From Atmospheric Tension to Embodied Regulation: A Mixed-Methods Study of Fear in VR and Non-VR Survival Horror Gameplay

1. Introduction

Virtual reality has expanded the cultural and commercial reach of survival horror by producing experiences in which threat feels spatially proximate, continuously surrounding, and difficult to disengage from. For players, this intensification is not confined to heightened sensory stimulation. VR experiences may influence how players engage with immersive environments beyond immediate gameplay. Previous research indicates that VR exposure may have short-term emotional consequences beyond immediate in-game reactions (Lin, 2017; Lavoie et al., 2021; Souchet et al., 2023). Empirical research on immersive media has likewise shown that highly immersive environments can elicit not only stronger presence but also more demanding affective and physiological responses, particularly when users confront aversive or threatening content (Cummings & Bailenson, 2016; Lemmens et al., 2022). These dynamics indicate that immersive fear is not exhausted at the point of exposure, but involves ongoing regulation across bodily engagement during play and interpretive adjustment after the experience ends. As VR horror becomes a mainstream entertainment form and a reference model for immersive experience design, understanding fear therefore requires attention to how it is enacted, managed, and stabilized over time.

Existing scholarship has made substantial progress in demonstrating that VR tends to enhance presence and emotional engagement compared to less immersive displays, with especially robust effects observed in interactive contexts (Caroux, 2023). However, much of this literature still approaches fear primarily as an outcome variable, emphasizing overall amplification or isolating specific features associated with stronger emotional response. Such an orientation leaves several analytically important questions under-specified for survival horror as an aesthetic and experiential form. Fear in this genre is not uniform: combat-driven escalation, slow psychological tension, and spatial disorientation are organized through distinct configurations of lighting, sound, spatial layout, and interaction demands, each of which may invite different forms of bodily engagement and regulation. Recent research has begun to show that design choices, narrative framing, and mode of interaction shape presence and affect in different ways, suggesting that immersive experience is better understood as a structured process than as a simple increase in intensity (Diemer et al., 2015; Steinhaeusser et al., 2023). In addition, player background is frequently reduced to a control variable, despite growing evidence that prior gaming experience, VR familiarity, and narrative grounding systematically influence how virtual environments are interpreted, how presence is stabilized, and how cybersickness or overload is managed (Lin et al., 2018; Weech et al., 2020). Finally, research relying predominantly on self-report measures risks overlooking the interpretive work through which players retrospectively align their narratives of control, vulnerability, or enjoyment with what their bodies enacted under immersive threat. This highlights the importance of examining immersive fear as a dynamic process rather than as a single outcome.

To address these issues, this study reconceptualizes immersive fear in survival horror as a process shaped by atmospheric configuration, embodied regulation, and post-play interpretation, rather than as a single emotional outcome. Empirically, it adopts a sequential mixed-methods design centered on Resident Evil Village. In Study 1, scene-based observation, synchronized gameplay recordings, and post-play interviews are used to examine how fear is enacted and negotiated across three contrasting atmospheric segments characterized by combat pressure, psychological ambiguity, and spatial disorientation. This phase examines embodied regulation and the relation between behavior and retrospective narration. Building on these insights, Study 2 employs a within-subject comparative experiment in which 30 participants experience both VR and non-VR versions of the game in randomized order. By comparing emotional ratings, scene-specific fear, immersion, and motion sickness across display modes, the second phase assesses whether VR not only intensifies fear but also reconfigures the experiential profile of survival horror gameplay. Taken together, the two-stage design allows the study to move beyond generalized claims about VR amplification and toward a more process-oriented account of how immersive fear is differentially organized, interpreted, and sustained across modes of play.

2. Literature Review

2.1. Atmospheric Configuration and Embodied Regulation

Game atmosphere is commonly understood as an experiential configuration through which audiovisual form, spatial composition, and interaction jointly shape how a game world is apprehended and acted upon, rather than as a neutral backdrop for gameplay. In survival horror, atmosphere functions as a structured arrangement of threat: visibility is restricted, auditory cues extend perception toward what remains unseen, and spatial layouts organize vulnerability by alternating exposure, confinement, and navigational uncertainty. From this perspective, atmosphere matters not only as representation but as a practical condition of conduct. More broadly, presence research has long suggested that mediated environments become experientially consequential when users no longer merely observe them but respond as if they were situated within them (Lombard & Ditton, 1997; Slater & Sanchez-Vives, 2016). Recent reviews further show that game design choices systematically shape presence and engagement, reinforcing the view that atmosphere operates as a multi-factor experiential system rather than as the isolated effect of any single sensory cue (Cummings & Bailenson, 2016; Freeman et al., 2018).

Virtual reality intensifies the coupling between environment and conduct by binding perception and action through embodied viewpoints, head tracking, and sensorimotor contingency. When place illusion and plausibility are sufficiently sustained, users tend to respond to virtual environments in ways that resemble everyday threat engagement (Slater, 2009). Within game research, immersive display conditions have repeatedly been associated with higher presence, stronger negative affect, and greater experiential intensity than non-immersive alternatives, although these effects vary according to content, interaction design, and prior player experience (Lemmens et al., 2022; Bouchard et al., 2017). Prior studies suggest that fear experiences vary according to players’ coping strategies and sense of self-efficacy (Bailenson et al., 2018; Kothgassner et al., 2019; Rosenberg et al., 2019). These findings suggest that immersive fear is best conceptualized not as a direct emotional output of technological intensity, but as an emergent outcome of how atmospheric structures recruit and constrain embodied regulation.

Multisensory and design-oriented accounts provide an additional explanatory layer for understanding this process. Presence in games is sensitive to a constellation of factors spanning audiovisual form, interaction mode, narrative framing, and the relationship between players’ expectations and the environment’s demands (Freeman et al., 2017; Slater et al., 2020). Yet even where immersive experience is treated in a more differentiated manner, fear is still often modeled as an aggregated emotional outcome rather than as a scene-specific pattern of bodily adjustment. As a result, there remain limited accounts of how distinct atmospheric logics within survival horror—such as combat pressure, psychological ambiguity, or spatial disorientation—differentially recruit hesitation, avoidance, scanning, or tactical advancement during play. Addressing this limitation motivates an atmosphere-typological approach that compares contrasting scene configurations to trace how fear is regulated through situated bodily action.

2.2. Experiential Mediation

Immersive fear in VR survival horror is shaped not only by atmospheric design but also by players’ experiential repertoires. Prior familiarity with VR conventions, genre schemas, and gaming interfaces conditions how environmental cues are attended to, interpreted, and acted upon during play. Rather than unfolding as a linear stimulus–response sequence, immersive gameplay is increasingly conceptualized as a dynamic process in which cognitive appraisal, affective fluctuation, and sensorimotor engagement jointly organize moment-to-moment involvement (Kuppens & Verduyn, 2017; Wan & Chiu, 2023). From this process-oriented perspective, identical atmospheric configurations may be encountered as manageable suspense by some players and as destabilizing threat by others, depending on how prior experience supports situational inference, predictive control, and attentional anchoring under uncertainty.

Research on presence further indicates that experiential factors shape not only the intensity of immersion but also its consequences. Although immersive systems often increase involvement, that involvement is not uniformly beneficial. Presence can coexist with cognitive workload, motion-related discomfort, perceptual strain, and reduced control, especially when users lack stable interpretive expectations or interactional familiarity (Weech et al., 2020; Yildirim, 2020; Liu et al., 2022). In such cases, elevated fear may partly reflect interactional overload rather than atmospheric effectiveness alone. Comparative studies therefore suggest that presence should not be treated as a straightforward proxy for better experience, but as a condition whose meaning depends on how users recruit prior repertoires to stabilize action and interpretation (Reer et al., 2022).

Beyond technical familiarity, genre experience and affective style further mediate how fear is interpreted through learned coping scripts and patterned expectations. Lin et al. (2018) show that self-efficacy in managing frightening content is central to the enjoyment of VR horror, while Weech et al. (2020) demonstrate that gaming experience interacts with narrative framing to shape both presence and cybersickness. This suggests that experience operates not merely as resistance to intensity but as a form of interpretive competence. Fear may thus be framed as tactical challenge, aesthetic thrill, or overwhelming exposure depending on how players situate atmospheric cues within established repertoires of action and meaning. Existing scholarship strongly implies such mediation, but VR horror research has rarely specified how experiential repertoires become visible when embodied behavioral traces are examined alongside post-play accounts.

2.3. Behavioral-Narrative Discrepancy and Residual Effects

Immersive fear in VR survival horror cannot be adequately captured through either observable behavior or retrospective narration alone because the linkage between real-time regulation and subsequent reports is systematically reconstructive rather than descriptive. Autobiographical memory research shows that retrospective reports are shaped by mechanisms of contextual reconstruction and integration that transform how original experiences are later described (Kensinger & Ford, 2020). Studies in cognitive affective neuroscience further demonstrate that negative and positive episodic memories are differentially reconstructed based on emotional salience, with remembered content disproportionately influenced by peak affect and contextual cues rather than the full experiential timeline (Williams, Ford, & Kensinger, 2022). In immersive gameplay, where bodily responses unfold under complex environmental pressures, this divergence means that the embodied trajectory of fear regulation may not be transparent in later verbal narratives.

Rather than treating such discrepancy as a methodological nuisance, it can be interpreted as evidence that fear unfolds across distinct temporal registers. One register is immediate and embodied, where players retreat, pause, scan, hesitate, or advance under threat constraints; another registers in retrospective memory, where these same actions are reorganized into coherent narratives of competence, tactic, or genre literacy. Research in memory and experience shows that remembered experiences are not simple reproductions of moment-to-moment states; rather, remembered emotional valence is often influenced by the structure of the experience and the prominence of salient momentary features (Smith & Mulligan, 2021). For VR survival horror, this suggests that micro-level vigilance and low-grade tension evident in behavioral sequences may be underrepresented in verbal reports dominated by narrative landmarks such as jump scares or climactic peaks.

Independent lines of work investigating memory for virtual experiences indicate that immersive environments shape both encoding and retention such that emotional intensity and presence contribute uniquely to how episodic elements are later recalled (Diemer, Alpers, & Peperkorn, 2015). Additionally, systematic reviews of VR and emotion confirm that immersive emotional experiences influence post-experience memory, attentional orientation, and affective disposition long after the formal session has ended (Riva et al., 2023). In the context of VR horror, lingering unease, intrusive recollections, or reluctance to re-enter similar environments should therefore not be treated as peripheral side effects but as part of the broader process through which embodied impressions are integrated into subsequent perceptual expectations and mnemonic frameworks.

Taken together, these insights support conceptualizing immersive fear as a process distributed across action, interpretation, and temporal dynamics. Fear is enacted through bodily regulation under atmospheric constraint, negotiated through retrospective narration, and partially carried forward through short-term aftereffects that recalibrate attention and anticipation beyond play. Methodologically, this perspective justifies triangulating behavioral traces with post-play accounts not to eliminate discrepancies but to treat their patterned divergence as a signal of how immersive fear actually operates. Conceptually, it positions immersive fear as an emergent outcome of embodied engagement, affective encoding, and mnemonic reconstruction rather than as a direct response to frightening stimuli.

3. Methodology

This study employed a sequential mixed-methods design to examine immersive fear as a process shaped by atmospheric configuration, embodied regulation, and post-play interpretation. The design consisted of two interconnected phases. Study 1 used scene-based qualitative analysis to identify how fear was enacted, negotiated, and narratively reconstructed during VR gameplay. Study 2 extended this analysis through a within-subject comparison between VR and non-VR gameplay, allowing assessment of whether immersive fear differed systematically across display modes. Study 1 was conducted first, and its findings informed the analytic logic of Study 2. Integration was conceptualized not as parallel description but as connected design in which exploratory conceptual distinctions were tested and refined in the comparative phase (Fetters et al., 2013; O’Cathain et al., 2008; Gibson, 2017). Data were collected in two rounds of fieldwork from March to June 2025.

3.1. Study 1: Exploratory Scene-Based Analysis

3.1.1. Participants

Study 1 involved eight participants (four male, four female) aged 21–45 (M = 27.1, SD = 4.3). Participants were recruited via online and offline gaming communities. The sample was intentionally heterogeneous in VR familiarity and horror-game experience to capture contrastive experiential patterns. Eligibility required adult status, ability to complete VR gameplay, and willingness to engage with horror content. Individuals with severe simulator sickness risk were excluded. Participants received a small honorarium.

3.1.2. Materials and Scene Selection

Resident Evil Village was selected for its diverse fear atmospheres and availability in both VR and non-VR modes. Three focal segments were defined based on distinct atmospheric logics: the Village Attack (combat pressure), the Baby Room (psychological ambiguity), and the Factory (spatial disorientation). Scene-based approaches have been validated in prior mixed-methods VR research for examining how affective and behavioral responses are structured across controlled gameplay conditions (Mekler et al., 2014; IJsselsteijn et al., 2013).

3.1.3. Data Collection

Gameplay sessions were conducted on PlayStation 5 with PlayStation VR2 under controlled indoor conditions. Each session lasted up to two hours, with voluntary pauses or termination permitted. Gameplay footage captured viewpoint and interaction, while an external camera recorded posture, orientation, and spontaneous reactions. Post-play interviews (30–40 minutes) were audio-recorded, transcribed verbatim, and translated into English when necessary. These multimodal data collection methods are consistent with prior VR behavioral research emphasizing triangulation between enacted and reported responses (Creswell & Plano Clark, 2018; Hesse-Biber, 2017).

3.1.4. Data Analysis

Data analysis proceeded via thematic interpretation and structured behavioral coding. Behavioral coding included movement regulation (defensive retreat, hesitant advancement, cautious reorientation), interaction regulation (weapon readiness, resource checking, pausing), and overt emotional regulation (postural tension, gaze avoidance, affective reactivity). Codes were compared with post-play narratives to examine alignment and divergence, following best practices for mixed-methods integration (Guetterman et al., 2015; Plano Clark et al., 2019). Two independent coders achieved strong intercoder reliability (Cohen’s κ = 0.82).

3.2. Study 2: Within-Subject VR/Non-VR Comparative Experiment

3.2.1. Participants

Study 2 recruited 32 participants; two withdrew due to severe cybersickness and were excluded. The final sample included 30 participants (15 male, 15 female; M = 30.17, SD = 8.94). Participants were stratified by VR experience (rich, medium, novice) and horror-game experience (high, medium, low), with balanced distribution across categories. All participants provided informed consent.

3.2.2. Procedure

Participants completed both VR and non-VR conditions in counterbalanced order. Each condition used the same PlayStation 5 system and identical content; only the display differed. Gameplay duration, scene order, interaction, and narrative pacing were strictly controlled. After each condition, participants completed measures of emotion, scene-specific fear, immersion, and motion sickness. Within-subject comparisons have been recommended for isolating mode-specific effects while controlling for individual differences (Weech et al., 2020; Slater, 2018).

3.2.3. Measures

Measures included background variables (age, gender, VR and horror experience), emotional state (SAM: valence, arousal, dominance), scene-specific fear ratings (0–10), and overall immersion and motion sickness ratings (0–10). The use of SAM followed its established value as a parsimonious instrument for capturing dimensional affective states across valence, arousal, and dominance (Bradley and Lang, 1994). More broadly, the decision to combine dimensional affect measures with focused experience indicators was informed by game-experience research emphasizing that player response is multi-dimensional and cannot be reduced to a single global rating (Johnson et al., 2018; Vanden Abeele et al., 2020). These measures therefore captured both general emotional structure and scene-level differences across gameplay modes.

3.2.4. Quantitative Analysis

Descriptive statistics and paired-samples t tests were used to compare VR and non-VR conditions across emotional and experiential variables, including scene-specific fear ratings. Additional descriptive comparisons examined how VR-related fear enhancement varied across experience groups. Analyses were conducted in SPSS 29.0, with significance set at p < .05 and effect sizes reported.

3.3. Ethical Considerations

The study was conducted in accordance with institutional ethical standards. Participants provided informed consent and could withdraw at any time. Sessions were monitored for distress or cybersickness and stopped when necessary. This precaution was important because contemporary VR research has repeatedly shown that immersive systems may produce motion-related discomfort, acute stress, visual fatigue, or mental overload in at least a subset of users, especially under affectively demanding conditions (Smith et al., 2019; Riva et al., 2019; Kelly et al., 2024). All participants were debriefed after participation.

4. Findings

4.1. VR Reconfigured the Emotional Profile of Survival Horror Experience

Paired-samples t tests (Study 2) revealed systematic differences between VR and non-VR conditions across all primary affective dimensions. Relative to non-VR gameplay, VR produced significantly lower valence, t(29) = -4.86, p < .001, d = -0.89, higher arousal, t(29) = 7.83, p < .001, d = 1.43, and lower dominance, t(29) = -4.86, p < .001, d = -0.89. Participants also reported higher overall fear (M = 7.99, SD = 0.77 vs. 6.20, SD = 1.01), t(29) = 16.42, p < .001, d = 3.00, alongside higher immersion, t(29) = 10.02, p < .001, d = 1.83, and higher motion sickness, t(29) = 8.55, p < .001, d = 1.56.

Taken together, the SAM dimensions indicate a coherent shift in emotional structure rather than isolated changes. VR gameplay was characterized by a combination of elevated arousal, reduced valence, and diminished dominance, suggesting an experiential profile marked by intensified activation under reduced perceived control. Importantly, increased immersion co-occurred with both heightened fear and increased motion sickness, indicating that experiential intensification was accompanied by measurable bodily cost rather than purely affective amplification.

These results suggest that VR reorganized survival horror experience into a distinct configuration combining heightened arousal, reduced control, and increased sensory and physiological demand.

4.2. Fear Intensification Varied Across Atmospheric Configurations

Scene-level paired comparisons showed that VR-related fear increases were consistent but not uniform across the three focal segments. Fear ratings were significantly higher in VR than in non-VR for the Village Attack (t(29) = 9.25, p < .001, d = 1.69), Baby Room (t(29) = 8.00, p < .001, d = 1.46), and Factory (t(29) = 9.01, p < .001, d = 1.64).

However, the distribution of fear differed by scene. The Baby Room produced the highest fear ratings under VR (M = 8.77), followed by the Village Attack and the Factory. This pattern indicates that VR effects were sensitive to the dominant atmospheric logic of each segment rather than operating as a uniform amplification mechanism.

Psychologically ambiguous and proximity-based environments showed the strongest intensification, whereas spatial disorientation produced a more moderate but still significant increase. These results suggest that VR selectively amplifies particular forms of atmospheric threat rather than uniformly increasing fear across contexts.

4.3. Distinct Patterns of Embodied Regulation Across Scenes

Behavioral coding in Study 1 identified three recurrent modes of embodied fear regulation aligned with the focal scenes. In the Village Attack, fear was primarily enacted through defensive retreat, rapid backward movement, and pre-emptive weapon readiness under compressed decision time. In the Baby Room, participants exhibited hesitant advancement, prolonged pauses, gaze aversion, and controlled perceptual exposure. In the Factory, fear was associated with cautious reorientation, including repeated head-turn scanning, non-optimal pathing, and frequent pauses for spatial reassessment.

These patterns were consistently observed across participants and coding dimensions, with high intercoder reliability (κ = 0.82), indicating that they represent stable behavioral tendencies rather than isolated responses. Across scenes, fear was not expressed as a uniform increase in intensity but as differentiated forms of bodily adjustment, including movement modulation, attentional regulation, and interaction pacing.

The scene-specific regulatory patterns provide a mechanism-level account of the quantitative differences observed in Study 2, linking variations in fear ratings to distinct modes of embodied engagement.

4.4. Fear Was Mediated by Experience and Extended Through Interpretation

Descriptive comparisons across experience groups indicated that prior experience shaped the magnitude of VR-related fear increase. Participants with lower VR experience showed larger increases in fear between VR and non-VR conditions (novice: +1.97 vs. experienced: +1.77), whereas participants with moderate to high horror-game experience showed greater amplification than low-experience participants. These patterns suggest that experiential background influenced how strongly VR conditions intensified fear, although no inferential tests were conducted due to sample size.

Behavioral and narrative data further revealed systematic discrepancies between enacted responses and retrospective accounts. Participants frequently reinterpreted avoidance behaviors as strategic positioning, while forward movement sometimes coexisted with reported panic. These divergences indicate that fear was not fully captured by either behavioral or self-report measures alone, but emerged through a combination of embodied action and retrospective interpretation.

Short-term residual effects were also reported. Six of eight participants described intrusive recall or mental imagery shortly after gameplay, and four reported heightened vigilance in real-world environments. Reluctance to replay the game alone was consistently low (M = 2.3/10). These patterns suggest that immersive fear extended beyond gameplay as a temporally distributed process involving post-experience cognitive and perceptual adjustment.

5. Discussion

The findings reposition immersive fear in VR survival horror from a question of emotional magnitude to a question of emotional organization. What matters is not simply that VR heightens fear relative to non-VR play, but that it reorders the relations among perception, action, and interpretation. Once fear is approached in this way, several assumptions that remain common in VR research become insufficient. Presence is no longer reducible to a stronger feeling of “being there,” atmosphere is no longer a sensory wrapper around gameplay, and player experience can no longer be treated as a background characteristic that merely strengthens or weakens a general effect. Instead, the results point toward a more relational account in which immersive fear is produced through the encounter between atmospheric structure, embodied regulation, and retrospective sense-making. This reframing is consistent with broader work showing that immersive media tend to strengthen presence, that the experience of VR horror depends partly on users’ confidence in managing mediated fear, and that emotional episodes are reconstructed in patterned ways rather than simply retained as raw experience (Cummings and Bailenson, 2016; Makransky & Lilleholt, 2018; Schacter et al., 2011; Wan et al., 2023). Recent evidence suggests that the temporal unfolding of fear is modulated not only by scene-specific cues but also by adaptive attentional strategies, highlighting how the interplay of prior experience and in-situ perception continuously shapes the affective trajectory (Riva et al., 2022).

5.1. Atmospheric Configuration Organizes Fear as a Mode of Embodied Regulation

The results indicate that fear in immersive horror is best understood as a patterned mode of bodily adjustment shaped by atmospheric configuration. This point matters because much of the literature on VR affect, even when sophisticated, still tends to model fear as an outcome to be increased, measured, and compared. That orientation is useful for establishing that VR often heightens arousal and presence, but it is less capable of explaining why distinct horror scenes produce different kinds of tension, different forms of hesitation, and different bodily responses under otherwise comparable technological conditions. Existing comparative work has shown that immersive formats can intensify negative emotion and presence relative to less immersive media, but it also indicates that these effects vary with content and interactional structure (Lemmens et al., 2022; Steinhaeusser et al., 2023). Similar observations have been reported in multi-modal VR simulations, where specific design features such as controlled auditory ambiguity directly influenced both physiological arousal and decision-making latency, emphasizing the role of atmosphere in constraining behavioral possibilities (Slater et al., 2021).

What emerges from the present study is a more differentiated picture. Combat-intensive scenes did not simply become “more frightening” in VR; they compressed temporal margins for decision and pushed players toward rapid defensive retreat, weapon readiness, and spatial repositioning. Psychologically ambiguous scenes did not merely elevate dread; they suspended action, slowed movement, and redistributed attention across partial visibility, delayed approach, and controlled exposure. Spatially disorienting scenes did not just create unease; they converted fear into repeated reorientation, cautious scanning, and unstable navigation. In other words, atmospheric configuration operated as a regulatory field that allocated what kinds of action were viable, risky, or temporarily suspended. Fear was therefore not only felt but enacted through specific adjustments in pacing, posture, orientation, and interaction. These nuanced patterns mirror findings in spatial cognition research, suggesting that environmental complexity systematically biases attentional allocation and embodied action even in non-VR horror contexts (Meehan et al., 2005).

This shifts the theoretical emphasis from stimulus intensity to action structure. Lighting, sound, spatial compression, and environmental uncertainty are consequential not because they directly inject fear into the player, but because they alter the conditions under which bodies can anticipate, orient, and respond. In this sense, atmosphere should be treated less as an expressive layer and more as a constraint-producing arrangement that organizes the practical management of threat. Such a view deepens existing atmosphere-oriented scholarship by showing that audiovisual-spatial design is not only affectively charged but operationally directive: it tells the player’s body how quickly it must respond, how much it can trust perception, and whether movement should accelerate, stall, or disperse. The importance of VR lies precisely in tightening this relation between environmental structure and embodied adjustment, which is why the same scene in VR and non-VR can share narrative content yet diverge sharply in experiential form.

5.2. Experience Mediates Fear not as Resistance to Intensity, but as Interpretive Competence

The findings also suggest that prior experience should be conceptualized less as a moderating demographic feature and more as a form of interpretive competence. This distinction is analytically important. In many empirical studies, user background is incorporated primarily as a control variable: more experience is expected to reduce uncertainty, dampen novelty, or weaken emotional response. Yet the present results do not support such a simple attenuation model. More experienced participants were not uniformly less affected; rather, they were differently affected. They more readily recognized atmospheric cues as structured design signals, anticipated how tension would unfold, and translated sensory ambiguity into actionable information. Less experienced participants, by contrast, were more likely to encounter the same cues as diffuse and poorly differentiated threat, which increased overload, narrowed control, and made fear more difficult to organize. This aligns with recent models of embodied cognition in VR that argue prior domain-specific knowledge supports the rapid appraisal of environmental affordances, directly mediating stress and engagement outcomes (Diemer et al., 2015).

This pattern is consistent with prior work on self-efficacy in VR horror, which shows that the enjoyment of frightening gameplay is closely tied to players’ confidence in coping with mediated fear rather than to fear reduction as such (Lin et al., 2018). It also aligns with research indicating that immersion and presence do not operate as invariant psychological effects; their consequences depend on how users recruit prior expectations and interactional stability under immersive conditions (Weech et al., 2020; Reer et al., 2022; Wan et al., 2023). The issue, then, is not simply whether a player is experienced, but what that experience allows the player to do with the environment. Experience provides a repertoire of interpretive habits: it makes some atmospheric features legible as pacing devices, others as tactical prompts, and still others as aesthetic constructions rather than immediate existential threat. Without such repertoires, the same environment becomes harder to parse and harder to regulate. Furthermore, the literature suggests that individual differences in affective regulation strategies modulate the translation of environmental cues into fear perception, highlighting the intersection between experience and physiological reactivity (Rothbaum et al., 2015).

Reframing experience in this way also clarifies the relation between presence and control. A persistent tendency in VR research is to assume that stronger presence straightforwardly enhances experience. The present findings suggest a more conditional relationship. Presence can intensify involvement, but without sufficient interpretive grounding it may also intensify workload, perceptual strain, disorientation, and motion-related discomfort (Yildirim, 2020; Souchet et al., 2023). What differentiates compelling immersion from aversive overload is not presence alone, but the user’s ability to sustain coherent expectations about how the environment behaves and what kinds of action remain available within it. This is why the same scene may be lived as suspense by one player and as destabilizing threat by another. The difference lies not in raw exposure, but in whether the environment can be read as structured, anticipatable, and therefore partially governable. Under this view, immersive fear is mediated through a dynamic fit between atmospheric demand and interpretive capacity rather than produced by technological intensity in any simple sense.

5.3. Fear Extends Beyond Gameplay Because It Is Reconstructed as Well as Enacted

Perhaps the most theoretically consequential implication of the study lies in the relationship between embodied action and retrospective narration. The discrepancy observed between what players did and how they later described what they did should not be treated as a methodological inconvenience to be corrected away. It is better understood as evidence that fear in immersive environments unfolds across temporally distinct but interdependent registers. One register is immediate and embodied: the body retreats, scans, hesitates, pauses, or presses forward under atmospheric constraint. Another is retrospective and narrative: the same actions are later reorganized into accounts of competence, tactical reasoning, aesthetic appreciation, or expected genre response. These are not redundant descriptions of the same event. They are different phases of how the event becomes meaningful. Prior research on affective memory reconstruction supports this interpretation, showing that peak-end moments disproportionately shape retrospective emotion reports even when micro-actions indicate sustained vigilance (Fredrickson & Kahneman, 1993).

Research on remembered experience supports precisely this point. Retrospective accounts do not simply recover prior states but rebuild them through selective retrieval and interpretive framing (Schacter et al., 2011). Related findings on peak-end effects show that remembered episodes are disproportionately organized around salient high-intensity moments and endings, rather than around the full temporal distribution of feeling (Kahneman et al., 1993; Ariely and Carmon, 2000). Once these insights are brought into dialogue with immersive horror, the behavioral–narrative discrepancy in the present study becomes theoretically legible. It shows that fear is not only regulated in the moment but also narratively stabilized afterward. Low-grade but behaviorally consequential tension—expressed through posture, gaze aversion, cautious movement, or momentary freezing—may be underrepresented in retrospective talk because it lacks the legibility of jump scares or dramatic encounters. Conversely, strongly visible avoidance can later be reframed as strategic agency. The point is not that participants are inaccurate; the point is that immersive fear has to pass through narrative reconstruction before it becomes reportable as experience. Integrating findings from extended VR exposure studies, it is clear that such temporal layering allows fear responses to generalize across contexts, implicating potential applications in adaptive horror game design and training scenarios (Riva et al., 2022).

This temporal layering also helps explain the short-term residual effects observed after gameplay. Intrusive recall, heightened vigilance, and reluctance to re-enter similar situations indicate that fear does not terminate when the headset is removed. Instead, the experience continues as a reorganization of attention and anticipation. Recent VR research suggests that immersive exposure may generate emotional carryover, stress-related aftereffects, and post-experience discomfort that remain analytically relevant beyond the moment of formal play (Lavoie et al., 2021; Pan & Hamilton, 2020; Slater & Sanchez-Vives, 2021). Residual unease is therefore not merely leftover arousal. It is part of the post-hoc integration of embodied impressions into everyday perceptual expectations. What persists is not only an emotional trace but a changed orientation toward possible threat.

Taken together, these findings support a processual conception of immersive fear as something enacted, reconstructed, and carried forward. Theoretical accounts that confine fear to momentary stimulus response miss this temporal extension. What VR intensifies is not only sensation, but the coupling of bodily regulation with retrospective and prospective meaning-making. Fear in immersive horror therefore cannot be reduced to what players feel at the moment of encounter. It includes what their bodies do before they can fully articulate it, how they later reinterpret those actions, and how the encounter continues to shape attention after the formal experience has ended. That is why behavioral traces and post-play accounts should not be forced into artificial agreement. Their divergence is not a defect in the data; it is one of the clearest indicators of how immersive fear actually operates.

6. Conclusion

Immersive fear in survival horror is best understood as a temporally distributed and structurally organized process rather than a uniform amplification of emotion. Fear emerges through the interaction of atmospheric configuration, embodied regulation, and players’ prior experiential repertoires, producing scene-specific emotional and behavioral profiles across both VR and non-VR modes. Environments characterized by psychological ambiguity and uncertainty, such as the Baby Room, amplify tension and elicit complex patterns of bodily adjustment, demonstrating how spatial, auditory, and narrative cues collectively shape the enactment and modulation of fear. Research on presence and affect in VR highlights that emotional engagement with the environment influences how users allocate attention and memory resources, suggesting that immersive contexts systematically alter affective processing rather than merely intensify it (Cadet & Chainay, 2020).

Distinct modes of embodied regulation—including defensive retreat, hesitant exposure, and cautious reorientation—reflect the ways in which players negotiate risk, control, and perceptual uncertainty in real time. Prior experience further mediates these responses, enabling some players to anticipate threat, optimize movement, and translate ambiguous cues into strategic action, while less experienced players exhibit more diffuse and reactive behaviors. Retrospective narratives reveal divergences from real-time behavior, highlighting that fear continues to be reorganized cognitively and affectively after gameplay, extending its influence into post-play attention, memory, and anticipatory perception. Systematic reviews of VR emotion and memory show that immersive emotional experiences shape subsequent mnemonic outcomes in ways that are distinct from non-immersive exposures, underscoring that emotional contexts within VR can persist beyond the moment of experience (Bailey et al., 2023).

Theoretical interpretation of immersive fear benefits from recognizing it as a relational phenomenon, where emotion, action, and cognition are co-constructed by environmental affordances and interpretive competence. Temporal interdependency between perceived presence and fear has been observed in repeated VR exposure, indicating that emotional experience and the sense of presence evolve together across time, which supports conceptualizing fear as a non-static affective trajectory rather than a fixed outcome of exposure (Slater et al., 2010). Such a perspective reframes fear as a process enacted, interpreted, and stabilized over time, rather than as a static outcome of exposure or technological intensity. These insights provide a foundation for future work to explore other immersive horror experiences, repeated exposure, and the interaction of environmental design with player expertise, offering a more nuanced understanding of how virtual environments structure affective engagement, regulatory strategies, and cognitive-emotional adaptation.

The study makes four interrelated contributions to the literature on VR-mediated horror experiences and immersive game research.

First, it advances theoretical understanding of immersive fear in VR horror by shifting the analytical focus from coarse emotional amplification to the differential reconfiguration of fear across multiple experiential dimensions. Unlike prior work that primarily treats fear as a unidimensional outcome, the present findings demonstrate how VR modulates valence, arousal, perceived control, immersion, and motion-related discomfort in scene-specific ways. This nuanced perspective reveals that fear is dynamically structured, contingent upon atmospheric configuration, player embodiment, and interpretive processes, thereby offering a more temporally and behaviorally grounded conceptualization of immersive affect. By linking scene-level environmental cues to participant-specific regulatory patterns, the study provides a framework for conceptualizing immersive fear as an emergent process rather than a static affective state.

Second, it extends atmosphere-oriented game studies by empirically showing that distinct fear atmospheres recruit differentiated bodily and attentional regulatory strategies. The integration of combat-intensive, psychologically ambiguous, and spatially disorienting scenes illustrates how VR heightens the visibility of these patterns, allowing for the identification of context-specific embodied adjustments such as defensive retreat, controlled exposure, and repeated reorientation. This provides concrete evidence that game atmosphere operates as a regulatory field shaping not only subjective experience but also observable behavior. These insights deepen theoretical models of presence, threat anticipation, and embodied cognition in immersive games, offering a bridge between perceptual design, player action, and affective response.

Third, the study contributes methodologically by demonstrating the value of integrating scene-based qualitative analysis with a within-subject quantitative comparison across VR and non-VR conditions. This combined design allows for the simultaneous assessment of experiential, behavioral, and affective outcomes at high temporal and scene-level resolution. It provides a robust template for future research aiming to disentangle technological affordances, narrative context, and player experience, moving beyond reliance on self-report alone. By linking mechanism-oriented interpretation with mode-based comparison, the study establishes a replicable approach for investigating how immersive technologies shape complex emotional processes in dynamic interactive environments.

Finally, the findings have implications for both game design and applied VR experiences. Understanding how atmospheric cues selectively modulate embodied fear responses can inform adaptive horror design, therapeutic VR interventions, and experimental paradigms in emotion research. The results suggest that designers and researchers should consider scene-specific affordances and player experience as integral components for shaping affective outcomes in immersive media.

Several limitations should be acknowledged. First, although Study 2 substantially strengthens the empirical basis of the article, the sample remains modest for fine-grained subgroup analysis, and the experience-based patterns should therefore be interpreted descriptively rather than as formal moderation effects. Second, the study focuses on a single title, Resident Evil Village, which limits the immediate generalizability of the findings across other horror subgenres, pacing structures, or VR systems.

Future research may build on these findings in at least two directions. One direction is to examine whether the same atmosphere–body–narrative process can be observed in other horror games or immersive genres with different symbolic structures and interaction logics. Another is to explore longer-term trajectories of fear regulation, including habituation, residual aftereffects, and repeated exposure over time. Such work would further clarify how immersive media shape not only momentary emotional response, but also evolving habits of interpretation and embodied anticipation.