Cognitive efficiency in VR simulated natural indoor environments examined through EEG and affective responses

Abstract

This study investigated the neurophysiological and affective responses elicited by nature-inspired indoor design elements, including curvilinear forms (CL), nature views (N), and wooden interiors (W), in a virtual environment, and their effects on cognitive performance. Thirty-six participants experienced one control and three experimental conditions in a within-subject design. Electroencephalography (EEG) was used to record neural activity, relaxation and valence ratings assessed affective states, and standardized tasks measured cognitive performance. The W condition elicited EEG patterns indicative of relaxed attentional engagement, including increased alpha-to-theta (ATR) and alpha-to-beta (ABR) ratios, and a decreased theta-to-beta (TBR) ratio. These neural patterns were associated with higher self-reported relaxation and positive affect, and with enhanced cognitive performance relative to the control condition. In contrast, the CL and N conditions did not improve cognitive performance, and the N condition showed elevated physiological arousal, likely due to heightened visual stimulation. Regression analysis identified ATR and relaxation as significant predictors of cognitive performance, emphasizing the role of emotional stability and neural balance in supporting task engagement. Overall, the findings highlight the potential of nature-inspired design to foster a synergy between psychological relaxation and cognitive attention, though further research is needed across diverse spatial typologies to isolate specific design parameters.

Introduction

In contemporary society, humans spend approximately 80–90% of their time in indoor environments, making the quality and characteristics of these spaces crucial determinants of human well-being and behavior¹. The relationship between individuals and their indoor surroundings manifests through various pathways, including physical, psychological, and social dimensions, ultimately contributing to the overall quality of life. The dynamic interplay between occupants and the built environment creates a reciprocal relationship where environmental conditions influence human behavior and well-being. Understanding these intricate interactions is paramount for creating spaces that optimize human health, productivity, and satisfaction in an era where indoor spaces increasingly define the human experience. Given the profound influence of environmental factors on human well-being, researchers have increasingly focused on understanding how indoor spaces can be designed to promote positive outcomes.

The potential of indoor environments to promote well-being has emerged as a key focus in built environment research^2–5. Indoor environments and well-being are intricately interconnected across various dimensions, including mental, psychological, and physical health, offering substantial potential benefits. Among the various benefits, “relaxation” has emerged as a key psychological outcome, highlighting the connection between human experience and environmental design. As a state of reduced physical or mental tension, relaxation fulfills the universal desire for comfort and tranquility in spaces^6–8. This psychological effect complements the functional and aesthetic aspects of indoor spaces, underscoring their multifaceted influence on well-being.

A well-established principle of relaxation is the positive influence of natural environments on physiological and emotional well-being^9–13. Two prominent theories, Attention Restoration Theory (ART)^14,15 and Stress Recovery Theory (SRT)¹⁶, emphasize the health and psychological benefits of nature exposure. Kaplan’s ART suggests that direct interaction with nature facilitates recovery from attention fatigue. Similarly, Ulrich’s SRT demonstrates that even indirect exposure to nature—such as viewing natural landscapes through a hospital window—can lower heart rates and accelerate physical recovery. Whether through direct engagement or indirect observation, nature consistently promotes relaxation, alleviates cognitive fatigue, and promotes psychological recovery^17–19.

Design strategies that incorporate nature-inspired elements, such as indoor plants^20,21, curvilinear shapes and forms reflecting botanical motifs, and the use of natural materials like wood, extend the relaxation benefits of natural environments to indoor settings^22–24. Humans inherently gravitate toward organic shapes and biophilic elements, with curved contours and rounded furniture reflecting nature’s aesthetics and promoting a sense of comfort and well-being^25–27. Additionally, research shows an association between spaces with wooden finishes and enhanced positive emotions and psychological comfort^28,29, while other studies further link wood in interior design to reduced heart rates and a greater sense of tranquility^30,31. These findings underscore the significant potential of nature-inspired interiors to induce relaxation and provide substantial psychological and physiological benefits.

Indoor environments influence human well-being through the interaction of environmental stimuli with human sensory and nervous systems, which trigger physiological, psychological, and behavioral responses^32–34. This interaction underscores the profound impact of indoor settings on shaping mental and physical states. For instance, specific environments can induce relaxation, enhance mood, and improve focus and productivity. While previous studies have explored these effects, a critical question remains: can the psychological benefits of relaxation-oriented environments extend to cognitive performance improvements?

Despite growing evidence of nature’s restorative effects on cognitive functioning, the direct cognitive outcomes of psychological relaxation in indoor environments remain unclear. Although some studies suggest that nature-integrated spaces can enhance cognition^35–37, empirical efforts to establish a systematic link between nature-inspired indoor environments and improved cognitive performance have been limited³⁸. Moreover, the mechanisms through which relaxation-oriented environments influence cognition remain insufficiently explored. Addressing this gap requires rigorous empirical studies that integrate emotional, physiological, and cognitive measures to clarify whether and how environments promoting psychological relaxation translate into tangible cognitive benefits.

Previous research has indicated that various natural design elements—such as curvilinear forms, nature views, and wooden materials—can contribute to users’ sense of relaxation and well-being. However, few studies have directly compared these different design strategies within a controlled experimental context. A systematic investigation of how users respond to different nature-inspired environmental features can deepen our understanding of their influence on relaxation and cognitive functioning, and guide future efforts to identify their underlying mechanisms.

The relaxation experienced in architectural spaces is shaped by complex interactions, making it challenging to quantify. Electroencephalography (EEG), a non-invasive method for measuring cortical electrical activity³⁹, provides an objective approach to assessing brain function, emotions, and psychological responses in built environments^40,41. EEG captures real-time neural activity, quantifying perceptions of comfort, emotion, and attention⁴¹, thereby bridging architecture and neuroscience to explore how indoor spaces influence cognitive and affective states^42,43. Previous studies have validated EEG’s utility in evaluating emotional responses to architectural features such as geometry, lighting, and natural elements, reinforcing its role in understanding human-environment interactions^44–46.

Advancements in virtual reality (VR) have further enhanced architectural research by enabling immersive simulations that eliminate spatial constraints. VR allows for precise control over environmental variables, minimizing confounding factors while systematically manipulating environmental attributes. Recent studies integrating VR with EEG and cognitive assessments have established its reliability in evaluating human responses to built environments^40,47–49.

To address a gap in existing research, this study directly examines the relationship between natural indoor environmental design and cognitive performance by employing quantitative neurophysiological affective, and behavioral data. By integrating EEG-based neurophysiological measures self-reported affective responses, and cognitive performance assessments, this study provides a comprehensive and objective evaluation of how relaxation-oriented environments influence cognitive function.

To achieve this, EEG is used to measure attentional control and relaxation-related neurophysiological states, while self-reports assess perceived relaxation and emotional engagement. Cognitive tasks evaluate whether these changes translate into measurable improvements in performance. Additionally, VR-based controlled indoor simulations allow for the precise manipulation of environmental variables while minimizing confounding factors. Through this approach, the study provides empirical evidence on the impact of natural design elements on cognitive performance.

Accordingly, this study aims to explore whether nature-integrated indoor environments are associated with improvements in cognitive performance through their relaxation effects. The study compares user responses to three representative natural design features—Curvilinear forms (CL), Nature view (N), and Wooden interiors (W)—against a neutral control condition (C). It examines how these design elements influence neurophysiological activity, emotional states, and cognitive outcomes. Data collection includes EEG recordings, self-reported affective responses, and cognitive performance assessments, enabling a multi-dimensional understanding of how relaxation-related processes may relate to cognitive functioning. By integrating these measures, the study seeks to contribute to the growing body of research on the psychological and cognitive impacts of natural design elements in indoor environments.

To address its research objectives, the study hypothesized that indoor environments incorporating natural design attributes would enhance cognitive performance by fostering improved neurophysiological and affective states. Specifically, it posits that a relaxed yet attentive mental state and positive affective responses will serve as key mediators in cognitive performance improvements. The following hypotheses guided the experiments.

H₁

Nature-integrated indoor environments will induce neural activity patterns associated with a relaxed and attentive mental state compared to a neutral indoor environment without natural elements.

H₂

Participants in nature-integrated indoor environments will report greater positive emotions and relaxation than those in the control condition.

H₃

Cognitive performance will be higher in nature-integrated indoor environments compared to the control condition.

H₄

Neurophysiological patterns associated with relaxation and attentiveness, along with positive affective responses, will correlate with higher cognitive performance.

Results

Repeated measures of analysis of variance (ANOVA) were performed to compare the experimental conditions with the control condition. Three EEG indicators—alpha-to-theta ratio (ATR) in frontal region, theta-to-beta ratio (TBR) in frontal region, and alpha-to-beta ratio (ABR) in occipital region—were analyzed to evaluate neurophysiological responses (NR). Affective responses (AR) were assessed using two dimensions: relaxation and valence. Cognitive task scores served as indicators of cognitive performance. Table 1 presents the means for each measure under different conditions and the statistical significance of the main effects of conditions. The results revealed that the condition had a statistically significant main effect across all measurements. To further explore these differences, post-hoc comparisons were conducted. The following sections detail the findings for each measurement category.

Table 1.

The means for each measure under different conditions and the statistical significance of differences across conditions.

(N = 36) Measurement		Mean (SD)				Within subjects effects
		C	CL	N	W	F	p	ηp2
NR	ATR	0.49 (0.37)	0.57 (0.32)	0.57 (0.39)	0.61 (0.37)	4.27	0.007	0.109
	TBR	9.7 (5.26)	8.31 (4.53)	8.6 (4.99)	7.52 (3.66)	4.19	0.008	0.107
	ABR	3.87 (2.02)	5.11 (2.01)	4.63 (2.28)	4.79 (2.36)	5.561	0.001	0.137
AR	Relaxation	6.08 (1.37)	5.46 (1.25)	5.6 (1.33)	6.68 (1.43)	7.94	< 0.001	0.185
	Valence	5.31 (1.5)	6.42 (1.81)	7.35 (1.51)	7.57 (1.17)	22.20	< 0.001	0.388
Cognitive task		0.73 (0.09)	0.71 (0.08)	0.71 (0.08)	0.78 (0.1)	31.60	< 0.001	0.475

Neurophysiological responses

Figure 1 summarizes the post hoc comparison results for neurophysiological responses, and Fig. 2 visualizes the distribution of mean differences between the control condition and the experimental conditions using a topographic map.

Fig. 1.

Comparison of mean differences across conditions for electroencephalography (EEG) indicators. Statistically significant differences are indicated as *p < 0.05, **p < 0.01, and ***p < 0.001.

Fig. 2.

Topographic maps illustrating the mean differences between each experimental condition (CL, N, W) and the control condition (C) for the ATR, TBR, and ABR indices. For ATR, the frontal region in the W-C comparison shows the most intense yellow, indicating higher ATR values in the W condition, which suggests a focused yet relaxed cognitive state. For TBR, the frontal region in the W-C comparison is the most intensely blue, representing the lowest TBR values, which suggests reduced cognitive load and enhanced attentional focus in the W condition. For ABR, the occipital region appears yellow in the CL-C and W-C comparisons, reflecting higher ABR values and a more relaxed state, whereas N-C appears blue, indicating lower ABR values and a less relaxed, visually stimulated state. These findings provide neural evidence of mental state differences depending on natural design elements in indoor environments.

• Alpha-to-theta ratio (ATR): A significant main effect of condition was observed, (F(3, 105) = 4.27, p = 0.007, η²_p = 0.109), indicating that ATR in the frontal region significantly varied across the conditions. The partial eta squared value suggests a moderate-to-large effect size, underscoring the meaningful impact of indoor environmental conditions on ATR. While the overall main effect was statistically significant, pairwise comparisons revealed that only the wooden interiors (W) condition showed a significant difference compared to the control condition (W–C; t(35) = 3.134, SE = 0.039, p = 0.021). In contrast, the curvilinear forms (CL) condition (CL–C; t(35) = 2.383, SE = 0.034, p = 0.136) and the view of nature (N) condition (CL–N; t(35) = 2.158, SE = 0.041, p = 0.227) demonstrated higher ATR values relative to the control condition, but these differences did not reach statistical significance.

  Higher ATR reflects greater internal attention and a balanced state of relaxation and alertness, which benefits cognitive efficiency. In contrast, lower ATR indicates increased mental fatigue or diminished attentional regulation. A significant increase in ATR in the W condition suggests that wooden interiors contribute to a more focused yet relaxed cognitive state, supporting sustained attention with reduced mental strain. While the CL and N conditions also exhibited higher ATR values compared to the control condition, these differences were not statistically significant. This lack of statistical significance indicates that while curvilinear forms and nature views may have some influence on attentional control, their impact is relatively modest and may not be strong enough to induce a consistent effect across individuals.

• Theta-to-beta ratio (TBR): A significant main effect of condition was found, (F(3, 105) = 4.19, p = 0.008, η²_p = 0.107), reflecting a moderate effect size, indicating a meaningful impact of the experimental environments. Post hoc comparison identified significant differences between specific conditions. The W condition exhibited a significantly lower TBR compared to C (t(35) = − 4.271, SE = 0.511, p < 0.001), while the CL and N conditions did not show statistically significant differences (CL–C: t(35) = − 2.437, SE = 0.571, p = 0.12; N–C: t(35) = − 1.101, SE = 0.781, p = 1).

  TBR serves as a neural marker of cognitive control, with lower values indicating reduced cognitive distraction and heightened focus. The significant reduction in TBR in the W condition suggests that theta activity, which reflects externally directed attention, decreased relative to beta activity, indicating that participants’ attention remained more internally focused rather than being drawn to external stimuli. This finding implies that wooden materials create a less distracting environment, thereby minimizing cognitive interference and enhancing sustained attentional engagement. In contrast, the CL and N conditions did not yield statistically significant differences compared to the control condition, suggesting that environments with curvilinear elements and nature views may have diverted attention outward, potentially increasing cognitive distraction and reducing sustained focus.

• Alpha-to-beta ratio (ABR): Analysis revealed a significant main effect of condition, F(3, 105) = 5.56, p = 0.001, η²_p = 0.137, indicating substantial differences in ABR across conditions. Post hoc comparisons revealed that the CL condition demonstrated the largest difference, t(35) = 4.246, SE = 0.292, p < 0.001, when compared to the control condition. Similarly, the W condition exhibited a significant difference from the control, t(35) = 2.921, SE = 0.315, p = 0.036. In contrast, the N condition showed no significant difference compared to the control condition, t(35) = 2.024, SE = 0.377, p = 0.304.

  Higher ABR indicates a state of relaxation and reduced strain. Additionally, an increase in ABR reflects heightened occipital alpha activity, which is associated with greater visual comfort. The significant rise in ABR in the W and CL conditions, indicating greater alpha power, suggests that these environments fostered both relaxation and visual ease, contributing to a more relaxed neurophysiological state. In contrast, the relatively lower ABR in the N condition indicates reduced alpha activity, suggesting that occipital alpha power desynchronization may be linked to the increased demand for spatial visual information processing in this environment. This finding implies that while participants perceived the W and CL conditions as visually more comfortable, the N condition may have required greater cognitive resources for processing visual stimuli, leading to a less relaxed state.

In summary, the experimental conditions exhibited distinct neurophysiological responses compared to the control condition. The W condition, in particular, demonstrated significant differences in ATR, TBR, and ABR. Specifically, the increase in ATR reflects greater alpha activity, suggesting enhanced attentional engagement, while the significant decrease in TBR indicates reduced theta activity, promoting cognitive stability and minimizing external distractions. Additionally, the increase in ABR signifies higher alpha activity, reinforcing the role of wooden interiors in fostering mental relaxation. These patterns suggest that the W condition supports a balance between relaxation and attentional control, potentially enhancing cognitive performance in sustained tasks. In contrast, while the CL and N conditions showed some trends in ATR and ABR, they did not produce significant neurophysiological changes, suggesting weaker and less consistent effects on attentional regulation and relaxation.

Affective responses

A significant main effect of condition was observed on relaxation (F(3, 105) = 7.94, p < 0.001, η²_p = 0.185) and valence (F(3, 105) = 20, p < 0.001, η²_p = 0.388), indicating substantial differences in affective responses across the conditions. The large effect sizes highlight the significant influence of environmental conditions on participants’ relaxation and emotional valence, revealing distinct affective experiences depending on the indoor environment.

Figure 3 illustrates the post hoc comparisons. For relaxation (relaxed versus aroused), participants reported significantly higher relaxation in the W condition compared to the CL (t(35) = 3.93, SE = 0.31, p = 0.002) and N conditions (t(35) = 5.095, SE = 0.211, p < 0.001). No significant differences emerged among the other pairwise comparisons, suggesting that the W condition uniquely enhanced relaxation relative to the other environments.

Fig. 3.

Comparison of mean differences across conditions for affective responses and cognitive performance. Statistically significant differences are indicated as **p < 0.01, ***p < 0.001.

For valence (positive or negative), significant differences across multiple conditions were observed. All experimental conditions (CL, N, W) elicited significantly higher positive emotions compared to the C condition (CL–C: t(35) = 3.028, SE = 0.367, p = 0.028; N–C: t(35) = 8.678, SE = 0.235, p < 0.001; W–C: t(35) = 7.272, SE = 0.311, p < 0.001). Participants in the W condition also reported significantly higher positive valence compared to the CL condition (t(35) = 4.323, SE = 0.266, p < 0.001), with the highest positive emotion scores observed. These results indicate that all experimental conditions promoted positive emotions, with the W condition showing a particularly distinct advantage.

To identify affective patterns, a scatter plot analysis was conducted, identifying dominant patterns associated with each condition. The results are as follows, as illustrated in Fig. 4. The CL and N conditions exhibited positive emotion coupled with aroused affect, while the W condition combined positive emotion with relaxed affect. These findings highlight the unique emotional states elicited by each environment. Notably, the N and W conditions evoked positive emotions but differed in their affective characteristics: the N condition generated relatively higher arousal, while the W condition promoted enhanced relaxation. Similarly, the CL condition, like the N condition, elicited positive emotion but with aroused affect.

Fig. 4.

Scatter plot of affective responses across experimental conditions.

Cognitive performance

A repeated measures ANOVA revealed a significant main effect of condition on cognitive performance (F(3, 105) = 31.6, p < 0.001, η²_p = 0.475), indicating substantial differences across the conditions. The large effect size underscores the significant impact of environmental conditions on participants’ cognitive performance.

Post-hoc comparisons, illustrated in Fig. 3, revealed that participants in the W achieved significantly higher cognitive task scores than those in other conditions (W–C, t(35) = 8.137, SE = 0.007, p < 0.001); W–CL: t(35) = 7.55, SE = 0.009, p < 0.001; W–N: t(35) = 6.367, SE = 0.011, p < 0.001). In contrast, no significant differences emerged among the C, CL, and N conditions, indicating comparable cognitive performance across these groups.

These results highlight the W condition’s superiority in enhancing cognitive performance, consistently producing the highest scores. By comparison, the CL and N conditions, while performing similarly to the control condition, were less effective in facilitating cognitive improvements. This finding suggests that incorporating wood elements into indoor environments offers distinct advantages for boosting cognitive performance compared to other design conditions.

Effects of neurophysiological and affective responses on cognitive performance

Regression analysis was conducted to investigate how neurophysiological and affective responses influenced cognitive task performance under various environmental conditions. The result, summarized in Table 2, revealed several key findings. The final model was statistically significant, F(5,138) = 4.774, p < 0.001, explaining approximately 14.8% of the variance in cognitive task performance (R² = 0.148, adjusted R² = 0.117). This reflects a moderate level of explanatory power, indicating that the predictors in the model account for a reasonable proportion of the variance in cognitive performance. Additionally, multicollinearity was not a concern, as all VIF values remained below 3 (ATR = 2.72, ABR = 2.45, TBR = 2.06, valence = 1.41, relaxation = 1.16, and Environment = 1.66).

Table 2.

Final regression model results on neurophysiological and affective effects on cognitive performance.

(N = 36) Independent variable		Unstandardized		Standardized	t	Sig.
		β	Std. error	β
	(Intercept)	0.788	0.031	NA	25.244	0.000
NR	ATR	0.053	0.021	0.206	2.595	0.010
AR	Relaxation	0.018	0.009	0.173	2.058	0.041
Environment	CL	0.003	0.021	0.013	0.125	0.374
	N	0.004	0.023	0.017	0.124	0.314
	W	0.076	0.024	0.353	3.219	0.002
Model Summary: R² = 0.148, Adjusted R² = 0.117, F(5,138) = 4.774, p < 0.001

Among the neurophysiological predictors, the ATR had a significant positive effect on cognitive performance, (β = 0.206, p = 0.01), indicating that higher ATR levels were associated with improved performance. This finding suggests that an increase in alpha-band power relative to theta-band power, which reflects a more efficient cognitive state characterized by reduced mental distraction and enhanced focus, may contribute to better task performance. This aligns with prior evidence suggesting that a balance favoring alpha activity over theta activity supports cognitive efficiency.

In contrast, neither the ABR nor the TBR showed significant associations with cognitive performance, suggesting that these indicators may not be as sensitive to cognitive task demands under the current experimental conditions. The absence of significance for these predictors could be attributed to their weaker theoretical relevance or potential variability across participants.

Among the affective responses, the relaxation was also a significant predictor (β = 0.173, p = 0.041), suggesting that higher relaxation levels contributed to enhanced cognitive task outcomes. This indicates that participants experiencing higher levels of relaxation tended to perform better on cognitive tasks. However, valence, representing the positivity or negativity of emotional states, did not show a significant relationship with cognitive performance. This lack of significance suggests that emotional positivity alone may not directly enhance cognitive task performance without accompanying relaxation.

Regarding environmental conditions, only the W condition showed a significant positive effect (β = 0.353, p = 0.002), indicating that cognitive performance was notably higher in this environment compared to the reference category (C condition). However, CL (β = 0.013, p = 0.374) and N (β = 0.017, p = 0.314) were not statistically significant, suggesting that these environmental conditions did not have a substantial impact on cognitive performance.

Overall, both ATR and relaxation emerged as significant contributors to cognitive task performance, and environmental factors, particularly the W condition, also played a facilitative role in performance improvement. The positive association of ATR highlights the importance of a balanced neurophysiological state that fosters relaxation while maintaining alertness, with higher alpha power relative to theta power for cognitive efficiency. Similarly, the positive effect of relaxation underscores the role of emotional calmness in optimizing task performance.

Discussion

This study investigated the neurophysiological and affective responses elicited by indoor environments incorporating natural attributes and their subsequent impact on cognitive performance. Among the tested conditions, the wooden interior (W) environment was associated with patterns of neural activity and affective responses indicative of increased relaxation, focused attention, and positive emotional regulation. These patterns coincided with relatively higher cognitive performance in this condition, suggesting a possible link between neurophysiological states and task outcomes.

The W condition was the only environment among those tested that showed statistically significant differences from the control condition (C) across all neurophysiological indices, including the alpha-to-theta ratio (ATR), alpha-to-beta ratio (ABR), and theta-to-beta ratio (TBR). In particular, the W condition exhibited relatively higher ATR values, reflecting the balance between alpha and theta activity, two frequency bands that are closely associated with cognitive functioning⁵⁰. Alpha activity has been linked to states of relaxation and internal focus^51,52, while theta activity is related to external cognitive processing. Reduced frontal theta power has been interpreted as a marker of heightened concentration and lower cognitive load^53,54. In contrast, excessive theta activity may indicate mind-wandering or drowsiness, which can lead to reduced cognitive engagement^55–57. In this context, the observed pattern in the W condition may reflect a neural pattern potentially supportive of cognitive regulation. However, these findings should be interpreted with caution, as they are based on limited comparisons across integrated environmental conditions and do not isolate the effects of individual design attributes.

A high ATR, marked by increased alpha and reduced theta power, indicates a balanced interaction between internal relaxation and external attentional processes⁵⁸, supporting cognitive functioning. In the present study, the elevated ATR observed in the W condition may reflect a distinctive neural pattern involving both internal tranquility and outward attentional readiness. This potential balance between relaxation and alertness appeared in the W condition without requiring deliberate cognitive effort and was less evident in the other experimental conditions (curvilinear forms [CL] and nature views [N]). These findings suggest that the W condition may foster a mental state supportive of cognitive engagement.

Decreased theta activity also has significant cognitive and neurophysiological implications, particularly its association with reduced cognitive load and enhanced cognitive engagement. The TBR metric reflects this dynamic by capturing the interplay between theta waves, indicative of cognitive load, and beta waves, associated with attentional focus. A higher TBR suggests a distracted or inattentive state, while a lower TBR, marked by reduced theta and increased beta activity, corresponds to enhanced focus and cognitive engagement^59,60.

The W condition showed a significantly lower TBR compared to the control, which reflect a neural state indicative of reduced cognitive strain and enhanced sustained attention. In contrast, the CL and N conditions did not show significant differences in TBR relative to the control, suggesting that the attentional effects observed in the W condition were not consistently present across all experimental settings. These findings point to a potentially favorable attentional state in the W condition.

The alpha-to-beta ratio serves as a neuropsychological indicator of emotional balance, reflecting the relative activity between alpha and beta waves. In this study, ABR was significantly higher in the occipital region under the CL and W conditions. Occipital alpha activity tends to increase during calm states requiring minimal conscious attention, such as with closed eyes, and directly reflects improved emotional relaxation and reduced stress⁶¹. In contrast, beta activity often increases during negative emotional states, including stress, anxiety, and tension^62,63. Therefore, a higher ABR—indicating greater alpha power relative to beta power— can reflect a state of emotional equilibrium characterized by relaxation. The elevated ABR observed in the W condition suggests that participants may have experienced a more emotionally relaxed state in this setting compared to the other environments.

Overall, the neurophysiological responses in the W condition, including increased ATR and ABR along with decreased TBR, suggest its potential to support a balanced mental state often described as “alert relaxation.” This state reflects a harmonious interplay between internal and external attentional processes, accompanied by reduced cognitive load. Such a pattern is generally associated with cognitive comfort and sustained focus and a relaxed emotional state with lower levels of stress and tension.

In addition, affective responses and cognitive performance outcomes in the W condition were broadly aligned with these neurophysiological patterns. Participants reported more positive emotional states and greater relaxation, and they showed significantly higher cognitive performance scores in the W condition compared to the C, CL, and N conditions. These findings suggest that the W condition may help facilitate a mental state characterized by internal calmness and outward cognitive engagement, which could contribute to enhanced cognitive performance.

A notable finding was the unexpected lower cognitive performance in the N condition, which contradicted the initial hypothesis. While the N condition, featuring a window view of natural scenery, is typically associated with enhanced psychological stability and attention restoration^9,17, the results revealed a contrasting pattern. Participants in the N condition exhibited reduced occipital ABR compared to both the CL and W conditions, indicating a more stimulated neurophysiological state that demanded increased visual attention⁴¹. This finding aligns with established research on alpha activity in the occipital cortex, which plays a crucial role in visual processing and attentional control. Specifically, occipital alpha activity demonstrates an inverse relationship with external stimulation - decreasing during heightened external visual processing and increasing during tasks requiring internal focus or minimal cognitive effort^64,65. The reduced alpha activity observed in the N condition indicates that participants experienced a more demanding attentional state.

The specific virtual reality (VR) stimuli in the N condition likely influenced this outcome. Unlike the simpler visual designs of the C, CL, and W conditions, the N condition featured a visually rich landscape, including diverse trees and flowers visible through the window. This complex and dynamic scene likely heightened external attentional processing, diverting participants’ focus away from internal cognitive tasks. The resulting heightened visual cognitive load may have disrupted the mental state required for sustained cognitive performance. This interpretation is supported by the absence of significant differences in TBR between the N and C conditions, suggesting that participants did not achieve the level of cognitive focus observed in the W condition. Rather, the visual complexity of the N condition appears to have captured participants’ attention toward external stimuli, potentially compromising the internal cognitive processes essential for optimal task performance.

Neural activity patterns observed in the N condition aligned with participants’ affective responses and cognitive performance. Participants reported feeling positive and slightly aroused emotions during affective evaluations, describing enjoyment and mild excitement. These arousal responses contrasted sharply with the relaxation reported in the W condition. Visual stimuli processed in the occipital region likely triggered these positive and engaging emotions in the N condition.

However, the visual stimulation in the N condition weakened participants’ attentional focus, as evidenced by its lowest cognitive performance scores among all conditions. While brief exposure to indirect natural environments in the N condition elicited pleasant emotions, it failed to sustain the attention restoration and focused cognitive states reported in previous studies^12,13,18,66. These findings suggest that short-term interactions with natural environments can provide emotional benefits but are insufficient to promote the cognitive improvements seen with longer or more immersive exposure.

Moreover, these findings highlight a paradox: although natural environments typically correlate with improved brain regulation and attention restoration, the visually rich stimuli in the N condition may have undermined these benefits. By imposing heightened cognitive demands through external attentional processing, the N condition likely disrupted the balance between relaxation and focus, which is essential for optimal cognitive performance. This finding highlights the importance of carefully balancing visual stimulation and cognitive demands when designing environments to support psychological well-being and cognitive efficiency.

Like the N condition, the CL condition also resulted in lower cognitive performance, with affective responses characterized by low relaxation levels and weak positive emotions. The findings from the CL condition, which included a concentric-elliptical ceiling and rounded furniture, did not show clear cognitive or neurophysiological advantages over other conditions. A possible explanation is that the curved elements, though intended to introduce salient curvilinear features, were spatially detached from the surrounding architecture and may not have been perceived as part of the immediate environment. This likely reduced their immersive quality and diminished their perceptual and emotional impact. Spatially integrated curvilinear features—such as continuous wall contours or ceiling-wall transitions—are more effective in eliciting aesthetic and affective responses⁶⁷. The implementation of curvature in the CL condition may therefore have lacked the spatial coherence needed to produce stronger effects.

On the other hand, the W condition demonstrated the highest cognitive performance alongside the highest relaxation levels. Comparing these results suggests that relaxation, rather than the emotional valence (positive or negative), plays a critical role in influencing cognitive performance.

Regression analysis further validated the complex relationships between neurophysiological and affective responses and their influence on cognitive performance. The results indicated strong associations between cognitive task performance and neurophysiological indicator of ATR, along with the affective indicator of relaxation. These findings underscore the significant role of emotional stability and the balance between alpha and theta activity in supporting effective task execution. A positive relationship was observed between higher ATR, reflecting a relaxed state of internal attention, and improved cognitive performance. Both ATR and relaxation serve as indicators of an enhanced state of relaxation, which supports sustained attention and efficient cognitive processing.

In conclusion, this study highlights the potential role of neurophysiological and affective relaxation induced by nature-integrated indoor design as contributing factors to enhanced cognitive performance. Among the tested conditions, the wooden interior (W) environment was associated with patterns of emotional stability and neural activity indicative of a relaxed yet focused mental state. These observations suggest that incorporating wood into indoor design may help foster psychological well-being and cognitive readiness, supporting sustained attention and mental clarity in certain contexts.

Building on this observation, the calming effects replicated through wooden materials appear to create atmospheres conducive to both emotional ease and cognitive functioning. Wooden interiors that incorporate indirect natural elements may serve as effective design interventions, offering benefits similar to those derived from direct exposure to natural environments. Such design elements hold promise not only for restorative settings but also for cognitively demanding spaces such as libraries or learning environments where sustained focus is essential.

However, while these findings are promising, they should be interpreted within the exploratory scope of this study. The experimental design does not allow for definitive conclusions about the relative superiority of one environment over others, nor does it isolate the effects of individual design features. Rather, the aim was to investigate how specific naturalistic design elements might differentially engage psychological and neurocognitive responses.

Each tested environment exhibited unique patterns. For instance, the nature view (N) condition effectively elicited positive emotional responses, but proved less effective for tasks requiring sustained cognitive focus. The curvilinear (CL) environment showed more subtle effects, neither strongly enhancing nor diminishing performance. These distinctions suggest that different environmental features may serve different psychological and cognitive functions. Natural views and curved forms may be better suited for spaces intended for relaxation and attentional recovery, while wooden interiors may offer advantages in settings demanding prolonged cognitive engagement.

Taken together, these findings support a more nuanced understanding of how diverse natural elements in architectural design can be strategically aligned with specific functional goals—whether relaxation, restoration, or cognitive performance. This underscores the value of intentional design in shaping user experiences and outcomes in indoor environments.

This study explored whether nature-integrated indoor environments are associated with improvements in cognitive performance through their relaxation effects. By examining neurophysiological responses, emotional states, and cognitive outcomes across three representative natural design features—curvilinear forms, nature views, and wooden interiors—compared to a neutral control condition, the study aimed to provide an integrated understanding of how relaxation-related processes may support cognitive functioning.

The findings offer preliminary insights into the potential role of specific natural design elements, particularly wooden interiors, in eliciting favorable neurophysiological and affective responses linked to cognitive performance. These results contribute to the broader discourse on the psychological and cognitive impacts of natural design features in built environments.

However, several limitations should be acknowledged. First, as the sample comprised only Korean college students, the findings may not generalize to broader populations, particularly given that responses to architectural environments can vary across cultural and demographic contexts⁶⁸.

Second, while the study reflected characteristics of real-world architecture, the visual extent of natural elements across conditions was not systematically quantified. The wooden interior used extensive wood finishes, the curvilinear condition applied limited curved elements, and the nature view was confined to one side through a window, resulting in varying levels of visual exposure and spatial immersion. Perceived differences in lighting ambiance across conditions may also have arisen from material-light interactions and outdoor scenery, potentially influencing affective responses⁶⁹. Therefore, the findings suggesting that wooden interiors enhance cognitive and neurophysiological engagement should be interpreted within this experimental context. Systematically quantifying visual design features in immersive virtual environments allows for more precise assessments of their emotional impact⁷⁰. Future research should systematically quantify the visual proportions of spatial elements, allowing for more standardized comparisons and broader investigations into the effects of nature-inspired elements across diverse environmental settings.

Third, While the controlled, time-limited VR-based experimental setup provided a high level of internal validity by isolating specific architectural features, it may not fully reflect the long-term or multisensory nature of real-world indoor experiences. Real environments include additional factors—such as tactile sensations, airflow, ambient temperature, and olfactory cues—that can influence psychological and physiological responses. Moreover, the use of a single, minimally furnished space may limit ecological validity by not capturing the spatial and contextual richness of real settings. Nonetheless, VR remains a valuable tool in architectural research, offering immersive and precisely controlled environments that elicit emotional and cognitive responses comparable to those in physical settings⁷¹. Although VR may not fully capture the complexity of real-world experiences due to some sensory discrepancies, prior VR studies⁷² and the present findings support the reliability of VR-based assessments.

To strengthen these insights, future research should combine VR simulations with real-world testing to more comprehensively examine the effects of natural indoor environments. Future studies should also investigate the cognitive effects of nature-integrated interiors across a wider range of spatial settings and extended exposure durations, while incorporating dynamic environmental factors such as lighting, to enhance ecological relevance.

Methods

This study employed a within-subject experimental design to investigate how interior design elements associated with natural attributes—curvilinear forms (CL), view of nature (N), and wooden interiors (W)—influence cognitive performance. The experiment was conducted in a controlled laboratory environment at a university in South Korea to ensure consistency and minimize external confounds.

All study procedures, including participant recruitment, experimental protocols, and methodologies, adhered to ethical standards and received approval from the Institutional Review Board of Yonsei University (no. 7001988-202209-HR-1687-02). All participants provided written informed consent for the publication of their data in this study. The data will be provided in accordance with institutional guidelines and ethical considerations to ensure participant confidentiality.

Participants

Participants were recruited through advertisements posted on campus bulletin boards and online community platforms at a university, where the experiment took place. All participants were Korean nationals. A total of 40 undergraduate and graduate students voluntarily participated in the study. Data from three participants were excluded due to EEG recording errors or poor data quality, and data from one participant were excluded due to missing cognitive performance data, resulting in a final sample of 36 participants (23 males, 57.5%; 13 females, 32.5%).

To ensure adequate statistical power, a priori power analysis was conducted using G*Power 3.1.9.7 for a repeated measures ANOVA within-subjects design⁷³. The analysis was based on an expected medium effect size (f = 0.25)⁷⁴, an alpha level of 0.05, and a statistical power of 0.95, with four repeated measurements. The results indicated that a minimum of 36 participants was required to achieve sufficient statistical power, confirming that the final sample size was appropriate for the planned analyses.

To ensure the reliability of EEG data, participants met specific eligibility criteria. They were between 19 and 29 years old Korean, in good physical health, free from neurological disorders, and accustomed to virtual reality (VR) viewing without experiencing motion sickness. Participants were required to self-report any history of neurological disorders before enrollment, ensuring that all individuals included in the study had no known neurological conditions. Additionally, all participants had normal or corrected vision, and no restrictions were placed on their academic majors. To optimize experimental conditions, participants were instructed to avoid excessive academic work, strenuous activity, and alcohol the day before the study and to maintain their usual condition on the study day. Each participant received a USD 20 gift card as compensation for their time.

Experimental stimuli

The experimental stimuli comprised four VR-simulated immersive virtual environments: a control condition (C) and three experimental conditions (CL, N, W), illustrated in Fig. 5. The virtual environment were designed to replicate an actual student lounge at the university where the experiment took place, with dimensions of 12 m (W) × 18 m (L) × 3 m (H). The space furnishings included individual tables and chairs, communal tables, sofas, and coffee tables to create a realistic lounge atmosphere.

Fig. 5.

Virtual environmental stimuli used in the experiment.

To enhance ecological validity, a pilot study was conducted with 10 student participants to assess the perceived realism of the VR environments. Based on their feedback, refinements were made to improve visual realism, ensuring that the virtual spaces accurately represented real-world indoor environments.

Each experimental condition was designed to incorporated distinct architectural elements while maintaining consistency in other environmental factors. The CL condition featured curved designs in the walls, ceilings, and furniture, highlighting the potential impact of organic forms on perception and cognitive responses. The N condition included windows with views of natural scenery, designed to isolate the effects of passive exposure to nature through architectural openings rather than the combined influence of indoor greenery. This approach aligns with Ulrich’s previous studies¹⁶ on the restorative benefits of external nature exposure in built environments, allowing for an independent evaluation of the effects of viewing nature through windows. Similarly, the W condition incorporated walls, ceilings, and tables finished with wooden materials, emphasizing the perceptual and neurophysiological impact of extensive timber use in interior design. Although real-world applications of wood elements are often more moderate, the experimental stimuli intentionally maximized wood coverage to enhance the detectability of its effects on cognitive and affective states.

The VR environments were developed using SketchUp 2020 for 3D modeling and Unreal Engine for rendering. To maintain consistency across conditions, all other environmental variables, including furniture arrangement, color schemes, and lighting levels, were kept constant. The environments were rendered as 360-degree videos with identical viewpoints across conditions, ensuring stable EEG recordings by minimizing participant movement while maintaining an immersive viewing experience.

To eliminate potential confounds, lighting was standardized through Unreal Engine’s rendering properties. Identical lighting sources were placed in the same locations across all conditions, with uniform illumination intensity applied to maintain consistent brightness levels. Additionally, visual exposure, sun position, and time settings were kept identical to prevent variations in screen brightness and perceived contrast across conditions.

Although lighting settings were kept consistent, the VR renderings may have appeared to differ slightly in lighting ambiance across conditions due to material-light interactions. Specially, warm-toned wooden surfaces in the W condition may have reflected light in a way that visually enhanced warmth, while natural scenery in the N condition may have altered the perceived color temperature due to light transmission through the virtual window. No additional adjustments were made to equalize the perceived color temperature across conditions. Instead, the focus was on maintaining technical consistency while preserving the ecological validity of each design condition’s material and contextual characteristics.

The VR simulation was executed on a high-performance computer equipped with an Intel Core i7 processor and an NVIDIA GeForce RTX 3070 GPU. The virtual environments were displayed using the Meta Quest 2 VR headset (Meta Platforms, Inc., Menlo Park, CA, USA). The headset features a resolution of 1832 × 1920 pixels per eye, a refresh rate of 90 Hz, and an approximate field of view (FOV) of 100 degrees, providing a high-fidelity immersive experience. To ensure optimal visual clarity and participant comfort, the headset’s interpupillary distance (IPD) was adjusted individually for each participant before the experiment.

Procedures

The experimental procedure, outlined in Fig. 6, began with a briefing on the study’s instructions and cognitive tasks. Participants were fitted with an EEG equipment and a VR head-mounted display, followed by an acclimation period to ensure familiarity with the VR environment. The VR display was adjusted for focal alignment, field of view, and eye level to optimize visual perception and minimize discomfort. Before starting the main experiment, participants closed their eyes and remained in a resting state to allow EEG signals to stabilize.

Fig. 6.

Experimental setup. (A) Experimental procedure, (B) Participant during experiment, (C) Cognitive task in VR.

Each participant first completed the control condition, followed by the three experimental conditions, with the same procedure applied for each. In each condition, participants viewed the VR-simulated environment for three minutes while their EEG was recorded. They were instructed to imagine themselves physically present in the university lounge depicted in the VR space. After viewing each environment, participants took a one-minute break before performing three cognitive tasks in sequence. To ensure familiarity, they completed one practice trial before starting the main tasks. To minimize carryover effects and fatigue, participants rested with their eyes closed for five minutes between conditions before proceeding to the next environment. They repeated this process for all four environments. Following the experimental sessions, participants removed the devices and rested for five minutes before completing a self-report questionnaire that assessed their subjective emotional experiences in each environment. The entire experiment lasted approximately 80 min.

Since this experiment was designed as a within-subject study, each participant was exposed to multiple conditions and performed multiple cognitive tasks. Therefore, appropriately controlling for order and learning effects was crucial. To minimize these effects, four problem sets of equal difficulty (A, B, C, D) were prepared and systematically assigned to each experimental condition. For example, Set A was allocated to the control condition, Set B to the CL condition, Set C to the N condition, and Set D to the W condition. This ensured that participants encountered different but equally challenging tasks in each condition, preventing familiarity with specific problem sets from influencing performance outcomes.

Additionally, to further control for order effects, six different sequences of experimental conditions were created. These sequences were evenly distributed among the initial 40 participants, ensuring a balanced assignment. Participants were randomly assigned to one of these sequences, preventing systematic influences of prior task exposure on subsequent performance. This approach ensured that repeated task performance did not confound the evaluation of environmental influences.

Measurement

Neurophysiological responses

The study used EEG to measure neurophysiological responses while participants viewed each VR environment. EEG signals feature amplitude and oscillatory patterns, with neural oscillations classified into frequency bands that correspond to specific neural activities and mental or cognitive states⁷⁵. This study focused on three primary frequency bands: theta (4–8 Hz), alpha (8–13 Hz), and beta (13–30 Hz), which reflect emotional, mental, and cognitive states.

EEG data were recorded using the Emotiv Epoc X (Emotiv Inc., San Francisco, CA, USA), a wireless 14-channel EEG headset with a sampling rate of 256 Hz and a spatial resolution of 14 active electrodes arranged according to the 10–20 international system. The electrodes were positioned over the antero-frontal (AF3, AF4 F3, F4, F7, F8, FC5, FC6), temporal lobes (T7, T8), parietal lobes (P7, P8), and occipital lobes (O1, O2), with reference channels placed on both mastoids. The device has a built-in notch filter to remove power-line noise at 50 Hz and an internal dynamic noise reduction system to enhance signal quality. The impedance of all electrodes was maintained below 10 kΩ throughout the recording.

Given the susceptibility of EEG data to noise, ensuring high signal quality is critical for obtaining reliable results. Therefore, a multistep EEG preprocessing was conducted in MATLAB using the open-source EEGLAB⁷⁶ toolbox, following the pipeline described by Rodrigues et al.⁷⁷. The preprocessing procedure began with a visual inspection of the raw EEG signals to remove excessively noisy segments. To detect and interpolate bad channels, statistical measures including probability, kurtosis, and spectral frequency were applied, with a z-value threshold set at ± 3.29. Channels exceeding this threshold were classified as outliers and interpolated using spherical spline interpolation to preserve spatial consistency across the dataset. A finite impulse response (FIR) band-pass filter was then applied with a high-pass cutoff of 1 Hz and a low-pass cutoff of 45 Hz to retain relevant neural signal components while eliminating slow drifts and high-frequency noise. This step ensured that the EEG data captured brain activity within the expected physiological frequency ranges.

To further isolate and remove artifacts, independent component analysis (ICA) was conducted to identify sources of noise such as eye blinks, muscle activity, and other physiological or movement-related interferences. Components were manually inspected and classified based on their topographical distribution and spectral characteristics. Components containing less than 80% brain-related activity, as well as those associated with ocular artifacts, muscle movements, or other non-neural sources, were removed.

The cleaned EEG data were transformed into power spectral density (PSD) estimates using Welch’s method⁷⁸. Continuous EEG data were segmented into one-second epochs, and a Fast Fourier Transform (FFT) with a Hamming window was applied to each epoch. PSD values were then computed for each frequency band and channel across all conditions. To ensure data integrity and minimize the influence of extreme values, PSD values were screened for outliers, with data points exceeding a Z-score threshold of ± 3 identified and removed.

Following this preprocessing, the PSD values obtained from EEG data were used to compute three key neurophysiological indicators: alpha-to-theta ratio (ATR), theta-to-beta ratio (TBR), and alpha-to-beta ratio (ABR). These metrics were selected based on their well-established relevance in cognitive neuroscience and psychophysiology, particularly in assessing attention engagement, cognitive control, and relaxation responses.

ATR, the ratio of alpha activity to theta activity, reflects attentional focus and cognitive engagement^50,79. Higher ATR indicate greater alpha power relative to theta power, reflecting increased internal attention and an optimal balance between relaxation and alertness⁸⁰, which is beneficial for cognitive efficiency. ATR was calculated as the mean alpha-to-theta ratio from frontal electrodes (AF3, F7, F3, F4, F8, AF4), as theta and alpha activity in these regions are strongly linked to attentional processes^59,81. TBR, the ratio of theta activity to beta activity, serves as a neural marker of attentional control and cognitive distraction^59,60,82. Lower TBR indicate reduced theta power relative to beta power, reflecting decreased cognitive distraction and enhanced sustained attention, making them particularly useful for assessing cognitive attention states. Like ATR, TBR was calculated as the mean theta-to-beta ratio from frontal electrodes, given their critical role in executive function and cognitive regulation⁸³. ABR, the ratio of alpha activity to beta activity, evaluates mental relaxation, stress regulation, and arousal levels⁸⁴. Higher ABR indicate increased alpha power relative to beta power, reflecting a more relaxed neurophysiological state, making it a key measure for assessing the calming effects of indoor environmental conditions. ABR was computed from occipital electrodes (O1, O2), as alpha activity is predominantly observed in the occipital region where it is strongly associated with relaxation and visual processing⁶⁵.

ATR, TBR, and ABR for each condition were statistically analyzed to examine how neural patterns associated with attentional and relaxation states change across various natural indoor environments.

Affective responses

To investigate how participants’ affective responses to natural indoor environment influenced their cognitive performance, a self-reported survey was conducted to assess their subjective experiences in the VR settings. The affective self-report scales were selected to capture emotional experiences related to relaxation and environmental perception. The survey employed the Circumplex Model of Affect, developed by Russell^85,86, a widely recognized framework for evaluating emotional responses to environmental stimuli and situations. This model categorizes emotions along two core dimensions: relaxation (relaxation–arousal) and valence (positive–negative).

Relaxation represents the level of activation or stimulation in emotional responses, such as how calm or energized a participant feels in a given environment. Relaxation was assessed using word pairs like “excited” and “calm.” Valence, on the other hand, measures the positivity or negativity of emotional responses, indicating how happy or unhappy participants felt about a particular environment, with word pairs like “happy” and “unhappy.”

Participants rated their feelings for each condition using items composed of five semantic differential word pairs for each dimension, with each scored on a ten-point scale (arousal: 0 = highly aroused, 10 = fully relaxed; valence: 0 = highly negative, 10 = highly positive). The average scores for relaxation and valence were calculated to evaluate participants’ affective responses to each condition.

Additionally, To analyze affective patterns, a two-dimensional scatter plot was constructed with valence on the horizontal axis and relaxation on the vertical axis. Data points for each experimental condition were plotted relative to the mean score of the control condition, and their distribution was analyzed to determine the quadrant with the most prominent clustering. The first quadrant represents positive-relaxed states, the second quadrant negative-relaxed states, the third quadrant negative-aroused states, and the fourth quadrant positive-aroused states.

Cognitive performance

The cognitive tasks were selected to examine whether neurophysiological and affective responses to natural indoor environments translate into measurable cognitive performance outcomes. These tasks assess executive function, attentional control, and inhibitory processing, which are cognitive domains known to be influenced by architectural environments. Based on prior research investigating the effects of architectural environments on task performance^45,87, the study employed three validated cognitive tasks: Error Detection Task, Stroop Test, and Go/No-Go Test.

To systematically assess differences in cognitive performance across environments, participants completed these tasks within the VR setting, where they were presented on a virtual screen, as shown in Fig. 6. Participants interacted with the tasks using a virtual keyboard operated through VR controllers.

Four problem sets of equal difficulty (A, B, C, D) were prepared for each task and assigned to the control and experimental conditions. Additionally, participants were randomly assigned to different sequences of condition exposure to prevent learning effects from prior tasks, ensuring that repeated task performance did not contaminate the evaluation of environmental influences. The Stroop Test and Go/No-Go Test were implemented using PsyToolkit, a widely used platform for cognitive-psychological experiments.

The first task, the Error Detection Task, assessed focus and concentration⁸⁷. The researchers presented participants with a passage containing ten intentional spelling errors extracted from reading passages from Korean college entrance exams. The participants read the passage, identified errors, and reported them verbally. Response times were recorded to evaluate reading speed and accuracy, calculating task score as a composite of five metrics: accuracy (correctly identified errors divided by total words), precision (correctly identified errors divided by total identified errors), recall (correctly identified errors divided by total actual errors), F1 score (2 × precision × recall / (precision + recall)), and standardized response time. Each metric received an equal weight (20%) and normalized using min-max scaling.

The second task, the Stroop Test, measured attention and executive function⁸⁸. Participants were shown words in four colors (red, yellow, green, blue), with some words’ meanings matching their displayed color and others not. For example, if the screen displayed the word “red” in the color green, participants had to select “green.” Performance was evaluated using the Stroop Effect, calculated as the difference between response times for correct trials in incongruent and congruent conditions⁸⁹. The Stroop Effect scores were normalized using min-max scaling for comparisons across environments.

The third task, the Go/No-Go Test, assessed inhibitory control and attention⁹⁰. Participants were instructed to respond only to “Go” stimuli and refrain from responding to “No-Go” stimuli. Performance was evaluated using three metrics: mean Go reaction time, Go error rate, and No-Go error rate. These metrics were averaged and normalized using min-max scaling.

The overall cognitive performance score for each environment was the average of the normalized scores from the three tasks.

Statistical analysis

To determine statistically significant differences between the control condition (C) and experimental conditions (CL, N, W) across neurophysiological responses, affective responses, and cognitive performance, repeated measures ANOVA (RM ANOVA) was conducted for each measure. Environmental conditions served as the independent variable, while the dependent variables were the measured indices: ATR, TBR, and ABR for neurophysiological responses, the mean scores of relaxation and valence for affective responses, and the mean cognitive task score for cognitive performance. For RM ANOVA, the Greenhouse–Geisser correction was applied to a violated assumption of sphericity. Effect sizes were evaluated using partial eta squared (η²_p) and interpreted as small effect (η²_p ≥ 0.01), medium effect (η²_p ≥ 0.06), and large effect (η²_p ≥ 0.14)⁷⁴. Post hoc comparisons with Bonferroni correction were used to identify specific statistically significant differences between conditions.

Additionally, a regression analysis was conducted to examine how the neurophysiological responses (ATR, ABR, TBR) and affective responses (relaxation, valence) influenced cognitive task performance under different environmental conditions (C, CL, N, W). The cognitive task score served as the dependent variable, while neurophysiological and affective indicators were the independent variables. Environmental conditions were included in the regression model as dummy-coded variables, with C condition set as the reference category. The stepwise selection method was applied to refine the model, and multicollinearity among variables was examined using Variance Inflation Factor (VIF) values to ensure that predictor variables did not exhibit high intercorrelations. The final model was evaluated based on estimated regression coefficients and their statistical significance. Model fit indices, including the coefficient of determination (R²) and the adjusted R², were examined to assess the explanatory power of the model. This regression analysis identified key neurophysiological and affective factors that influenced cognitive performance and provided a comprehensive assessment of how each environmental condition contributed to performance improvement.

All statistical analyses were performed using the R programming language (version 4.3.1; R Core Team, 2023)⁹¹, with the afex package for repeated measures ANOVA and the lme4 package for regression modeling. Statistical significance was set at p < 0.05 for all analyses.

Author contributions

S.K. conceptualized the research, designed the methodology, conceived and conducted the experiment, curated and analyzed the EEG data, generated figures and tables, visualized the results, wrote the original manuscript and contributed to the review and editing process.

Funding

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education. (RS-2023-00239734).

Data availability

The datasets generated and analyzed during the current study, excluding sensitive information that could identify individuals, are available upon reasonable request from the corresponding author. The data will be provided in accordance with institutional guidelines and ethical considerations to ensure participant confidentiality.

Competing interests

The authors declare no competing interests.