Nature in Indoor Environments: An EEG Study Comparing Natural Greenery with Nature-Derived Visual Alternatives

Abstract

Integrating greenery into built environments plays a significant role in enhancing indoor space, nurturing mental health, and reinforcing overall well-being. Yet, many indoor settings lack direct access to natural greenery, raising questions whether nature-derived visual alternatives, such as chromatic properties or visual complexity could evoke neural responses comparable to those induced by natural greenery. Four indoor window-view stimuli were presented: White Color (control), Green Color, Natural Greenery, and Shadows of Greenery. We used electroencephalography (EEG) to monitor brain electrical activity, with a particular focus on frontal alpha asymmetry (FAA) as an indicator of approach/withdrawal motivation and affective engagement. Natural Greenery showed the strongest asymmetry, with the most negative FAA values. Natural Greenery was also rated highest in comfort and likability. Additionally, a two-way repeated measures ANOVA indicated significantly greater right-frontal engagement in the Natural Greenery view. No comparable neural response was observed for Green Color or Shadows of Greenery.

1. Introduction

Architectural research has increasingly emphasized concerns for occupants’ health and well-being in indoor environments [1,2,3,4]. Consequently, integrating greenery as a natural element has been one way to improve indoor environmental quality and support mental health outcomes [5,6]. The innate connection of humans with nature, known as “biophilia,” was popularized by Edward O. Wilson in 1984 [7], who argued that this connection might be attributed to the genetic past of human evolution in the natural environment. Wilson relates this human tendency to connect with nature as a biological need integral to the physical and mental development of the human species. Even before “biophilia” was defined as a term, researchers explored related ideas about how natural settings influence cognitive development, esthetic preferences, behavior, and even physical healing [8].

1.1. Restorative Environments and Attention Restoration Theory

When William James [9] distinguished between two types of attention (involuntary and directed attention), he described certain natural elements, such as moving things, wild animals, and blows, as examples of involuntary attention. Similarly, Attention Restoration Theory (ART), originally developed by Kaplan, extensively explores the restorative quality of nature [10]. ART contends that nature captures our attention effortlessly and serves as the best resource for resting directed attention, reducing fatigue, and restoring mental function and effectiveness [10]. Kaplan and Kaplan [11] identified four essential components of restorative environments, including fascination. They suggest that fascinations with nature, such as clouds, sunsets, and rustling leaves, can be called “soft fascinations” and allow for a reflective mode. A stimulus that elicits fascination draws involuntary attention, making it critical for a restorative experience—it attracts people, keeps them from boredom, and allows them to function without directed attention.

1.2. Visual Access to Nature in Indoor Settings

Originally a theoretical framework, biophilia has evolved into a foundational concept in design discourse. In particular, the emerging paradigm of biophilic design translates biophilic principles into strategies for shaping the built environment. Certification frameworks, such as LEED [12] and WELL standards [13], integrate biophilic principles to enhance environmental quality and human health. These frameworks emphasize the role of daylight, greenery, and natural materials as a critical strategy for creating biophilic design.

Research has increasingly examined the restorative impact of visual access to nature in indoor environments [14]. Studies associate access to greenery views in the indoor environment with higher relaxation states [15], improved well-being [16], reduced stress [17], enhanced cognitive performance [18], and improved sleep and mood regulation [19,20]. During the COVID-19 pandemic, green window views from within the home have been associated with increased life satisfaction, subjective happiness, self-esteem [21], and reduced depressive/anxiety symptoms [22]. Moreover, biophilic environments have been shown to facilitate faster recovery from stress compared to non-biophilic settings [23]. However, many of these previous studies have relied on subjective questionnaires as a general assessment approach. The subjective nature of such reports can limit the researchers to the subjective perception of occupants’ satisfaction [24,25]. Consequently, an unbiased and objective assessment is important to maximize its benefits and improve design practices. To address this limitation, architectural research has increasingly incorporated neuroscience to deepen the objective assessment of these interactions.

1.3. Neuroarchitectural Approaches and EEG-Based Assessment

Neuroarchitecture portrays a new frontier in architecture, lying between the interrelation of neuroscience and architecture. A key area of this field is examining how architectural elements influence cognitive and emotional states by drawing on insights from cognitive neuroscience. Several authors [26,27,28,29] have highlighted that the value of collaboration between architecture and neuroscience lies in generating empirical evidence as a core component of the research and design process.

Electroencephalography (EEG), a well-established neuroscientific method for measuring brain activity owing to its high temporal resolution and non-intrusive nature, has been widely used in neuroarchitectural studies [30]. EEG measures electrical activity in the brain through voltage fluctuations across regions [31,32]. These signals are categorized into frequency bands: delta, theta, alpha, beta, and gamma—each linked to different states of awareness [33]. Notably, alpha waves (8–13 Hz) have been linked to relaxation and stress reduction [34,35,36,37]; higher frontal lobe alpha activity is considered a sign of cognitive recovery, and a positive correlation to emotional states and motivation, particularly when subjects had views towards greenery [38,39,40,41,42,43,44,45].

1.4. Frontal Alpha Asymmetry (FAA) in Nature Processing

In the wake of research on the relationship between the frontal lobe and alpha waves, a specific interest has emerged in the phenomenon of asymmetric modulation of alpha waves in the prefrontal cortex, known as cerebral asymmetry or frontal alpha asymmetry (FAA). FAA reflects variations in alpha activity between the right and left hemispheres and serves as a neural marker of emotional processing, particularly in relation to individual differences in positive and negative affect, and motivational tendencies [46]. Davidson et al. [47] presented evidence that the difference in frontal alpha hemispheres represents the approach and withdrawal motivation. They found that greater relative right frontal activation (higher left alpha frequency) was significantly associated with negative responses and withdrawal. In contrast, left frontal activation was associated with approach motivation and positive responses (Figure 1). Subsequent studies have supported this model [45,46,48,49,50], including research on exposure to greenery [38]. Other evidence, however, has reported conflicting findings, indicating that exposure to greenery views increases right frontal activation [51,52]. This suggests an engagement of involuntary attentional processes rather than solely eliciting emotional effects [45,53,54,55,56]. These contradictory FAA results, which support both affective and attentional interpretations, highlight a gap to which the present study contributes additional evidence by examining neural responses to different forms of greenery in architectural environments.

1.5. Do Nature-Derived Visual Features Elicit Similar Neural Responses?

We spend more than 90% of our time in indoor environments [57,58,59]. As a result, there is a growing effort to elucidate the relationship between environmental stimuli and neural activity to maximize the effective use of greenery. In settings where natural greenery is unavailable, it is important to explore potential substitutes that might evoke similar restorative effects. In some cases, architects and interior designers use green hues or greenery patterns to replicate natural greenery. Therefore, an investigation into different types of greenery representation is crucial for expanding the knowledge on greenery integration. However, until now, FAA has been mainly analyzed in studies comparing natural and non-natural settings (e.g., urban landscapes vs. natural greenery) [34,38,54,60,61,62,63,64]. To close this gap, this study examines whether greenery alternatives, such as color and pattern, can elicit similar neural responses to those of natural greenery, particularly in environments where real plants or natural views are limited.

A common experience in nature is the irregular patterns of light and shadow created when sunlight filters through a tree canopy, known as dappled light. Even in the absence of a direct view of nature, these dynamic light patterns serve as subtle reminders of natural surroundings. In this study, this effect is referred to as “Shadows of Greenery”. This phenomenon relates to fractal geometry, explored by mathematician Benoit Mandelbrot, who observed such patterns in nature (including tree branches, cloud formation, and coastlines) [65]. The fractal dimension (D) describes the scaling relationship of these structures by measuring their repetition at different scales. A D value near one indicates a smooth, sparse pattern, while a D value closer to two represents a highly intricate, detailed structure [66]. Research has shown that both natural and mathematical fractals can influence esthetic preferences and human brain dynamics [66]. Furthermore, fractals can induce alpha responses, indicating a state of relaxation [67]. Augustin S. [68] suggests that a preference for such patterns may be an adaptive trait that signals favorable environmental conditions, such as the presence of nearby trees that can serve as shelter. More recent research finds that shadows of greenery projected onto room surfaces through a window, particularly those with medium to high complexity, were significantly more preferred and induced higher levels of relaxation compared to simpler patterns [69].

Another commonly used alternative to natural greenery is the color green, often applied through green-colored walls or other forms of interior design elements intended to evoke associations with nature through color. However, limited research has been completed on the effects of green color on brain activity indoors. One study found that the color green was associated with higher levels of alpha and theta bandwidth [70], while another study that investigated the effect of perceiving color in an indoor environment found that a green room was the only room that had a depression score of 0.00 [71].

As discussed above, existing research suggests that fractal patterns and green color can increase relaxation and preference for indoor environments. However, it remains unclear whether these substitutes from nature activate neural processes comparable to those elicited by actual natural greenery. This question is directly relevant to architectural design because these substitutes are often used where indoor environments lack views of natural greenery. To the best of our knowledge, no previous study has directly compared the neural responses to Natural Greenery, Shades of Greenery, and Green Color using FAA, a neurophysiological marker of approach-withdrawal motivation. While earlier research primarily examined the broader relationship between nature and urban environments, the present study directly compares neural responses to natural greenery and commonly used design substitutes (green color and shadow patterns), providing a neurophysiological perspective on how different visual representations of greenery are processed in indoor environments. We hypothesize that Natural Greenery will elicit the strongest FAA response. Then we examine whether Shadows of Greenery and Green Color produce FAA patterns that are relative to natural greenery or remain distinguishable from it. Together, these analyses extend FAA research by examining nuanced representations of greenery, offering insights that may inform future biophilic design strategies. The conceptual research framework and analytical workflow of the study are illustrated in Figure 2.

2. Materials and Methods

2.1. Participants

In this study, 25 healthy volunteers (mean age ± SD 32.88 ± 8.68 years; 13 female) participated in all experimental conditions. All participants reported normal or corrected-to-normal vision and no history of neurological or psychiatric disorders. Participants who normally used corrective lenses were allowed to wear them during the experiment. Participants were recruited from the university community. Written informed consent was obtained from all participants, and all data were anonymized and protected in accordance with the Code of Ethics. Before the screening, we informed participants that the research involved investigating esthetic judgments and measuring brain waves, but we did not mention the experimental goals. The study was approved by the Ethics Committee of the University of Tokyo (approval date: 13 May 2021; reference number: 21-17).

Power analysis was conducted using G*Power 3.1 [72]. For the one-way repeated measures ANOVA with four conditions, assuming a medium effect size (f = 0.3), an alpha level of 0.05, a correlation among repeated measures of 0.5, and a non-sphericity correction ε = 0.75, the achieved power with this sample was 0.89. For the two-way repeated measures ANOVA with factors Trigger (four levels) and Hemisphere (two levels), under the same assumption, the achieved power was 0.99. Therefore, the study was adequately powered to detect medium-sized effects, although nonsignificant results should be interpreted with appropriate caution.

2.2. Stimuli Set

Four types of visual stimuli were prepared to systematically examine the effects of color, natural elements, and light-shadow patterns.

White Color (Control): The White Control condition consisted of a plain white background in the window view, eliminating visual input related to color, natural elements, or shadow patterns. This neutral baseline was created to ensure that the differences observed were due to the stimuli themselves, not to external visual factors.

Green Color: The green color used in the Green Color stimulus was sampled from the Natural Greenery image. To quantitatively characterize the chromatic properties of the stimulus, mean RGB values were extracted from a centrally defined region of interest (ROI; 500 × 500 pixels), covering approximately 47% of the image area, within the Green Color stimulus using Fiji/ImageJ (ImageJ version 1.53c; https://imagej.net/software/fiji/). The resulting mean color values were R = 131, G = 147, and B = 83 (8-bit scale).

Natural Greenery: We selected a high-resolution photograph depicting a natural setting with trees to represent a direct view of natural greenery. This provided a reference for an ecological visual input, whereas the image remained static throughout the experiment.

Shadows of Greenery: We used the same natural greenery image to capture the dappled-light effect and processed it using Fiji/ImageJ. We first converted the image to 8-bit grayscale. Then, we binarized it using the threshold function to increase contrast between light and shadow. Lastly, we applied fractal box counting, yielding a fractal dimension of D = 1.8. This dimension represents a medium-to-high fractal complexity. Similar fractal dimensions (approaching D ≈ 1.9) have been used to represent highly detailed tree and vegetation branching patterns in experimental studies of fractal perception [66]. Moreover, as recent research found that shadows of greenery projected onto room surfaces through a window, particularly those with medium to high complexity, were significantly more preferred and induced higher levels of relaxation compared to simpler patterns [69], the choice for this dimension was intended to approximate naturally occurring shadow patterns of vegetation. Since the binarization process produced a high-contrast black-and-white image, the opacity of the shadow regions was reduced to achieve a more neutral grayscale image. This helped reach a balanced perception of the stimulus without excessive contrast effects. Like the natural greener condition, this stimulus remained static throughout the experiment. The stimuli set is illustrated in Figure 3.

For this study, we designed 24 room variations with views toward the windows. The overall dimensions of the rooms remained constant while the window shapes and placements were altered. All four stimuli (White Color (control), Green Color, Natural Greenery, and Shadows of Greenery) were presented in every room variation, resulting in 96 final variations. Figure 4 shows all 24 room variations and the use of all four stimulus conditions within a single room variation.

The purpose of testing the stimuli under multiple layouts was not to compare these window configurations as independent variables, but rather to confirm that the observed effects of the stimuli were not dependent on a single room configuration. We left the rooms intentionally non-materialized so that interior design features would not impose additional effects beyond the window views themselves. While some influence of room layout cannot be entirely ruled out, the study was designed to maintain a clear and focused investigation on greenery-related visual stimuli.

2.3. Experimental Design and Data Collection

Data collection took place in a laboratory-controlled environment on the premises of the University of Tokyo. Participants were instructed to passively sit on a chair placed 1 m away from the viewing screen and follow the screen instructions (Figure 5).

For this experiment we used a within-subject block design. We divided the critical trials into eight blocks of 12 stimuli, each beginning with a 10 s baseline. After each block, participants were encouraged to rest before starting the next at their own pace. Each participant completed 96 trials in total (8 blocks × 12 trials). To control the sequence, we used E-Prime 3.0 (Psychology Software Tools, Inc., Pittsburgh, PA, USA), which delivered the visual stimuli and automatically recorded the onset time of each segment.

The four stimulus conditions (White Color, Green Color, Natural Greenery, and Shadows of Greenery) were presented in a randomized order within each block. EEG stimulus presentation durations vary across studies, depending on research objectives and methodological constraints. Previous research shows that frontal alpha asymmetry (FAA) differences can emerge within milliseconds (300 ms) [73]. Given that our study focused on neural responses, rather than sustained emotional states, we chose an 8 s EEG stimulus duration. The stimulus duration was selected to be sufficiently long to allow stimulus processing while remaining short enough to minimize common EEG artifacts, such as eye blinks and attentional drift. The 8 s viewing period was followed by a 1 s fixation cross (Figure 6).

First, we asked participants to carefully read and sign the informed consent paperwork. Then, we equipped them with the EEG cap and gave them a short period of time to adjust to the equipment. A brief test recording was conducted to verify signal quality, and to familiarize participants with the experimental procedure. After that, we informed them of the tasks they would perform during the experimental session.

After viewing all conditions, participants rated each condition on comfort and likability using a 7 point semantic differential scale. Subjects were asked to judge the opposites presented by scores of +3 to −3, ranked in the following order: Very (+3/−3), Moderately (+2/−2), Slightly (+1/−1), Neither (0).

The four indoor environmental conditions (White Color, Green Color, Natural Greenery, and Shadows of Greenery) were grouped and analyzed according to their assigned visual trigger, independent of room variation, which was kept consistent across conditions.

2.4. EEG Acquisition and Preprocessing

Electroencephalography (EEG) acquisition and reporting followed current community recommendations for EEG/MEG methods reporting, including explicit specification of sensor layout, reference/ground, sampling rate, and filter settings [74]. EEG was recorder in a controlled laboratory environment using g.Nautilus Research wireless EEG headset (g.tec medical engineering GmbH, Schiedlberg, Austria), equipped with eight dry g.Sahara electrodes, positioned according to the international 10–20 system for electrode placement (Fz, F3, F4, Cz, Pz, P4, P5, Oz). During acquisition, signals were referenced to the right-ear clip and grounded using the ear-mounted ground electrode, consistent with the g.Nautilus acquisition configuration that uses dedicated mastoid/ear reference and ground clips [75]. This reference was maintained throughout the entire preprocessing pipeline, as the reference signal was subtracted at each time point during acquisition; thus, no additional re-referencing was applied during analysis, in accordance with g.tec’s acquisition protocol. EEG data were sampled at 250 Hz, which is a typical hardware-supported sampling rate for the g.Nautilus platform and provides sufficient bandwidth for analysis focused on frequencies below 50 Hz. Online filtering included a 0.5–50 Hz bandpass and a 50 Hz line-noise notch (implemented as a 48–52 Hz bandstop), consistent with g.Nautilus example pipelines for 50 Hz power-line environments, such as Eastern Japan (Tokyo), where the experiment was conducted. Data were then exported for analysis using g.tec’s tools and MATLAB (Version R2024a, MathWorks, Natick, MA, USA). EEG signals were amplified by a g.Nautilus amplifier.

Preprocessing decisions were defined a priori and reported in accordance with EEG/MEG reporting recommendations, including artifact criteria and documentation of rejected data [74,76]. Trials were screened for ocular and myogenic contamination using a two-stage procedure: (1) automated amplitude-based flagging and (2) visual confirmation. Amplitude-based flagging used a peak-to-peak threshold of ±75 µV, a commonly used criterion in EEG epoch screening and consistent with published examples for threshold-based artifact rejection.

To ensure data quality, flagged trials were then visually inspected for confirmation following established EEG preprocessing guidelines [77]. Trials containing clear artifacts were excluded for further analysis. Because artifact prevalence can differ across experimental conditions [74], the number of retained and rejected trials was tracked per participant and condition.

In line with common EEG preprocessing practice [77], participant-specific adjustments were applied in cases of unusually high EEG amplitude variance. For four participants showing consistently high alpha amplitude without behavioral or visual evidence of artifact contamination, the amplitude threshold was adjusted to ±100 µV. This adjustment reduced unnecessary rejection of physiologically plausible high-amplitude alpha activity while maintaining artifact control.

2.5. Alpha Power Extraction and FAA Calculation

Alpha power was obtained using a FFT-based bandpass filter between 7 and 13 Hz implemented in the g.Nautilus processing pipeline (g.tec medical engineering GmbH, Schiedlberg, Austria). The software computes spectral power over 0.5 s sliding windows (125 samples) with a 124-sample overlap. For each participant stimulus condition, all retained 8 s trials were averaged to yield a single mean alpha power value (in µV²) per electrode (F3 and F4). The entire 8 s segment was treated as one analysis unit; no further temporal subdivision or segmentation was applied within the epoch. Thus, one alpha power value per electrode was obtained per participant per condition.

The extracted alpha power values were transformed by calculating the natural logarithm (ln) of µV² values. Frontal alpha asymmetry (FAA) was calculated as the difference between the natural log of alpha power at the right frontal site (F4) and the left frontal site (F3), i.e., FAA = ln(F4) − ln(F3). This log-transformed metric, essentially the log ratio of right to left alpha power, is a widely used and validated approach in EEG research to quantify hemispheric asymmetry [47,48,50,78].

We interpret FAA as an index of relative hemispheric differences in frontal alpha-band power during stimulus viewing, computed using the log-difference metric that is commonly used in FAA methodology [79].

In the present study, FAA serves as a comparative neural index across within-subject conditions rather than as a biomarker of attention or affect. Although FAA has been linked to motivational and affective processes in some literature, ongoing theoretical and methodological debates persist, and conclusions are sensitive to analytic and referencing choices.

2.6. Statistical Analysis

We used IBM SPSS Statistics (v 29.0) to perform all the analyses. Given the within-subject experimental design and the comparison across multiple stimulus conditions, repeated measures ANOVA was selected as an appropriate statistical approach to account for correlations among repeated observations within the same participants. For the EEG, frontal alpha asymmetry (FAA = ln F4 − ln F3) was first examined with a one-way repeated measures ANOVA across the four window-view triggers (White Color (control), Green Color, Natural Greenery, and Shadows of Greenery). To maximize statistical power, a complementary two-way repeated measures ANOVA was then run with Trigger (four levels: White, Green Color, Natural Greenery, Shadows of Greenery) and Hemisphere (two levels: ln F4, ln F3) as within-subject factors. When Mauchly’s test indicated a violation of sphericity, we applied Greenhouse–Geisser corrections to the F-tests and Bonferroni adjustments to all pairwise contrasts.

For the semantic-differential questionnaire, comfort and likeability ratings were analyzed in separate one-way repeated measures ANOVAs across the same four triggers. We used identical assumption checks and Bonferroni-corrected post hoc comparisons. All tests were two-tailed, with α = 0.05 after the relevant corrections, and descriptive statistics are reported as mean ± SD. The overall EEG acquisition and analysis workflow is illustrated in Figure 7.

3. Results

3.1. Frontal Alpha Asymmetry (FAA): One-Way Repeated Measures ANOVA

A one-way repeated measures ANOVA was first conducted to test whether there were statistically significant differences in frontal alpha asymmetry scores (FAA = ln(F4) − ln(F3)) across the four visual conditions: White, Green Color, Natural Greenery, and Shadows of Greenery. The analysis revealed a significant main effect of condition, F(2.124, 50.968) = 4.591, p = 0.013, partial η² = 0.161, indicating variation in FAA across stimuli. Normality was met for most conditions, with mild deviations observed in two measures (lnF3 for Green Color and Shadows of Greenery). Sphericity violations were corrected using the Greenhouse–Geisser adjustment.

Although Bonferroni-adjusted pairwise comparisons did not reach conventional significance (lowest adjusted p = 0.062 for Natural Greenery vs. Green Color), the Natural Greenery condition consistently exhibited the most negative FAA values, suggesting a trend toward increased right-frontal activation (i.e., a right-dominant FAA score), as shown in Figure 8.

To further explore lateralized activity, a two-way repeated measures ANOVA including Hemisphere (left vs. right) as a within-subject factor was conducted. Descriptive statistics for each condition are presented in

Table 1. Descriptive statistics for frontal alpha asymmetry (FAA = ln(F4) − ln(F3)) across visual conditions (N = 25).

Condition	Mean FAA	SD
White Color (Control)	−0.168	0.321
Green Color	−0.063	0.386
Natural Greenery	−0.266	0.379
Shadows of Greenery	−0.174	0.324

.

3.2. Frontal Alpha Asymmetry (FAA): Two-Way Repeated Measures ANOVA

To examine hemispheric differences in alpha activity across visual conditions, a two-way repeated measures ANOVA was conducted with factors Trigger (White, Green Color, Natural Greenery, Shadows of Greenery) and Hemisphere (F3, F4) on log-transformed frontal alpha power. Normality assumptions were met for most conditions, with minor deviations in two. Sphericity violations were corrected using the Greenhouse–Geisser method.

The analysis revealed a significant interaction between Trigger and Hemisphere, F(2.124, 50.968) = 4.59, p = 0.013, partial η² = 0.161, indicating that the effect of visual stimuli varied by hemisphere. This prompted further examination of simple effects within each condition.

In the Natural Greenery condition, right frontal alpha power (F4: 2.818 ± 0.275 ln µV²) was significantly lower than left (F3: 3.084 ± 0.465), F(1, 24) = 12.29, p = 0.002, partial η² = 0.339, consistent with increased right-hemisphere activation—a neural marker linked to approach/motivation. We observed a similar pattern in the Shadows of Greenery condition, F(1, 24) = 7.20, p = 0.013, though it did not meet Bonferroni-adjusted significance (α = 0.0125). The White (control) condition also showed right-dominant asymmetry, F(1, 24) = 6.80, p = 0.015, but again fell short of corrected thresholds. No significant hemispheric difference was observed in the Green Color condition, F(1, 24) = 0.66, p = 0.424, partial η² = 0.027 (Figure 9). Frontal alpha asymmetry (FAA) analysis is shown in Figure 10. Descriptive statistics for each condition are presented in

Table 2. Descriptive statistics for log-transformed frontal alpha power.

Condition (Trigger)	F4 (Mean ± SD)	F3 (Mean ± SD)
White Color (Control)	2.905 ± 0.376	3.072 ± 0.436
Green Color	2.904 ± 0.426	2.967 ± 0.393
Natural Greenery	2.818 ± 0.275	3.084 ± 0.465
Shadows of Greenery	2.833 ± 0.383	3.006 ± 0.439

Note: values represent mean ± standard deviation of log-transformed frontal alpha power recorded at right (F4) and left (F3) frontal electrodes across the four visual conditions (n = 25).

, and the results of the two-way repeated measures ANOVA are presented in

Table 3. Two-way repeated measures ANOVA for log-transformed frontal alpha power.

Effect	df	F	p	Partial η2
Trigger	1.557, 37.36	0.693	0.472	0.028
Hemisphere	1, 24	7.224	0.013	0.231
Trigger × Hemisphere	2.124, 50.968	4.591	0.013	0.161

Note. Greenhouse–Geisser correction was applied where the sphericity assumption was violated.

.

3.3. Subjective Ratings: Semantic Differential Questionnaires

One-way repeated measures ANOVA was used to assess scores of the semantic differential questionnaire for likeability and comfort ratings across the four window-view conditions (White Color, Green Color, Natural Greenery, and Shadows of Greenery).

Likeability: Natural Greenery received the highest likeability ratings, showing a significant difference between both the Green Color and White conditions. Mauchly’s test indicated that the assumption of sphericity was violated (W = 0.585, χ²(5) = 12.18, p = 0.033); Greenhouse–Geisser correction was therefore applied (ε = 0.741). The effect of Trigger was highly significant, F(2.22, 53.25) = 43.60, p < 0.001, partial η² = 0.645. Mean ratings (± SD) were 1.08 ± 1.38 for White, −1.08 ± 1.73 for Green Color, 2.48 ± 0.87 for Natural Greenery, and 2.28 ± 0.94 for Shadows of Greenery. Bonferroni-adjusted pairwise contrasts showed that Green Color was rated less likable than every other view (p < 0.001 in all cases), and Natural Greenery exceeded both White (p = 0.001) and Green Color (p < 0.001). In contrast, Shadows of Greenery differed from Green Color (p < 0.001) but did not differ from Natural Greenery or White (both p = 1.000).

Comfort: Comfort ratings followed a similar pattern, with Natural Greenery and Shadows of Greenery rated significantly more comfortable than Green Color. Mauchly’s test indicated that the assumption of sphericity was not violated (W = 0.779, χ²(5) = 5.68, p = 0.339). The ANOVA revealed a robust main effect of Trigger, F(3, 72) = 37.71, p < 0.001, partial η² = 0.611. Mean comfort scores were 0.96 ± 1.40 (White), −0.88 ± 1.72 (Green Color), 2.36 ± 0.95 (Natural Greenery), and 2.00 ± 0.91 (Shadows of Greenery). Bonferroni comparisons indicated that Green Color was judged significantly less comfortable than all other views (p < 0.001); Natural Greenery was more comfortable than White and Green Color (both p < 0.001); and Shadows of Greenery exceeded Green Color (p < 0.001) but did not differ significantly from White after correction (p = 0.017) (Figure 11).

4. Discussion

Our study investigated frontal alpha asymmetry (FAA = InF4 − InF5) and hemispheric frontal alpha power while participants viewed four indoor-view conditions: White Color (control), Green Color, Natural Greenery and Shadows of Greenery. FAA differed across conditions, with Natural Greenery consistently showing the most negative values. Although Bonferroni-adjusted pairwise comparisons did not reach the corrected threshold, this directional pattern was reinforced by the Trigger and Hemisphere analysis reported below.

In the two-way repeated measures ANOVA, the interaction between Trigger and Hemisphere showed significant results, consistent with greater right-frontal engagement in the case of Natural Greenery (lower right frontal alpha power (F4) relative to left (F3)). White Color and Shadows of Greenery did not survive Bonferroni correction but showed similar directional tendencies. Green Color showed no hemispheric difference. Taken together, these findings indicate that Natural Greenery elicits a clearer right-left differentiation in frontal alpha activity than simplified green cues.

4.1. Attention Restoration Theory (ART) and Neural Evidence

According to ART, when natural scenes evoke “soft fascination” they capture involuntary attention with minimal effort, thereby allowing directed attentional resources to rest and recover from fatigue [11,80]. In the present study, Natural Greenery produced lower alpha power at F4 relative to F3 (negative FAA), consistent with greater right-frontal engagement during passive viewing. Given previous evidence that right hemisphere engagement, particularly in frontal regions, supports attentional vigilance and monitoring [51,81], (and in some contexts, threat-related orienting [82]), our results correlate more closely with an attentional processing rather than a valence-based interpretation of FAA. This interpretation is also consistent with the experimental paradigm used in the present study. Participants passively viewed static indoor scenes for a relatively short stimulus duration (8 s), in addition to the picture being static, which more likely engaged perceptual monitoring and attentional orientation rather than sustained affective states. Adding to this perspective, neuroimaging work has also reported the recruitment of right-lateralized networks involved in bottom-up attention during nature exposure [51,52]. EEG and fMRI studies grounded in ART have described changes in brain wave activity consistent with reduced executive demand during natural scene viewing [53,54,55,56]. Within ART, such patterns are compatible with “soft fascination,” whereby natural features hold attention effortlessly without sustained executive control.

By contrast, other lines of research emphasize affective calming and stress reduction rather than attentional engagement. For example, FAA differences during passive viewing have been interpreted as reflecting mood or approach-related tendencies [83], and neuroimaging studies have highlighted reduced stress-related limbic activity, following nature exposure [84]. In indoor biophilic design research, greenery has also been associated with lower-arousal electrophysiological profiles [85,86], suggesting relaxation mechanisms that may differ from attentional accounts. These differences may reflect variation in paradigm, outcome measures, and whether attention is engaged implicitly or assessed during tasks.

Previous work has relied on comparisons between contrasting conditions, such as urban versus natural environments, rather than systematically differentiating between different types of greenery. By testing Natural Greenery, Green Color, and Shadows of Greenery in the laboratory setting, the present results suggest that frontal alpha asymmetry can help distinguish which nature-related visual features elicit the strongest hemispheric engagement, which, in return, helps apply restoration-oriented theories into indoor design strategies.

Although the findings are discussed in relation to Attention Restoration Theory (ART), the present study did not directly measure attentional restoration or emotional states. Therefore, these interpretations should be considered as theoretical context rather than direct empirical evidence.

4.2. Divergence Between Neural and Subjective Responses

In the present study, Natural Greenery was associated with greater right-frontal engagement, while participants rated this condition highest in comfort and likeability relative to the other viewing conditions. If FAA is interpreted through an affective/motivational lens, relatively greater right-frontal activation is often discussed in relation to withdrawal-oriented or negatively valenced processing [45,46,47,48,49,50]. This expectation is not supported by the present pattern. This contradiction suggests that in our experiment, FAA may not be indexing valence alone. Notably, the positive subjective ratings of comfort and likability observed for the Natural Greenery condition further argue against a negative affective interpretation of right-frontal activation. Instead, the right-frontal activation could be context-dependent, linked to attentional monitoring during scene processing. This is consistent with ART’s attentional interpretations of effortless engagement during nature simulation exposure and could have implications for interpretations of nature restoration. However, as attention and fatigue recovery were not directly assessed, this is presented as a theoretically informed interpretation instead of evidence of restoration per se. More broadly, it is plausible that subjective appraisals and FAA capture partially distinct components of the experience, which should be tested directly in future work using combined neural, behavioral, and self-report measures.

4.3. Chromatic and Visual Complexity Considerations

Our study addresses important questions about the role of chromatic composition and visual complexity in neural response to greenery. Although previous studies have shown that green color promotes relaxation, influences brain activity, and reduces depression [70,71], our results suggest that chromatic hues alone are insufficient to elicit the same neural pattern noticed with natural greenery. Likewise, while fractal patterns commonly found in nature evoke positive esthetic responses [66,67], our findings show that, as an isolated factor, greenery patterns devoid of chromatic distinction might not be sufficient to achieve the same impact. The combined effects of chromatic hues and the dynamic complexity of natural forms may play a critical role in modulating attentional and affective responses, pointing to the need for further research to examine these factors.

4.4. Design Implications: What This Means for Architectural Practice

The findings from this study indicate that indoor green design features are not processed equivalently at the neural level. In practice, green paint, simplified plant motifs, or shadow-like patterns are often used as substitutes for real greenery. Here, Natural Greenery produced a more distinct frontal alpha pattern than Green Color, while Shadows of Greenery showed a similar directional tendency that did not survive correction. This suggests that visual access to actual vegetation, through indoor planting, green wall systems, or views to outdoor greenery, may be more likely to elicit the neural profile observed during Natural Greenery viewing than simplified green cues alone.

Consistent with the neuroarchitecture perspective that links the built environments to brain function and experience [26,28,87], these results illustrate how EEG measures such as FAA and hemispheric alpha power can complement self-report when evaluating indoor environmental features. Combined with established restoration and attention outcomes, such approaches can support evidence-based biophilic design strategies.

4.5. Limitations and Future Directions

4.5.1. Methodology

Methodological differences may have affected the overall EEG results, as previous studies with similar approaches typically compared natural to non-natural settings [34,61,62,63,64], whereas our study compared triggers representing different associations with nature.

4.5.2. Stimulus Design

The present experimental study has been represented by a single stimulus for each category (one natural green image, one green color stimulus, and one shadow pattern). Consequently, the observed effects may partly reflect properties specific to these particular images rather than the broader stimulus categories. Moreover, viewing nature on a screen is a proxy for real environments. Previous research suggests that digital nature stimuli can evoke measurable psychological and physiological responses, but their effects may differ in response to real environments due to the absence of multisensory cues such as depth, movement, and environmental context [88,89]. Future studies should therefore include multiple stimulus exemplars within each category or systematically manipulate visual parameters to improve the generalizability and external validity of the findings. Additionally, integrating complementary physiological or behavioral indicators, such as heart rate variability or attention-based tasks, would provide a more comprehensive assessment of restorative responses.

4.5.3. Room Layout

We used 24 room variations with different window shapes and placements to avoid tying the stimuli to a single spatial configuration. Window layout was not treated as an independent experimental factor in the statistical analysis. Because each stimulus condition was presented across the same set of room layouts, window configuration was not systematically associated with any single stimulus condition. Future studies systematically examining multiple window configurations across architectural variables, such as window size or placement, could help determine whether these spatial factors influence neural or subjective responses.

4.5.4. Participant Variability

Participants’ cultural backgrounds varied, which may have influenced responses across all visual conditions. Differences in esthetic preferences, familiarity with natural cues, and culturally shaped associations with greenery or specific colors may have contributed towards variability in both subjective and neural responses. In addition, individual differences such as prior exposure to natural environments or biophilic tendencies were not systematically assessed in the present study and could be explored in future research. These factors may partly limit the generalizability of the findings across populations.

4.5.5. Sample Size and Screening

Although a power analysis indicated that n = 25 is sufficient power to detect medium effects for the present experimental design, the modest size may have reduced sensitivity for detecting smaller condition-specific differences, particularly in the context of EEG research examining frontal alpha asymmetry (FAA). This may partly explain why Bonferroni-corrected pairwise comparisons did not reach statistical significance despite the overall effect observed in the repeated measures ANOVA. Therefore, the neural interpretations presented in this study should be considered with caution. Future research with larger samples is needed to further verify the robustness and generalizability of the observed patterns. Additionally, while participants were screened for major cognitive impairments and mental health conditions, pre-existing cognitive and emotional states were not directly assessed. Lastly, although participants reported normal or corrected-to-normal vision, formal ophthalmological or cognitive screening tests were not conducted prior to the experiment. Future studies could incorporate standardized visual and cognitive assessments to further control for potential individual differences.

5. Conclusions

This study affirms the critical role of visual access to natural greenery in an indoor environment. We used EEG to examine whether green visual cues derived from natural greenery can work as substitutes in indoor window views. We compared Natural Greenery with Green Color and Shadows of Greenery. Natural Greenery showed the clearest hemispheric differentiation in frontal alpha activity and the most pronounced FAA tendency, alongside the highest likability and comfort ratings.

In contrast, while Shadows of Greenery showed a similar direction, Green Color did not produce a comparable pattern. The evidence from this study implies that nature-related visual features are not processed equivalently, and that real greenery elicits a distinct neural pattern compared with simplified green cues. From a design perspective, these findings support the importance of incorporating actual vegetation or direct visual access to greenery in indoor environments rather than relying solely on symbolic green elements. Such evidence contributes to the development of evidence-based biophilic design strategies aimed at improving user experience and well-being in built environments.

These findings may also have implications for architectural certification frameworks. For example, they can support the criteria used in certification frameworks such as LEED and WELL, which both incorporate biophilic principles to improve environmental quality and human health. Furthermore, these findings provide a modest contribution to the database needed to understand how biophilic principles affect the human brain. Finally, they could serve as evidence to help architects justify integrating natural greenery into built environments.

Lastly, these findings could bridge the gap between neuroscience and practical applications of it in indoor design, highlighting the value of integrating neuroscientific methods into architectural practice—an emerging discipline often referred to as neuroarchitecture.