attention allocation capable to select between the current spatial representations. E.g.: While a man stands in an uptight postural position and his right hand is fixed to a distant target in front of his center of the visual space and his body is suddenly tilted to the right side, his hand is automatically moving toward the left from the fixed target. The role of this compensative hand drift (CHD) is to keep his hand toward the target and secondary to keep his postural balance. The person in VR gives a real perceptual and motor response to the constructed visuospatial environment as if it may be a proper reality while the individual mentally keeps the visuospatial maps of his physically real environment at the same time. From this point of view regarding spatial representation, a person's presence in the local space can be defined through more different or transient frames of reference (Meilinger and Vosgerau 2010). This means that a dual egocentric representation can be considered when a person is present in a VR environment.

Hypothesis
In a VR two mental representations are constructed from ongoing events at the same time. The outside area of the VR environment is the physical reality where the VR device is installed. The inside area of the VR environment is a computer-generated space, where the participant is immersed and a sensory visuospatial field has been constructed and mediated. In both cases, the contingency-based automatic multimodal integration configures the posture and the body schemata for responding to synchronous or asynchronous conflicting stimuli flow. On the other hand, postural automatisms support the person's stability (harmony of hands, head, and trunk movement) and generate contralateral or ipsilateral balancing movements (Wulf 2013). The postural maneuvers ensure that the person has a coherent experience of being in the environment and confirms the adequate functioning of different components (head, hand, etc.) of the body schema, and strives to solve problems between contradictory visuospatial inputs. This study aimed to explore the nature of this dual and rival spatial representation. We suppose that the rivalry for the attention allocation resources between the two frames of reference reflects a spatial representation conflict that is constrained when VR and physical reality coincidence are provoked. We assume that the lower extent of the compensatory contralateral hand drift (CHD) that is induced by the tilting of the VR room is associated with a lower rate of presence experienced since the person strives to maintain close contact with the outer area of the VR environment. In contrast, for participants with a high presence, the degree of compensatory CHD was expected to be greater.

Participants
In this study, 50 right-handed volunteers were enrolled from a Facebook group for undergraduates (females = 26, mean age = 22.6 years, standard deviation (SD) = 2.7 years; males = 24, mean age = 22.6 years, SD = 2.7 years). The sample size was supported by computing estimated statistical power (with H1 ρ 2 = 0.3 and 1-β = 0.8 the minimum required sample size was 38) using the G*Power software v3.0 (Faul et al. 2007). The handedness of all the participants was checked by the Edinburgh Handedness Inventory (EHI) (Oldfield 1971): minimum criterion for right-hand dominance was 70% or higher, handedness mean = 93; SD 7.8; the EHI scores ranged from 100-70%. No participant had any previous psychiatric or acute somatic illness. To control the incidentally manifesting visual or movement acute or chronic cognitive deficits as a screening test, the Trail making B (TM_B) task was used (Lezak 1995;Reitan and Wolfson 1985). A lower score indicated better and intact performance in executive function. In the present study, the TM_B mean = 55.96 and SD = 13.2. A clinically significant high score in TM_B was not found. Each participant was blind to the hypothesis tested by the study. Participation in the study was voluntary, and informed consent was obtained for the participants. The privacy rights of subjects have been observed. The participants received a small monetary compensation for participation (equivalent to 10 USD). The study adhered to the principles of the Declaration of Helsinki and was approved by the Regional Research Ethics Committee of the Medical Center of the local university (Ethical Allowance No. 14 6732 PTE/2017).

VR device and programming
The software of the experimental design was created in the 2019.3f1 version of the Unity game engine. The application's target hardware for rendering was the HTC Vive 2018 head-mounted wireless display (HMD) and its controller. The development of the software was made in the C# programming language, and the user interface was created with Unity's own UI kit, which fits the interface on all screen sizes proportionally. The application employed a custom user interface designed specifically for laboratory experiments. The interface was designed to be as simple as possible and utilize all the features embedded in the application. The software used a hybrid virtual-physical calibration approach to fit the coordinate system of the virtual space with the physical location. On the first screen (the calibration and data input screen), the conductors of the experiment could calibrate the virtual room with HTC Vive controllers by entering calibration mode and physically attaching one of the handheld controllers to a cross-shaped target point in the physical room. The physical target point and its exact pendant in virtual reality were presented in the same x-y coordinates. In this way, a calibration was performed where the two target points were synchronized and located in a unified origin point of a Cartesian coordinate system. This was a crucial step to maximize the precision of the automatized data recording. The calibration data were saved into a JSON file on the local hard drive of the PC that was used to run the experiment. This calibration file was loaded every time the software started up. All experimental data, the location of the target points, and the room size were stored locally in JSON files, making them accessible to the laboratory so that new virtual setups and circumstances could be created. The main drivers of the experiments were the position, rotation, and relative rotation of the HMD and the actuator of the controller. The data measured for the devices were stored in a CSV file. The data sampling frequency rate of the device states was 4 Hz.

Interpersonal reactivity index (IRI)
The IRI (Davis 1983;Kulcsár 1998) scales contain 28 items rated on a 5-point Likert scale ranging from 1 (= does not describe me well) to 5 (= describes me very well). The questionnaire involves PT, fantasy, empathic concern, and personal distress scales. Considering our examination hypothesis, PT was applied to assess participants' personal tendency to assume a more visuospatial perspective so that they could understand an alternative potential meaning of a current scenario. Individuals with a higher score on the PT scale spontaneously try to adopt more perspectives. While they are scanning a scene, their attention focus frequently fluctuates between an inner egocentric or another person's perspective. In this study, the internal consistency of the IRI questionnaire and the PT scale (IRI_PT) was good (Cronbach's α for: IRI = 0.852; and PT = 0.776).

Presence experience
The participants were immersed in the VR environment and immediately after finishing the room tilting task were asked to provide a self-report on the experience of the presence; on their body presence (inner space of the VR environment) and also were asked to report about the right-hand presence that was fixed on an invisible target in the outer space of the VR environment (outer VR presence). In both cases, a scale was used verbally and ranged from 0% (= no presence) to 100% (= absolute presence). Considering the individual differences in inner VR and outer VR attentional allocation, a derivative score was calculated that originated from the ratio of inner (body) presence/outer (hand) presence and was used as the inner/outer divided presence index (DPI).

Procedure
Under normal conditions, postural instability is experienced while traveling on moving airplanes, ships, and other means of transport or while climbing a high wall where the egocentric frame of reference and the expected stability of the surroundings do not correspond. When a postural imbalance is induced by a natural or VR environment, a compensatory postural response is available to help individuals recover their balance (Oude et al. 2009). For instance, while traveling on a turbulent sea, travelers realize that the ship is constructed on horizontal bases with a stable Euclidean geometric structure, and their posture is adapted to this stable vestibular and visual reference. To compensate for the turbulent tilting effects, travelers need to change their postural position to maintain postural stability, which is possible via repeated compensatory body-centered, arm-centered movements. Similar balance movements can be elicited during a video game and in VR-induced visual space manipulation. Under the present experimental conditions, two egocentric spatial representations were generated at the same time.
In the first step of the experiment, the outer VR egocentric frame of reference was established. Guided by an assistant, the participant stepped into the physical room of the laboratory without wearing the HMD. The laboratory room was considered a classically defined static Euclidean threedimensional space with white walls; its dimensions were 8 m in length, 7 m in width, and 5 m in height. After freely exploring the room, the participant was guided to stand upright in a defined location of the room and asked to fixate on a cross mark that was located on the laboratory wall via the laser pointer of the controller; the mark was 1.5 m from the floor and 4 m away from the participant. Afterward, the HMD was placed on the participant's head. The participant was also provided with a VR controller in his or her right hand. After a short period of relaxation, the blank screen changed, and via the HMD the virtual representation of the laboratory room with the target on the wall was presented (as a matter of course, the participant's targeting hand with the controller was visible to the subject). The participants were instructed to target the cross with the controller. The targeting time was 5 s. While the target was fixed on the cross, the following instruction was conveyed via the earphone: "Please continue to fix the target of the controller on the cross and fix your mind on the same place in the room, irrespective of what you see next in the VR environment".
In our experimental design, a lateral modification of the horizon in VR has been generated by a VR room tilting. This postural imbalance induces ipsilateral and contralateral movement automatisms that are manifested in head tilting in the direction of the room tilting, and the compensative balancing movement is manifested in hand drift toward the contralateral direction of the head and room tilting. The room tilting-induced head tilt is a relatively consciously controlled compensative act that moves along a curve from the vertical axis of the body starting from midline 0 points of the body axis and moving toward the right or left shoulder. The head tilt magnitude can be measured in degrees. In contrast, hand drifting is a spontaneous compensative movement that is realized along a horizontal line in front of the body and can be measured in hand drift magnitude from starting point to the end point of the hand drift in millimeters (see Fig. 1). The head tilt and hand drift move different paths and their movements representable in a body-centered and a target centered x, y Cartesian coordinate system.
After finishing the mental recording of the outer VR egocentric frame of reference, the construction of the inner VR egocentric frame of reference began. (The terms of the outer VR egocentric and the inner VR egocentric references and representations reflect the theoretical focus of this study, i.e., we analyzed the mental connection and representation transmission between VR and physical reality). The outer VR reference focused on the visuospatial cues of the laboratory room where the examination took place. The inner VR reference focused on the visuospatial cues of computer-mediated VR rooms, while the attention allocation for the outer VR room remained partly irrespective. After fixing the target in a VR environment, the HMD screen became black. After a VR copy of the previously shown laboratory room with pale blue walls was presented, however, the virtual room had no contained target, and the targeting hand and the controller were not visible. When the participants viewed the presented VR, they were instructed as follows: "You have entered a VR in which you will shortly experience the room tilting towards its right and left for a short period. However, corresponding with previous instructions, please note that your hand should target the fixed point in your mind, with the controller". In this context, the frame of the room tilting (right or left) depending on the participant's postural positions, and the HMD transmitted visuospatial cues were perceived in an egocentric spatial frame of reference (inner VR egocentric). The time scale of the VR-induced postural instability task. (see Fig. 2).
Consequently, during the tilting room stage of the experiment, two rival egocentric spatial frames of reference were present at the same time to represent the event in VR. The effect of the outer VR egocentric representation may be manifested in the horizontal drift of the right hand. However, the inner VR egocentric representation may be embodied in the room tilt-induced lateral head tilting. Across the whole course of the examination, the spatial positions of the VR hand controller and the HMD positions were recorded automatically and uploaded (4 Hz sample taking) into a CVS database. After the task accomplishment, two state-related presence experience ratings were given verbally, ranging from 0-100%. The first indicated the presence state rating for the room tilting experience Head and body (inner) Present VR). The second indicated the presence state rating for the presence of the targeting hand while completing the VR room-tilting task (Hand (outer) Present VR). This means that in the experimental design, two presence experiences were induced. Body presence VR refers to inner VR egocentric representation, and hand present VR refers to the reference frame of the outer VR egocentric representation. The first is linked to the head and body and the second is related to the right hand. The ratios of the head, body presence, and hand presence were accounted for as a divided presence index (DPI) that reflects the rate of the difference in presence experiences related to the dominant use of inner VR egocentric or outer VR real egocentric reference frames while a person performs a conflicting VR task. During the examination and VR room tilt induction, considerable adverse or unpleasant events were not detected.

Data analyses
The position of the head and the hand are defined in an x, y coordinate system where the participant's head was the base when the participant was standing upright. The head tilt-induced deviation from zero was defined in degrees where the right-side deviation had a positive score and the left-side deviation had a negative score. The position of the tilt-induced hand drifting was defined in mm. The rightside drift had a positive score, and the left-side drift had a negative score.
First, we tested the VR lateral room tilt effect on posture changes as defined by head tilt and hand drift. We used onesample Student's t-tests to compare the head tilt to 0 degrees (upright position) in four conditions (room tilted to the right or left, and two trials each). We used paired samples Student's t-tests to evaluate the difference between the baseline hand position and the hand position after the tilt, again in four conditions. Then, we tested head and hand laterality sensitivity to VR-induced postural instability. For this purpose, we calculated the absolute magnitude of the mean head tilt and hand drift values of trials one and two for both sides. Using paired samples Student's t-tests, we compared the VR-induced magnitude of the head tilt to the right and left and the VRinduced magnitude of the hand drift to the right and left.
After this, we examined the role of PT and focus of attention allocation in balancing contralateral hand movements and in the spatial construction of the VR environment. We created three equal groups based on the participants' DPI scores. We used ANOVAs to test group differences regarding the right/left-hand drift ratio and the right/left head-tilt ratio. We used Person's correlation to examine the relationship between DPI scores and the right/left-hand drift ratio and the right/left head-tilt ratio.
Finally, we examined the effect of PT personality bias on hand drift and head-tilt scores. First, we calculated the mean value of the absolute magnitude of right-and left-hand drift scores and the mean value of the absolute magnitude of right-and left-hand head-tilt scores. We used linear regressions, where the independent predictor was the IRI_PT scale with either the hand drift or head-tilt scores as the dependent Fig. 2 The time scale design of the VR-induced postural instability task. Note: The right and left tilting frames, plus the time of the horizontal zero breaks required 51 s. Four horizontal room-tilting positions were defined and applied to tilt the room toward the right or left side. RIGHT: 1. Horizontal frame without tilting 0° for 3 Sect. 2. Room tilting to the right 20° for 3 Sect. 3. Staying there, tilted at 20° for 3 s. Room tilting back from 20° to 0° for 3 s. LEFT: 1. Horizontal frame without tilting 0° for 3 Sect. 2. Room tilting to the left 20° for 3 Sect. 3. Staying there, tilted at 20° for 3 s. Room tilting back from 20° to 0° for 3 s. The left and right tilting sequences were repeated in sequential order two times. The mean speed of the tilting was 6.67 degrees/sec variable. Age and sex were entered as random effects in both models.
For all significant results, we report the Cohen's d (pairwise comparisons), the partial eta squared (ɳ p 2 , ANOVAs), or the R 2 and the standardized estimate (β, regressions) values as effect sizes. All analyses were performed using JAMOVI version 2.0 (The Jamovi Project 2021).

Results
For the detailed descriptive statistics of personality and presence scores, please see Table 1.

VR lateral room tilt effect on posture changes (head tilt and hand drift to the lateral or contralateral space)
The head tilted parallel to the direction of the VR room tilt shows that the tilting-related ipsilateral postural change is significant (Table 2). However, the hand movement during the VR room tilt drifted contralaterally to the direction of the VR room tilt. The mean rate of the change from a resting position to the final stage of hand drifting is described in Table 3.
The contralateral movement patterns in Trials I and II in the right and left VR room-tilting probes are shown in Fig. 3.

Analysis of head and hand laterality sensitivity to VR-induced postural instability
The within-subject comparison showed laterality differences in hands drifting toward the contralateral hemispace parallel to the VR-generated room tilting. Generally, when the VRgenerated room tilted rightward or leftward, the balancing hand drifted in the opposite direction as the room and the head tilted. The amplitude of this mediolateral compensative hand drift was significantly higher when the contralateral compensative hand drifted to the right side of the constructed VR space. Other lateral VR-generated postural movement differences between the left and right tilts were not found. Please see Table 4 for the exact statistical values.

Table 3 VR room-induced CHD (in mm)
Note: After VR room tilting, the participant's hand in the absence of a visible target cannot exactly revert to the next starting position, which is based on the middle axes of the body. Consequently, in the starting position during the next room tilting phase, the hand spatial position is relative. In that way, the scoring contained the route (in mm) that was achieved during the initial room tilting and final room tilting periods. Hand drifts toward the right side of the hemispace were scored as positive, and drifts toward the left hemispace were scored as negative

3
The lateral deviation of the horizontal balancing movements showed elevated rightward hand drift compared to the leftward hand drift in the VR. In head movement data, a similar difference could not be detected. This means that when a participant was instructed to use his or her right hand for fixing a target that was located in the outer space of the VR environment, the VR environment-induced postural instability, and the compensating CHD was stronger in the right hemispace of the VR environment than in the left hemispace of the VR environment. Therefore, while an individual is standing in an upright position and with the right hand targeting a recalled object from the outer area of the VR-that is not visible to the participant-the VRinduced visuospatial room tilting resulted in systematic contralateral compensative stabilization by a hand drift, similar to what materializes in a physically real environment. In right-handed individuals, the rate of compensatory drift was higher in the right VR hemispace than that in the left hemispace.
Furthermore, we explored the associations of this divided inner and outer VR focus with the intensity and laterality of the detected CHD. The response to these questions comes from the analysis of the relations of presence experiences. Two different presence scores were applied: one referred to the body presence (inner VR focus) that originated from the visuospatial induction of the participant's body-related sensations and the current cues from the VR visuospatial environment. The other presence referred to a part of the body, namely the right hand that was present as a transmitting link between the VR environment and the physically real environment.

Role of PT and attention allocation in balancing contralateral hand movement and the spatial construction of a VR environment
Examining the association between the rate of inner and outer focused presence, a significant relation was not found (r (sex, age controlled) = 0.280, n.s.). This result reveals that the divided attention allocation between the body and hand-related presence may depend on a personal variable. Some participants inhibited the manifested VR-triggered  presence experiences, while others were immersed in VR.
Considering the individual differences in VR inner and VR outer attentional allocation, a derivative score was calculated that originated from the ratio of body inner presence/hand outer presence and was used as the DPI. A higher score in the DPI indicated a higher immersion tendency of the inner visuospatial cues of the VR environment. A medium score referred to a harmonized attention allocation between the VR inner and VR outer spaces. Furthermore, a lower score can be considered to indicate a lower focus on the VR inner space and an elevated focus on the outer VR to the physical real environment. Our additional aim was to test the connection between the inner/outer attention allocation strategy and the postural instability VR task as defined by the right/left hand-drift and head-tilt ratios.
The correlational analysis revealed that DPI indicated differences between the locus of the attention allocation while a participant was present in a tilted VR room. When the participant's presence intensively focused on the visuospatial cues of the inner VR, contralateral hand drifting was greater in the right hemispace of the VR (r = −0.346, p = 0.017). In contrast, when the participant focused on the outer area of the VR environment rather than the physically real space, the CHD was greater in the left hemispace of the VR environment. The correlation for head tilt ratio was nonsignificant (r = −0.131, p = 0.382).
The analysis of the VR room tilt data showed that the degree of tilt-induced hand drift on the x-axis indicated a smaller left or right-hand drift in individuals with a higher score on IRI_PT scales. This means that the spatial location of the outer VR target was retrieved correctly in individuals with high PT skill scores compared with individuals with low scores on this PT scale.

Discussion
This study focuses on the relationship between VR presence experiences and egocentric spatial representation in VR. It was supposed that the induced spatial representation conflict induces a postural instability that generates contralateral head and hands postural compensative movements. Furthermore, the manifesting compensative movements are linked to PT personality predispositions and inner VR and outer VR presence experiences while a person is exposed to a VR room-tilting task. The VR task involved the construction of two rival frames of reference for attention allocation at the same time. In addition, an analysis was conducted on the role of PT personality bias, which may be a potential score to evaluate the direction of the attention allocation in virtual reality, while an intensive spatial representation conflict invades postural stability.
The first part of the analysis focused on the effect of VR room tilting on postural instability and the induced compensative movements to maintain balance. The participants standing in an upright position were able to follow the tilt of the room by lateral or contralateral head movements or by hand movements along the x and y axes in a Cartesian coordinate system. Our data showed that room tilting-induced postural instability was attended with an ipsilateral head tilting and the imbalance was compensated by a contralateral hand drifting (CHD). When the VR room was tilted to the right, participants' heads tilted rightwards, and their targeting hands drifted toward the left side of the hemispace Fig. 4 Association between the PT subscale of the IRI and the absolute hand drift magnitude (left) and the absolute head tilt magnitude (right) during VR room tilting-induced postural instability of the VR. Alternatively, when the room tilted to the left, participants' heads tilted toward the left, and their hands drifted toward the right hemispace of the VR environment. This movement pattern is a standard mechanism to compensate for postural instability in a physically real environment (Horak and Macpherson 1996;Ivanenko and Gurfinger 2018). Our presented data have attested that the mentioned head tilt and CHD manifest during VR room tilting similarly.
Hand drifting plays a dual role in the solution of the VR-generated conflict between the construction of the two frames of reference. In the first case, the individual focuses on the inner visuospatial cues of the VR environment. In the second case, the individual focuses on the outer cues of the VR, which represent the environment where the VR was embedded. The applied experimental design allowed users to focus on the inner VR and outer VR reference at the same time. Our data showed that usage of the dual reference at the same time in the visuospatial construction of the current representation of reality induced splitting between the sense of the inner and outer VR-related presence experiences. It was supposed that the intensity and the laterality rate of the targeting hand drift would reflect the manner of the solution of this dual reference frame conflict. During the room tilt, despite the participants' intention, the hand was unable to remain in a defined position in the real physical space. The drift of the hand showed an essential intensity and laterality difference while the individual tried to adapt to the demands of the VR task. The rate of compensatory drift was more intensive in the right VR hemispace than in the left VR hemispace. Therefore, the tilt-induced horizontal CHD is asymmetrical. These results can be interpreted in several ways. The role of brain hemispheric lateralization in visual and spatial construction in a VR environment may be a potential agent. However, the number of experimental reports on the relations between VR construction and brain lateral activity patterns is incomplete, and the possible lateralization effect is debatable (Mishra et al. 2021). Our presented examination in this part of the investigation included no controlled neuropsychological data. Therefore, we cannot participate in the VR vs. brain cognitive functional laterality debate.
The second part of the data analysis focused on the relations between the CHD and the quotient of the inner VR/ outer VR presence (DPI). The presented data showed that the rate of compensatory drift is predicted by the current self-reported presence experience. While a participant was exposed to a postural imbalance task in the VR, the lateralization error for attention allocation in the inner or outer VR environment depended on the rate of the DPI. When the participant's presence intensively focused on the visuospatial cues of the inner VR, contralateral hand drifting was greater in the right hemispace of the VR. Conversely, when the participant focused on the outer area of the VR, the physically real space, the CHD is greater in the left hemispace of the VR.
As a third perspective, it was stated that the rate of VR room-induced hand drift was greater when participants had not practiced the skills of PT. However, individuals with high scores on the PT scale can easily move from one spatial or mental perspective to another, shifting the focus of attention from one perspective to another. In contrast, the lower rate of PT contains narrowed reference-changing capability. Taken together, PT plays a role in the manifestation of the intensity of absorptive, immersive tendencies and mobilizes a contrasted cognitive strategy pattern that is activated when a person is immersed in a VR environment. Considering the definition of PT (Herrera et al. 2018;Pulos et al. 2004;Zaki 2014), alterations in mental focusing are advantageous in resolving complex visuospatial problems and maintaining a stable and comprehensive spatial environment for adjustment to task requirements. Following this definition, the difficulty in maintaining a parallel, more spatial reference is associated with decreased agency experiences in VR because the personal attention allocation resources are focused on two visuospatial reference frames at the same time and same place. Consequently, the elevated outer VR presence attenuates the articulation for the perception of the visuospatial cues in the VR and augments the visuospatial construction at the same time in the environment where the VR has been embedded. We suggest that intensive lateral deviation of the hand while a VR room is tilting in the left or right direction may be considered a clear index of the sensitivity to the rate of the presence and immersion into the virtual environment.
In a general view, our presented results suggested that inner (body-focused) presence is associated with the current visually created VR while the outer physical environment is neglected. On the other hand, the outer (targeting handfocused) presence is associated with low engagement in the creation of visual-spatial cues in VR. This experimental design was designated to induce a representational conflict between the inner and outer frame of reference simultaneously while the participant's body was in the same place (inner and outer VR at the same time). Consequently, when the attention is focused on the outer territory of the VR, the processing of visuospatial cues in VR may have been unarticulated, and in contrast, increased attention to the inner place of the VR induces an unarticulated, inhibited representation of the outer territory of the VR. The target of the attention allocation in VR depends on the intensity of the immersed individual's presence experience (Sanchez-Vives & Slater). Therefore, the attention is shared between inner and outer space in the VR and parallels the mental representation depending on the degree of the presence competing for the attentional resources. The divided attention between the rival representation induces an adjustment in the place 1 3 of attention allocation, which can be found in other conflicting representation paradigms as well where a targeting hand position is rivaled with the postural defined body schema (Botvinick and Cohen 1998). This perceptual inhibition tries to block one of the rival egocentric reference-based representations depending on the individual bias to control inner or outer body-related cues. To control this type of induced postural instability can be realized by automatic compensative maneuvers and trained and aware self-directed actions that can be manifested in a different hand, leg, and trunk movements (Moseley et al. 2012;Wulf 2013).
Our results are also compatible with Bayesian theories of perception and action (Berniker and Kording 2011). The VR environment could increase the uncertainty of the sensory information, which may result in slippage in motor control. The interference is dependent on individual differences (such as the inner/outer divided presence index). The greater perceived uncertainty in the VR environment induces an increased slippage between the motor control of behavior (e.g., drift) and the perceived design of the digital environment. The mismatch between the perceptual and motor components attenuates and the matching augments the experience of reality (Pailard 1991), and plays a role in the regulation of the external or internal attentional allocation when the body and its surrounding-related cues are adjusted (Wulf 2013). With this perception-action congruence (Pailard 1991;Princz 1990), the correct time and space multimodal integration provide the substance of the physical, mental, and software construction for reality (Spence 2011).
In summary, the VR design we constructed yields a competitive effect in the spatial representation of the current environment, which induces personality bias and an attentional focus-dependent compensative postural balance pattern. Understanding the nature of different forms of presence experience and PT skills in the visuospatial construction of VR personally may be beneficial. Conscious training in attention focus and developing a compensatory motor system yielded decreased sensitivity to postural instability, which was advantageous in neurorehabilitation and mental stability (Chander et al. 2019;Franson et al. 2019). The stronger fixation on the physical reality and the associated lower degree of presence restrict the evolution of contralateral balance movements and attenuate the intensity of the conflict between the representations of two egocentric references. The postural system is not static; rather, it is a relatively fluent problem-solving unit with several interactive body parts that respond to conflicting environmental challenges (de Vignemont 2010). The coincidence of the two egocentric spatial references used in VR is a nonconventional representational problem. Considering the data from inner VR and outer VR egocentric-related presence scores, the association analysis showed that these presence values indicated a relatively different state of mind. Our results showed that a higher score in PT ability is associated with an outer VR presence and a lower rate of lateral compensative hand drift. This means that participants intensively focused on the outer VR environment and felt less engaged in the current VR environment. High PT has an essential characteristic; it is a highly conscious attention allocation function, and its focus frequently fluctuates among different spatial positions (Davis 1983;Wallach et al. 2010). Therefore, supporting another finding (van Loon et al. 2018) it may be stated that the degree of the perspective-taking via the presence experience may be an essential agent in reality control when integration of virtual and physical reality is required.

Limitations
This study had certain limitations. This research project revealed numerous new data, but several questions remained open that need additional investigation to explore the effect on the self-coherence and postural instability of the mentioned double reality representation conflict. Considering the human and VR cognitive and behavioral interaction, the forthcoming question is the duration of the fusion of the double representation and its consequences that affect the coherence of the body schema and the stability of the postural system. To avoid data redundancy only a 4 Hz sample taking frequency was applied to measure the lateral head tilting from the vertical to the horizontal direction. To avoid inconvenient bodily sensations a gentle room tilting was applied. The pilot study has no induced motion sickness symptoms so in this study the assessment of motion sickness symptoms was unnecessary.

Conclusions
While a participant focuses on a VR-mediated environment, the control of the possible actions is shared between the affordance of two realities. The representation of the constructed VR and the fixed physical reality is assimilated, and the VR is in part adjusted to the physical environment, while the other part of the surroundings has been neglected. The cost of resolving the representation conflict manifests in a decreased "sense of being there". The presented results highlight the presence of the common, personal, physically real environment where the VR system is embedded. This last sentence is addressed to the VR constructing designers. It is only a courteous suggestion for consideration of the physical environment where the computer-generated reality will be used when installed in the physically real environment. Practicably, the provider should provide practical advice for customers on the application of VR. To call attention to the exact spatial and temporal structure of the surrounding where the VR will be installed, to construct a map with metrical parameters and about the potential obstructions and disturbing agents or persons. Since VR is only one of the different variations of reality. Be aware of where you are!