Exploring Simulation Sickness in Virtual Reality Pedestrian Scenarios: Effects of Gender, Exposure, and User Perceptions

Abstract

Simulation sickness (SS) remains a challenge in virtual reality (VR) applications, especially in pedestrian safety research. This study investigates SS symptoms in VR environments, focusing on gender differences, exposure time, and user perceptions. A total of 145 participants were exposed to two VR pedestrian scenarios: a crosswalk and a sidewalk. The Simulator Sickness Questionnaire (SSQ) was used to assess symptoms of nausea, oculomotor disturbance, and disorientation. Results showed that female participants reported significantly higher SS symptoms than males, with the sidewalk scenario inducing greater overall SS. Additionally, perceived realism in the VR environment was associated with reduced symptoms, while perceived disengagement led to increased discomfort. These findings highlight the importance of user perceptions in mitigating SS and suggest that VR scenarios should be designed with attention to gender differences and environmental realism to improve user experience and safety.

1. Introduction

1.1. Background

Simulation is widely applied across various research fields, including healthcare, manufacturing, education, psychology, and engineering. Simulation-based technologies provide valuable opportunities to investigate human factors within environments that are both safe and well controlled. By replicating complex real-world scenarios, these tools allow researchers to investigate behaviors and decision-making processes without exposing participants to actual risks. This capability is particularly valuable in fields like transportation, where studying interactions in potentially dangerous situations such as conflicts would be difficult to conduct safely in real-world settings.

In traffic safety research, simulation tools such as driving simulators and virtual reality (VR) are increasingly used to study complex interactions among various road users, including pedestrians, cyclists, e-scooter riders, and drivers. Driving simulators provide a valuable means to examine intricate driving behaviors within a safe and controlled setting, particularly when real-world experimentation would be impractical, unsafe, or ethically challenging [1,2]. In addition, driving simulators are also used for training purposes to practice new traffic rules and avoid risky behaviors such as speeding [3]. Similarly, VR serves as an effective tool for assessing how elements of the built environment influence pedestrian safety, by enabling researchers to explore pedestrian behavior, decision making, and interactions within realistic and dynamic traffic scenarios [4].

In addition to research and behavioral assessment, VR has also been widely adopted in safety training applications across various high-risk domains. For example, virtual reality is increasingly utilized in the mining sector to support activities such as complex data visualization, accident reconstruction, safety simulation, risk assessment, and enhanced hazard awareness and training [5]. In the South African mining context, Van Wyk and De Villiers [6] developed non-immersive VR training prototypes that realistically simulate underground environments, enabling miners to practice hazard recognition and remedial actions in a safe and controlled virtual setting. Similarly, NIOSH researchers developed affordable VR software to train U.S. miners in evacuation procedures and hazard awareness using realistic mine simulations [7]. Extending beyond industrial safety, Sharma and Otunba [8] have employed VR in a series of theme-based educational games and evacuation drills, including simulations for aircraft, buildings, subways, university campuses, and entire virtual cities [9,10,11,12], as well as a collaborative VR environment for multi-user emergency response scenarios such as fire, smoke, and active shooter training [13].

Despite their many advantages, simulation-based technologies are not without limitations. One of the most commonly reported issues is SS, a byproduct of modern high-fidelity visual simulation technology [14]. Notably, SS continues to be a significant concern, despite considerable advancements in VR presentation technologies over the past few decades [15]. SS can reduce how comfortable or focused a person feels while using VR. This situation happens when there is a mismatch between what the eyes see and what the body feels. Reason and Brand [16] explained this phenomenon through the Sensory Conflict Theory, which proposes that SS arises when the visual, vestibular, and proprioceptive systems receive mismatched inputs. In the context of driving or pedestrian simulations, such conflicts may occur when the user’s perception in the virtual environment does not align with expectations formed from prior real-world experiences indicating that the deviation in sensory information intake can increase the likelihood of SS. Moreover, personal factors such as vision problems (e.g., myopia, astigmatism), health conditions, and the VR equipment setup can all affect SS symptoms [17,18]. Some studies have identified key causes of SS, such as sensory conflict, visual scaling issues, and balance instability, which can lead to symptoms like nausea, blurry vision, and dizziness [19]. They also observed that inconsistent sensory inputs can increase vection and cause disorientation, making people feel uncomfortable and less engaged during VR past experiences. Lackner [20] emphasized that poor coordination between body senses and vision, lack of vestibular support, and personal differences all play a role in creating discomfort during VR exposure.

Visually induced motion sickness is another form of SS that can occur with little or no physical movement, often resulting in symptoms, ref. [21] suggested that when visual scenes move quickly, the eye muscles can become tense, contributing to discomfort. This tension can affect the vagus nerve, and potentially causes headaches and visual fatigue. Ebenholtz [22] further explained that certain eye movements, specifically optokinetic nystagmus and the vestibular ocular reflex, are involved in the development of SS. These movements help stabilize vision during motion, but when they are strained or disrupted in VR environments, they can cause symptoms like headaches, eye strain, and difficulty concentrating. SS is commonly associated with symptoms such as oculomotor disturbances and a sense of disorientation [23]. In more pronounced cases, individuals may experience nausea, eye discomfort, cold sweats, fatigue, dizziness, and even vomiting [24,25,26]. Moreover, the degree of SS that users feel within the simulated environment can influence their responses, potentially introducing bias into the behavioral data collected [27].

SS is often assessed using subjective tools such as questionnaires. One of the most widely used tools is the SSQ, developed by Kennedy et al. [14], which evaluates three primary symptom categories: nausea, oculomotor disturbance, and disorientation. Nausea includes symptoms such as sweating, trouble concentrating, and stomach discomfort, and is defined as the sensation preceding emesis (vomiting), which can be triggered by factors such as motion sickness, chemotherapy, or post-operative conditions [28,29]. Oculomotor disturbance includes symptoms such as headaches, eye strain, and blurry vision, and is associated with abnormalities in eye movement control, such as asymmetry in the vestibulo-ocular reflex, saccadic function disturbances, and prolonged latencies of eye motion [30]. Disorientation is characterized by sensations such as head fullness, dizziness (with both open and closed eyes), and vertigo [15,23].

1.2. Related Works

Several studies have examined the causes and severity of SS in interactive simulation environments, especially in driving-related applications. One of the most important factors shown to influence SS is user demographics, particularly age, gender, and prior experience with simulated environments, such as VR and driving simulators. Research has consistently shown that females and older adults are more vulnerable to experiencing SS symptoms. For example, Park et al. [31] found that older adults, particularly older female drivers aged 70–90, experienced higher SS compared to younger individuals aged 21–50. Additionally, they reported a higher dropout rate among older drivers (37.3%) compared to younger drivers (13.7%), and with female participants exhibiting more severe symptoms than males, indicating a potential gender-related difference in susceptibility to SS. In another study by Luong et al. [32], who studied 837 participants aged 18 to 80, they also concluded that women were more prone to cybersickness than men. Similarly, Rangelova et al. [33], with a sample of 62 participants, found that gender had a significant effect on SS in VR, with women experiencing more discomfort. Ramaseri-Chandra and Reza [34] reported that demographic factors such as age, gender, and prior VR exposure could predict the likelihood and intensity of simulation sickness. Reddy and Kim [35] also observed that older participants and women experienced more symptoms, based on data from 54 people evenly divided across gender and age groups. Similar findings were reported by Almallah et al. [23], who conducted a driving simulator study involving 132 participants from a multicultural society. Their results indicated that female participants reported higher levels of SS, and that increased age was associated with more severe symptoms.

Virtual speed has also been shown to influence SS. Hughes et al. [36] found that participants experienced more severe symptoms at a simulated speed of 120 mph compared to 60 mph. Another study by Hughes et al. [37] showed that higher speeds increased vection onset, duration, and presence. Similarly, Kemeny et al. [38] noted that while users tolerated low speeds, high acceleration levels caused symptoms in inexperienced participants. The type of display used in VR simulations can also affect the level of discomfort. The two main display types are head-mounted displays (HMDs) and standard screens like LED or TV displays. Guna et al. [39] found that TV displays were linked to lower sickness levels compared to HMDs. This was further confirmed by Rangelova and Andre [40], who noted that discomfort from HMDs remains a challenge despite improvements in VR technology.

Efforts to mitigate SS have highlighted the importance of combining multiple sensory inputs. According to Yeo et al. [41], synchronized sound and motion reduced visually induced motion sickness (VIMS), and higher levels of VIMS were linked to increased alpha and theta brain wave activity. Similarly, Kim et al. [42] reported a reduction in SS symptoms when both motion and visual cues were presented simultaneously, noting that mismatches in pitch angles were correlated with heightened symptom severity. Beyond multisensory inputs, several temporal and behavioral factors have also been identified as critical contributors to SS. For instance, Kennedy et al. [43] observed that long exposure periods made SS symptoms worse, but repeated exposures helped reduce overall sickness over time. In addition. head motion and sensory delays are additional factors that influence. Draper [44] reported that head rotation movements and incomplete simulation of real-world motion contributed to SS due to changes in vestibulo-ocular reflexes. Symptoms were especially noticeable after 30 min of exposure. DiZio and Lackner [45] also demonstrated that delays between head movement and visual update, especially those longer than 40 ms, led to motion sickness and loss of balance. In their study, 28% of participants withdrew after experiencing a 254 ms delay. Jung et al. [46] further noted that first-person viewpoints and horizontal head movements caused more severe symptoms. In some cases, the effects of SS persisted well beyond the simulation period, with Kennedy et al. [47] reporting symptoms lasting up to 12 h post-exposure.

Moreover, the level of realism in simulated environments also plays a critical role in influencing SS severity. Visual-vestibular discrepancies, along with the type of motion cues presented, can significantly affect user response. Kim et al. [42] emphasized that both motion and visual cues reduce simulator sickness compared to visual cues alone, with nausea scores decreasing from 54.59 to 31.27 when motion was included.

1.3. Research Gap

Although numerous studies have investigated SS within driving simulation environments, existing research has also examined SS in virtual reality (VR) settings using head-mounted displays (HMDs), particularly across content types such as 360 videos, games, and minimalist environments, as summarized in recent reviews [48]. However, these studies have not explored SS in relation to pedestrian behavior within realistic traffic contexts. VR delivered through HMDs is now increasingly used to simulate pedestrian interactions in roadway environments for behavioral research and traffic safety applications [49,50,51], yet SS within this context remains unexplored. However, to the best of our knowledge, no studies to date have specifically investigated SS in pedestrian-oriented VR traffic scenarios, such as walking, navigating sidewalks, or crossing streets while wearing HMDs. This knowledge gap becomes even more critical when considering diverse cultural contexts, where road user behaviors and perceptions may vary significantly. The State of Qatar, for instance, presents a unique case due to its highly multicultural road user population, with varying cultural and perceptual backgrounds [52,53,54].

1.4. Research Objectives and Questions

Building on the identified research gap, this study aims to examine the factors influencing SS in pedestrian-oriented VR scenarios using HMDs. The focus is on how simulation conditions, exposure, demographic characteristics, and user perceptions influence the severity of SS symptoms. Specifically, the study seeks to answer the following research questions:

RQ1:

How does exposure to different VR scenarios, specifically road-crossing and sidewalk walking, affect the severity of simulator sickness (SS) symptoms? These two scenarios were selected because research on pedestrian behavior often focuses either on road-crossing actions or interactions among pedestrians and other sidewalk users.

RQ2:

How do individual demographic characteristics influence the severity of SS symptoms in pedestrian VR environments?

RQ3:

To what extent do user perceptions, including perceived realism and disengagement, contribute to the severity of SS symptoms in pedestrian VR environments?

RQ4:

Are there significant relationships among the SSQ subscales in measuring SS symptoms?

Figure 1 illustrates the key focus areas of the study, highlighting the multidimensional approach taken to investigate SS in VR environments.

2. Materials and Methods

2.1. VR Simulation

A VR system developed at the simulation Lab of the Qatar Transportation and Traffic Safety Center (QTTSC) was used to carry out the experiment (Figure 2). The system operates using the HTC VIVE Pro (HTC Corporation: Taoyuan City, Taiwan) and the updated SteamVR Tracking 2.0 technology, which allows up to 16 base stations to be connected to form a single tracking area. For instance, connecting 8 base stations can cover an area of up to 20 m². The system tracked the position of each participant’s head using sensors, and this information was updated on the display in real time. As a result, participants experienced the feeling of walking through a virtual environment that included other users of shared spaces, such as pedestrians, cyclists, scooter riders, and vehicles.

2.2. Simulated Environment

The study involved two different VR pedestrian scenarios representing different experimental conditions (details shown in Figure 3). The first scenario simulated a crosswalk environment, where pedestrians were required to cross a two-way two-lane road, using a zebra crossing while observing approaching vehicles. Before crossing, the pedestrian stood at the edge of the crosswalk and turned their head to the right and left to observe approaching vehicles and make a crossing decision. The road had a posted speed limit of 50 km/h, and the lane width was 3.65 m, based on the standard specified in the Qatar Traffic Control Manual (QTCM, 2021). This road was designed to replicate a typical residential street in Qatar, and all the vehicles used in the simulation were sedan-type cars. The second scenario represented a sidewalk condition where pedestrians walked along a path parallel to the road and encountered approaching e-scooters and cyclists (See Figure 4). The sidewalk was 2.5 m wide and approximately 13 m long, designed in line with the guidelines from Qatar’s Ministry of Municipality and Urban Planning (MMUP). The virtual setting recreated a residential street sidewalk commonly found in Qatar. E-scooters approached from three different directions: directly from the center, from the participant’s left (roadside), and from the participant’s right (wall side). The e-scooters moved at constant speeds of either 10 km/h or 30 km/h throughout the simulation. This setup was intended to assess the participant’s response based on the position and speed of the approaching e-scooter in a controlled virtual environment. The environment and road features were created using Unity software (Unity 2021.3.3) to reflect realistic road and traffic conditions.

The sidewalk and crosswalk scenarios were chosen because they represent two different types of pedestrian activity commonly examined in traffic safety research: walking alongside traffic with potential micro-mobility interaction and actively crossing a road while monitoring vehicle movement. The crosswalk scenario was shorter in distance and required participants to check for approaching vehicles. Crossing in front of moving vehicles tends to demand more attention and induce a higher level of user involvement/engagement. In contrast, the sidewalk scenario involved a longer walking distance with continuous visual flow and interactions with oncoming e-scooters and cyclists. These elements introduced visual-vestibular conflicts that are known contributors to SS. While VR delivered through HMDs has been shown to cause SS in general, its effects have not been specifically examined in these pedestrian contexts. Understanding how SS may emerge in such scenarios is important for interpreting behavioral data accurately and for guiding the design of future VR-based traffic studies.

It is worth noting that participants walked approximately 9–10 m in the crosswalk scenario and around 13–14 m in the sidewalk scenario. In both scenarios, each participant completed 30 walks.

2.3. Simulation Sickness Questionnaire

In this study, SS was assessed subjectively using the Simulator Sickness Questionnaire (SSQ) developed by Kennedy et al. [14]. Although originally designed for aviation simulators, the SSQ has since been validated across a range of simulation environments. Notably, Balk et al. [55] confirmed its three-factor structure nausea, oculomotor disturbance, and disorientation using data from nine driving simulator studies, supporting its robustness in land-based contexts. In addition, Uğur et al. [56] validated the Turkish-translated version of the SSQ in a VR-HMD setting during a roller coaster simulation, reinforcing its applicability to immersive virtual environments.

The SSQ consists of 16 items, with each of the three subscales (nausea, oculomotor disturbance, and disorientation) containing 7 items. Some items are categorized in more than one subscale. After completing the VR experiment, participants rated the severity of each symptom using a four-point scale (i.e., 0 represents no symptom, and 3 represents the highest severity of the symptom). This standardized approach helps quantify SS symptom intensity within the context of the simulation, without serving a clinical diagnostic purpose [57].

2.4. Procedure

The experiment received ethical approval from the Institutional Review Board (QU-IRB) at Qatar University (QU-IRB 1915-EA/23). When participants arrived at the VR laboratory, they were welcomed and asked to sign an informed consent form. They were informed that they could withdraw from the study at any time without providing a reason. Before beginning the simulation, each participant completed a pre-questionnaire using the Qualtrics online platform, which collected basic demographic information (e.g., age, gender, education level, and nationality). Eye calibration was then conducted for each participant before starting the practice and simulation walks. Participants were given 10 min for a practice walk to become familiar with the virtual environment and the VR headset. Afterward, each participant completed six simulation walks in random order. They were encouraged to walk as they normally would in real-life situations. Following the VR experiment, participants were asked to complete the SSQ and a post-questionnaire (PQ) using Qualtrics. Both questionnaires took approximately five minutes each to complete.

2.5. Participants

A total of 145 subjects voluntarily participated in the experiment. Recruitment was conducted through direct invitations extended to Qatar University students and staff, as well as through social media channels to reach individuals outside the university. Interested participants scheduled their sessions through an online registration platform hosted by Qatar University (http://qds.qu.edu.qa/ accessed on 11 June 2025). Of these, 70 participated in the crosswalk scenario, while 75 took part in the sidewalk scenario, which involved encounters with a cyclist and an approaching e-scooter. Adhering to the guidelines established by Kennedy et al. [14] for the SSQ, participants were instructed to abstain from consuming food and beverages (with the exception of water) for a minimum of two hours before participating in the experimental session. Since no participants exhibited signs of SS during the experiment, all were observed to be unaffected and were asked to complete the post-experiment questionnaire and SSQ. The demographic characteristics of 145 participants are summarized in

Table 1. Descriptive of demographic.

Variable	Level	Sample Sidewalk	Sample Crosswalk	Total Sample
Gender	Female	30.67%	27.14%	28.97%
	Male	69.33%	72.86%	71.03%
Ethnicity	Non-Arab	66.70%	64.30%	65.52%
	Arab	33.30%	35.70%	34.48%
Educational level	Secondary education	56.00%	2.90%	30.34%
	High school	4.00%	55.70%	28.97%
	Diploma	28.00%	1.40%	15.17%
	Bachelor’s degree	12.00%	28.60%	20.00%
	Master’s or doctorate	0.00%	11.40%	5.52%
Age	Minimum	17.00	17.00	17.00
	Maximum	54.00	54.00	54.00
	Mean	23.60	24.76	24.01

. The mean age of the total sample was 24.01. While the sample was male dominated, consisting of 71.03% males and 28.97% females, this reflects the broader demographic distribution in Qatar, where males make up approximately 71.5% of the total population [58]. In terms of ethnicity, 65.52% of the participants were non-Arab, while 34.48% were of Arab origin, representing over 20 different nationalities. The participants’ educational levels were distributed as follows: 30.34% had a secondary-level education, 28.97% had a high school education, 15.17% held a diploma, 20.00% had a bachelor’s degree, and only 5.52% had a master’s or doctoral degree.

2.6. Analysis

Statistical analyses were conducted using two software sources such as SPSS version 27.0 and Smart-PLS software (version 4.1.0.9). Smart-PLS was employed for structural equation modeling [59]. An independent t-test was conducted to examine the symptoms of the SS subscale and the total score across genders. Pearson correlation was conducted to examine the relationships among the subscale of SS. Additionally, a structural equation model (SEM), as shown in Figure 5, was utilized to examine the impact of demographic characteristics and conditions on five perception-related questions concerning participants’ experiences, safety, realism, and behavior in the VR environment. These perception-related questions served as mediators, while the SS subscale score and total scores functioned as the dependent (endogenous) variables. SEM provides a theoretical approach for investigating complex systems with multiple variables and hypothesized causal pathways [60]. SEM evaluates both the direct causal associations between demographic factors and the mediating variables, as well as the relationships between these mediating variables and the SS subscale and total scores. Additionally, it identifies the indirect influence of demographic factors on SS outcomes through the mediation of the five perception-related questions. For the assessment of both direct and indirect effects, the bootstrapping approach was employed. PLS-SEM, as a variance-based method, is particularly suitable for analyzing data that do not follow a normal distribution, as it operates independently of normality assumptions [61].

For SEM, the demographic variables were coded as follows: Age was treated as a continuous variable, while gender was categorized as 0 = “Female” and 1 = “Male”. Education level was classified as follows: 1 = “Secondary education,” 2 = “High school,” 3 = “Diploma,” 4 = “bachelor’s degree,” and 5 = “Master’s or Doctoral degree”. Ethnicity was coded as 0 = “non-Arab” and 1 = “Arab”. Additionally, the “Condition” variable represented the experiment type, with 0 indicating the sidewalk experiment and 1 indicating the crosswalk experiment.

3. Results

3.1. Descriptive Statistics of SS

The bar chart in Figure 6 depicts the comparative mean scores for nausea, oculo-motor disturbance, disorientation, and the total simulation sickness score under the sidewalk and crosswalk conditions. The results indicated that participants consistently reported higher mean scores in all categories when in the sidewalk condition compared to the crosswalk condition. A detailed examination of the distributions reveals several key patterns. Nausea symptoms showed a relative difference between conditions with sidewalk participants reporting 32% higher mean scores (M = 11.70) compared to crosswalk participants (M = 8.86). Specifically, oculomotor disturbance exhibited the largest difference, with a mean score of approximately 16.6 for the sidewalk condition and 14.3 for the crosswalk condition, with an 11% increase in the sidewalk condition. Similarly, the total simulation sickness score was higher in the sidewalk condition (M = 17.1) compared to the crosswalk condition (M = 14.4). These findings suggest that the sidewalk experience induces greater symptoms, with oculomotor disturbance and overall simulation sickness being the most affected.

3.2. Comparison of Simulation Sickness Symptoms Between Males and Females

To compare SS symptoms between males and females, an independent t-test was conducted after removing outliers. The results indicated that female participants experienced greater symptoms of simulation sickness compared to males, particularly in oculomotor disturbance, disorientation, and the total simulation sickness score. The analysis revealed that females reported significantly higher scores in oculomotor disturbance (M = 19.49, SD = 25.45) compared to males (M = 12.38, SD = 14.00; p = 0.034). Similarly, they also exhibited significantly greater disorientation (p = 0.014) and total SS score (p = 0.021) than males. These findings suggest that females experienced greater levels of disorientation and overall simulation sickness compared to males.

Similarly, nausea scores were higher in females (M = 12.94, SD = 20.29) than in males (M = 8.97, SD = 10.86); this difference was not statistically significant (p = 0.131), indicating that gender differences in nausea were less pronounced. The box plots shown in Figure 7 visually reinforce these findings, showing that females consistently exhibited higher median scores and greater variability in simulation sickness symptoms across all subscales compared to males. The box plots in Figure 7 illustrate the distribution characteristics underlying these statistical differences. Across all SS subscales, females consistently show higher median values and notably wider interquartile ranges compared to males. The female distributions exhibit more extensive upper whiskers, particularly evident in oculomotor disturbance where several female participants reported scores exceeding 50 points. In contrast, male participants demonstrate more compact distributions with fewer extreme values, suggesting less variability in SS responses within this group.

3.3. Structure Equation Model

Table 2. Path analysis results for significant effects.

Path	β	STDEV	T Value	p Values
Condition -> Past experience	−0.381	0.079	4.850	<0.001
Condition -> Perceived disengagement	−0.379	0.182	2.079	0.038
Educational Level -> Perceived disengagement	−0.220	0.103	2.138	0.033
Gender -> Vehicle speed perception	−0.384	0.170	2.255	0.024
Gender -> Perceived realism	−0.405	0.181	2.242	0.025
Perceived disengagement -> Nausea	0.158	0.074	2.147	0.032
Perceived disengagement -> Oculomotor disturbance	0.182	0.087	2.087	0.037
Perceived disengagement -> Disorientation	0.218	0.073	2.971	0.003
Perceived disengagement -> Total scores	0.208	0.078	2.665	0.008
Perceived Realism -> Nausea	−0.210	0.089	2.361	0.018
Perceived Realism -> Oculomotor disturbance	−0.221	0.109	2.032	0.042
Perceived Realism -> Total scores	−0.226	0.105	2.163	0.031

presents the SEM estimation results for only significant relationships. It emphasizes the key impact of demographic factors and experimental conditions on perception-related mediators, as well as the influence of these mediators on simulation sickness scores. The results indicate that experiment conditions had a significant negative association with VR past experience (β = −0.381, p < 0.001) and perceived disengagement (β = −0.379, p = 0.038). This suggests that participants in the sidewalk condition were more likely to report having prior VR past experience (VR past experience coded as 1 = Yes, 0 = No) and perceived themselves as more disengaged in the VR environment compared to real-world situations in the crosswalk condition.

Regarding demographic factors, educational level negatively impacted perceived disengagement (β = −0.220, p = 0.033), indicating that participants with higher education levels reported lower perceived disengagement in the VR environment. Gender also played a significant role, influencing vehicle speed perception (β = −0.384, p = 0.024) and perceived realism (β = −0.405, p = 0.025), suggesting that perceptions of vehicle speed and movement and VR environmental realism varied by gender. Although males reported higher ratings on VR vehicle speed and movement perception question, they also exhibited greater disengagement in the VR environment compared to females. This suggests a potential difference in risk perception or behavioral adaptation between genders in virtual settings.

Regarding the direct impact of perception-related questions on simulation sickness scores, perceived disengagement revealed significant associations with simulation sickness symptoms. Higher perceived disengagement in the VR environment was linked to increased nausea (β = 0.158, p = 0.032), oculomotor disturbance (β = 0.182, p = 0.037), disorientation (β = 0.218, p = 0.003), and the total SS score (β = 0.208, p = 0.008), indicating that participants who perceived themselves as more disengaged in VR were more likely to experience symptoms of SS. In contrast, perceived realism negatively influenced nausea (β = −0.210, p = 0.018), oculomotor disturbance (β = −0.221, p = 0.042), and the total SS score (β = −0.226, p = 0.031), suggesting that greater perceived realism of the VR environment was associated with reduced SS symptoms. These findings highlight the role of perceived behavioral adaptation and environmental realism in shaping the intensity of simulation sickness effects in VR settings.

Overall, the SEM results demonstrate distinct pathways of influence, with certain demographic and experimental factors showing significant associations with specific perception-related factors. Education level was significantly associated with perceived disengagement only, while gender influenced both vehicle speed perception and perceived realism. Experimental conditions showed influence on past experience and perceived disengagement. Among these mediators, perceived disengagement and perceived realism significantly influenced SS symptoms. However, no significant indirect impacts of demographic or experimental conditions on SS symptoms were found.

3.4. Correlations Among Simulation Sickness Subscales

Table 3. Pearson correlations among simulation sickness subscales and total score.

SS Subscale		Nausea	Oculomotor	Disorientation
Nausea	Pearson correlation	1	0.630 **	0.644 **
	Sig. (2-tailed)	-	<0.001	<0.001
Oculomotor disturbance	Pearson correlation	0.630 **	1	0.833 **
	Sig. (2-tailed)	<0.001	-	<0.001
Disorientation	Pearson correlation	0.644 **	0.833 **	1
	Sig. (2-tailed)	<0.001	<0.001	-

** Correlation is significant at the 0.01 level (2-tailed).

shows the relationship between subscale of SS. The results revealed statistically significant and strong positive correlations between all subscales. Nausea was strongly correlated with oculomotor disturbance (r = 0.630, p < 0.001) and disorientation (r = 0.644, p < 0.001). The strongest association among the subscales was found between oculomotor disturbance and disorientation (r = 0.833, p < 0.001), indicating a high degree of co-occurrence between these two symptoms. These findings suggest that while the subscales represent distinct symptom dimensions, they are closely interrelated components of the overall simulation sickness experience.

4. Discussion

This study aimed to investigate SS experienced by pedestrians within an immersive VR environment using HMD. The research focused on identifying factors influencing SS subscales and overall SS during the simulation, comparing SS symptom manifestations between genders, and exploring interrelationships among SS symptom subscales. Participants engaged with two distinct VR scenarios designed to simulate real-world pedestrian experiences: a crosswalk scene involving vehicular interactions and a sidewalk scenario featuring approaching e-scooters and cyclists. After completing each experiment, self-reported data related to SS were collected using self-reported questionnaires using the SSQ developed by Kennedy et al. and Schubert et al. [14,62]. This questionnaire enabled a systematic and quantitative evaluation of participants’ physiological and perceptual responses to the VR experiences.

The results showed that SS symptoms were higher across subscales and overall SS in the sidewalk scenario, with the most notable differences observed in oculomotor disturbance and the total SS score. The sidewalk scenario involved continuous forward motion and the presence of approaching e-scooters and cyclists, which may have increased the intensity of visual stimuli. Additionally, participants had to walk for a slightly longer duration in this scenario compared to the crosswalk scenario. The setup of the sidewalk scenario likely introduced a stronger mismatch between visual and vestibular cues, commonly associated with SS. On the other hand, the crosswalk scenario required more stationary observation before crossing, which may have reduced sensory conflict. This outcome suggests that environmental complexity and exposure type significantly influence SS in VR pedestrian simulations [43]. Building on this observation, design specifications such as sidewalk width, movement direction, and scene layout likely contributed to the severity of SS symptoms. The narrower sidewalk compared to the crosswalk may have increased spatial confinement and reduced peripheral vision, particularly where a wall bordered the right side. These constraints, combined with continuous forward motion, may have intensified sensory mismatch and discomfort, leading to stronger oculomotor and disorientation symptoms. Such design-related aspects offer valuable guidance to be considered when developing future VR scenarios, particularly in configuring pedestrian environments like sidewalks or crosswalks in order to minimize simulator sickness symptoms during behavior-focused studies. In addition to environmental complexity, physical aspects of the VR hardware such as the weight and fit of the headset may also influence participant comfort and potentially amplify SS symptoms. The headset used in this study weighed approximately 800 g with the head strap, and such weight may contribute to increased neck muscle strain and fatigue during extended usage. Prior research has shown that heavier or poorly balanced headsets increase neck muscle activity and visual-vestibular strain, which can exacerbate SS responses in immersive simulations [63,64].

Gender differences also emerged as an important factor influencing SS symptoms. Female participants consistently reported higher SS scores than male participants, particularly in the subscales of disorientation and oculomotor disturbance, as well as the overall SS. These findings support earlier research showing a higher incidence of SS symptoms among females in both driving simulators and VR-based systems [23,31,32,33]. The higher SS symptoms observed in female participants may be linked to hormonal influences. Bannigan et al. [65] found increased physiological arousal during ovulation in naturally menstruating females, suggesting that hormonal fluctuations could contribute to greater sensitivity in VR environments.

Regarding the direct impact of demographic factors and experimental conditions on perception-related measures assessing VR past experience, involvement, and realism, the results showed that participants in the sidewalk condition were more likely to report prior VR past experience. This may partly explain why they also reported higher levels of perceived disengagement in the sidewalk scenario compared to the crosswalk scenario. Educational level was also found to negatively correlate with perceived disengagement, indicating that participants with higher education may be more attentive or cautious in virtual environments. Greater educational attainment has been associated with higher task completion and retention rates, indicating increased engagement, willingness, or commitment to the assigned activities [66]. Males were more likely to agree with the statement, “I think the speed and movement of approaching vehicles are realistic,” indicating a higher perception of speed realism compared to females. However, despite perceiving the vehicle movement as more realistic, males also reported higher levels of perceived disengagement in the VR environment. This contrast may reflect differences in risk perception or behavioral adaptation strategies between genders, where males may exhibit more confidence or desensitization to virtual stimuli, potentially leading to less cautious interactions. Conversely, females often report lower levels of comfort, confidence, and perceived capability particularly in highly technical and visually complex virtual environments [67]. These findings underscore the importance of considering gender-based perceptual and behavioral differences when designing and evaluating virtual environments for realism and safety-related assessments.

Regarding the direct impact of VR past experience, realism, and movement perception in the virtual environment, the results emphasize the importance of user experience, perception, and behavior in the development of SS symptoms. The findings indicated that participants who perceived themselves as more disengaged in the virtual environment were significantly associated with increased symptoms across all SS subscales, including nausea, oculomotor disturbance, disorientation, and overall SS scores. These findings suggest that perceived disengagement, characterized by behaviors such as acting less attentively or seriously in the virtual environment compared to real life, may contribute to increased sensory conflict or reduced situational awareness. Furthermore, a notable finding in this study was the impact of perceived realism on SS outcomes. The results clearly showed that greater perceived realism in the VR environment was associated with significantly lower scores for nausea, oculomotor disturbance, and the overall SSQ score. While some studies have shown that highly realistic VR past experiences may intensify symptoms over short durations due to increased sensory conflict [68,69], our findings suggest that in low-motion, pedestrian-oriented environments, higher perceived realism may reduce symptoms by improving immersion and perceptual coherence. The scenarios in the pedestrian VR environment should be designed to be highly engaging while minimizing SS symptoms. Moreover, the strong correlation among the SS subscales indicates that an increase in symptom severity in one subscale is likely to be accompanied by increased severity in the other two subscales [23]. The analysis revealed no significant indirect effects from demographic factors and experimental conditions on SS. This indicates that while perception appears to be a critical factor, the influence of demographic variables may be more directly manifested, as evidenced by the observed gender-related impact. The implications of these findings are important for researchers and developers working with pedestrian VR simulations. The significant effects of realism and user engagement suggest that improving environmental fidelity and ensuring participants are well oriented and focused during simulations can help reduce SS symptoms. Similarly, understanding how different demographic groups perceive and react to VR environments can help tailor designs to diverse user needs. For example, providing additional onboarding or adaptation time for females or older users may help mitigate discomfort.

Limitations and Future Research

It is important to acknowledge certain limitations of this study. First, although VR is a useful tool for controlled testing, the weight of the headset may have impacted participant comfort and reduced the perceived realism of the simulation.

One limitation of the study is the skewed age distribution of participants, with a predominance of young adults. However, this reflects broader population trends in Qatar, where approximately 73% of the population is under the age of 39, according to the National Planning Council [58]. Additionally, since the study was conducted at Qatar University, most participants were university students who received an invitation via email to take part in the experiment. Younger individuals are generally more familiar with technology, which may have influenced their ability to adapt to the VR environment more easily compared to an older population. Moreover, the gender distribution in this study was imbalanced, with 69% male and 31% female participants. This discrepancy aligns with the broader demographic structure of Qatar, where 28.5% of the population is female and 71.5% is male, as reported by the National Planning Council [58]. While this demographic profile reflects the broader population, the imbalance in gender and age may limit the generalizability of the findings, particularly in analyses involving gender-based comparisons. Future studies should recruit more balanced and diverse samples, including adequate representation of female and older adult participants, to enhance the generalizability of findings and to allow more robust subgroup analyses.

In this study, SS was the primary outcome measure. Although we incorporated SSQ subscales and user perception variables to provide additional insight, future studies should consider including behavioral and physiological outcome measures to strengthen the findings and ensure a more comprehensive evaluation of VR-based pedestrian experiences. For instance, incorporating objective metrics such as vital signs including heart rate, skin conductance, or postural stability may offer deeper insight into participants’ physiological responses. Behavioral metrics such as crossing reaction times or decision-making accuracy could also help reveal the performance impacts of SS. Additionally, adopting a design of experiments framework could enable systematic variation of key design parameters such as sidewalk width, motion speed, or exposure time to better understand their individual and combined effects on SS and user behavior. While our findings emphasize the importance of SS mitigation for enhancing comfort and engagement in VR pedestrian environments, the current study does not establish a direct empirical link between reduced SS symptoms and improvements in real-world pedestrian decision making or safety outcomes. Future research should explore this relationship by integrating behavioral performance metrics such as hazard anticipation, risk avoidance, or reaction time into the study design to determine whether reducing SS leads to measurable improvements in pedestrian safety performance.

Although the SSQ provides a validated framework for assessing SS in non-clinical research, clinical interpretation of subscales such as oculomotor disturbance and disorientation was not guided by consultation with optometrists or other medical professionals. Future research may benefit from interdisciplinary collaboration to strengthen clinical insight. Additionally, examining how various VR device specifications influence simulation sickness may offer valuable insights for optimizing VR-based pedestrian research. Future studies may also explore the effects of individual scenario parameters, such as exposure duration and environmental complexity, by isolating them through controlled scenario design.

5. Conclusions

This study investigated the impact of demographic and perceptual factors on SS experienced by pedestrians within an immersive VR environment using HMD. The results revealed that gender differences influenced SS symptoms, with female participants reporting higher levels of disorientation, oculomotor disturbance, and overall SS. This highlights the importance of designing VR environments that accommodate diverse sensory responses and promote comfort and accessibility for all users. Moreover, participants exposed to the sidewalk condition featuring dynamic interactions with e-scooters and cyclists reported higher SS symptoms compared to those in the crosswalk condition. Strong positive correlations were observed among SS subscales, indicating their interrelated nature. Perceptual factors also played a critical role: higher perceived disengagement in the VR environment was associated with more severe SS symptoms, while greater perceived realism was linked to reduced SS.

These findings reveal the need for VR design strategies in road safety research, particularly for exploring pedestrian behavior, that are sensitive to individual differences, including gender-based variations in SS susceptibility. Given that SS symptoms were more pronounced in the sidewalk scenario, where participants were exposed to longer durations, future VR scenarios should consider shorter exposure times and include brief breaks to reduce sensory overload. Furthermore, the significant role of perceived realism and disengagement highlights the importance of visual and behavioral fidelity in enhancing participant engagement while minimizing the adverse effects of SS severity.