Immediate Exposure – Environmental Effects in the Range of Minutes to Seconds

Disentangling Additive and Supra-Additive Effects of a Multimodal Forest Exposure
Nature's Effects on the Neural Pathways of Pain Perception
Examining the Neural Pathways Behind Nature Sounds and Their Effects on Cognitive Functioning
Capturing Lived Experience at Scale: A Large-Language Model (LLM)-Based Chatbot for Descriptive Experience Sampling
Visualizing House Facades Linked to Human-Rated House Attributes Using Generative Adversarial Networks
Form Preferences in Explicit Interior Design Choices
Examining Implicit Response Tendencies to Interior Environments
Comparing Neural Representations in Responses to Natural and Artificial Features in Indoor Environments
Conclusion

To understand the processes and mechanisms by which the physical environment affects mental and brain health, it is helpful to further restrict the timeframe of interest to immediate exposures. Here the methodological challenges are to maintain ecological validity even in lab settings and to capture participants' lived experience. To address these, CEN employs virtual reality (VR) scenarios, which permit immersive experiences while enabling experimental manipulation and thereby control over environmental parameters. The relevance of greenspace and sound exposure as observed on the observational/correlational as well as on the interventional/causal level in long-term and short-term focused studies is tested further in relation to immediate exposures. To extend the toolkit to adequately assess first-person experience, CEN is developing novel methods to assess this by means of interviews utilizing large language models.

So far, the research reported by CEN has solely focused on the outdoors. However, modern humans spend the most significant portion of their daily lives—80–90%—indoors in built environments. Accordingly, some of the immediate exposure projects presented here attempt to delineate which features of indoor environments and house facades contribute to brain and mental health. This extends CEN's purview into the novel topic of architectural neuroscience, a sub-branch of environmental neuroscience.

Disentangling Additive and Supra-Additive Effects of a Multimodal Forest Exposure

Forest-bathing, the act of spending time in a forest, is practiced in support of mental and physical well-being. Yet many people cannot easily access real forests. While VR can simulate natural experiences, it is still unclear whether integrating multiple senses can enhance emotional and cognitive recovery after acute stress.

In this study, participants first underwent a stress-induction procedure—as often done in environmental psychology studies—and were then immersed in a virtual Douglas fir forest under one of four conditions: visual, auditory, olfactory, or multimodal (combining all three senses) (Figure 1). Researchers assessed changes in positive and negative affect, connectedness to nature, and cognitive performance, including working memory, inhibition, and executive function.

Across all experimental conditions, participants showed recovery from stress and negative affect. In addition, the multimodal forest condition showed selective supra-additive benefits—that is, effects where the multimodal condition showed larger effects than the sum of effects of the unimodal conditions—namely for positive affect and nature connectedness. Cognitive improvements were limited, with working memory improving modestly in the multimodal condition, while the holistic experience showed no consistent advantage for the other domains.

These findings indicate that VR forest-bathing can provide meaningful emotional restoration, and that combining multiple sensory cues offers additional benefits. For individuals confined to indoor settings such as frail, elderly, or imprisoned populations, multisensory VR may be particularly valuable.

Key Reference

Nature's Effects on the Neural Pathways of Pain Perception

Positive effects of nature exposure have also been reported in the acute perception of pain. However, what is not clear yet is whether nature induces genuine analgesia or whether it has a domain-general effect, e.g., on general affective or cognitive processes, and is therefore rather the result of subjective reporting biases. To test this, CEN and collaborators conducted a neuroimaging study on healthy participants (N = 49), who received electrical shocks while in the MRI scanner as they were exposed to virtual videos of physical environments. Participants reported reduced pain when exposed to virtual nature videos compared to matched urban or indoor control settings. This subjective analgesic effect was primarily associated with reductions in nociception-related rather than domain-general cognitive-emotional neural pain processing regions in the brain. This result demonstrates that virtual nature exposure enables genuine analgesic effects through changes in low-level nociceptive and somatosensory processing. The fact that an analgesic effect can be achieved with easy-to-administer virtual nature exposure has important practical implications for clinicians, showing how nature might be used as part of non-pharmacological pain treatment programs.

Key Reference

Examining the Neural Pathways Behind Nature Sounds and Their Effects on Cognitive Functioning

Based on a previous behavioral study that demonstrated benefits from listening to natural soundscapes on working memory, CEN set out to investigate the neural pathways of the effects of natural and urban soundscapes on cognitive performance. Stobbe et al. (2024) conducted a pre–post-test within-subject fMRI study in which 35 participants were given affective questionnaires and two working memory tasks (dual n-back, backward digit-span task) before and after being presented natural and urban soundscapes in an MRI scanner. Higher functional connectivity was observed in the auditory, cinguloopercular, and somatomotor hand and mouth networks when listening to nature as compared to urban sounds, suggesting enhanced direct processing of the sound input and recall of associated action representations. Notably, this higher functional connectivity was positively associated with a gain in cognitive performance, suggesting that the positive effects of nature sounds on cognition may in part be driven by neural processing of the auditory stimulus but importantly may involve motor imagery.

Key References

Capturing Lived Experience at Scale: A Large-Language Model (LLM)-Based Chatbot for Descriptive Experience Sampling

Assessing the nuances of immediate lived experience is essential in psychology and neuroscience, but it remains a methodological challenge. The Ecological Momentary Assessment (EMA) approach prompts participants at random moments to complete brief questionnaires, but it is likely that fixed non-individualized items may flatten out subtle phenomenological differences in momentary experiences. Descriptive Experience Sampling (DES) offers a richer alternative: a trained interviewer probes what occurred in participants' experience at each sampled moment (when the participant hears an administered "beep"). Through careful, non-leading questioning, the interviewer helps participants articulate the precise character of their inner experience. This method yields detailed phenomenological descriptions but requires interviewers who are highly skilled in bracketing suppositions and therefore does not scale.

This ongoing project is developing a chatbot that conducts DES interviews autonomously, combining the depth of qualitative inquiry with scalability. The system employs an LLM-based pipeline that classifies participant responses, assesses them along quality dimensions drawn from expert DES practice, selects appropriate interviewing strategies, and generates non-leading follow-up questions. An experience registry tracks evolving descriptions across the interview. CEN aims to use this DES tool, once tested and validated, to capture immediate first-hand lived experience in different environmental contexts.

Visualizing House Facades Linked to Human-Rated House Attributes Using Generative Adversarial Networks

This work investigated whether concepts of house facades, such as whether a house looks "inviting," could be visualized using a generative image model. To achieve this, CEN utilized a generative adversarial network (GAN) to model how people evaluate house facades. GANs, widely used for image creation and manipulation, offer a controlled way to synthesize realistic stimuli.

Pohlmann, Tawil, et al. (2025) trained a GAN on a dataset of 2,000 house images, each accompanied by human ratings, which were acquired in an online study in which 204 participants rated the dataset on five dimensions: hominess, safety, invitingness, relaxation, and price. After training, the GAN was able to generate examples of an "uninviting" house and an "inviting" one. In a second online study, another group of 204 participants rated these GAN-generated images on the same dimensions to test whether the model could successfully embed targeted architectural characteristics.

Statistical analyses verified that participants' evaluations of the synthetic images corresponded to the intended psychological attributes used during image generation. Figure 2 shows examples of the generated images for "invitingness." These findings highlight the promise of generative image models as a methodological tool for architectural and environmental psychology research enabling researchers to visualize subjective rating dimensions of humans.

Key Reference

Form Preferences in Explicit Interior Design Choices

In this study, the researchers examined form and style preferences in interior design. Prior research has frequently reported a general preference for curved over angular forms, yet previous findings have often relied on predefined and static stimuli. To address this limitation, the present work introduced an interactive 3D web application (Figure 3) that allowed participants to furnish a virtual living room according to their own preferences while maintaining some experimental control. Participants (N = 196) selected items from a standardized set of furniture that varied systematically in form (angular or curved) and style (modern or classic).

The resulting choice patterns revealed a small but reliable overall preference for angular furniture, contradicting earlier studies, in which fixed stimuli had to be evaluated. The effect was driven by male participants, who chose angular items more frequently; female participants showed no systematic form preference. No general preference emerged for modern versus classic styles. Exploratory analyses linking design choices to personality traits measured using the Big Five Inventory 2 did not yield significant associations.

These findings demonstrate the utility of interactive 3D environments for investigating aesthetic preferences in settings that preserve both ecological validity and experimental control.

Key Reference

Examining Implicit Response Tendencies to Interior Environments

Do visual forms in interior spaces shape our immediate behavior? Participants report preferring curved over angular designs, but it is unclear whether this simply reflects conscious preferences or in fact deeper, automatic tendencies to approach or avoid a space. Early theoretical accounts suggest that angularity triggers avoidance due to perceived threat, while others posit that curvature is inherently pleasant and invites approach. 

To investigate these tendencies, 219 online participants were shown images of rooms featuring either curved or angular design elements. The researchers measured several implicit responses related to approach–avoidance, including attention, automatic associations between form and behavioral tendencies, and movement responses when attending directly to form and to other image features (i.e., orientation) (Figure 4).

Although attentional and general approach-avoidance responses were not clearly affected by form, automatic associations (IAT) and movement responses (SRCT) when attending to form revealed consistent patterns: curved designs were linked with approach and angular ones with avoidance. Moreover, curved designs were approached faster than they were avoided, whereas responses to angular designs were more comparable, indicating that the effect was primarily driven by attraction to curvature rather than an aversion to angularity.

These findings show that even simple design features can influence how welcoming a space feels beyond the conscious level. Future studies of CEN aim to address how these effects extend to immersive or real-world environments, and, crucially, how they are represented in the brain. This knowledge will be essential for guiding evidence-based architectural strategies aimed at promoting well-being in built environments.

Key Reference

Comparing Neural Representations in Responses to Natural and Artificial Features in Indoor Environments

This study asked whether naturalistic versus artificial architectural features in indoor environments influence subjective evaluations and brain responses beyond early visual processing. To address this question, Sztuka et al. (2025) combined behavioral measures (similarity judgments, subjective ratings) and fMRI. Thirty-six participants viewed images of naturalistic and artificial indoor architectural scenes while undergoing fMRI, completed a spatial arrangement task to indicate perceived similarity between scenes, and rated each scene on naturalness, preference, desire to be there, and arousal. This multimodal approach allowed researchers to relate subjective experience, visual features, and neural representational spaces using Representational Similarity Analysis (RSA). 

Behaviorally, naturalness had minimal influence on subjective ratings such as preference or desire to be in a space, and it did not reliably predict core dimensions underlying similarity judgments. Instead, similarity judgments were driven primarily by low-level visual features, such as brightness, edge density, and fractal properties. At the neural level, RSA revealed that distinctions between naturalistic and artificial interiors emerged predominantly in the ventral visual stream, with no robust extension into regions commonly associated with affective or restorative processing, such as the prefrontal cortex, hippocampus, or amygdala. Post-hoc analyses further showed that neural representations were more closely aligned with low-level visual features than with subjective naturalness ratings. 

Taken together, these findings challenge assumptions of biophilic design (that natural elements in design directly impact affect and cognition), suggesting instead that perceived naturalness in indoor architecture does not straightforwardly map onto higher-order cognitive or affective brain networks. Instead, early visual processing plays a dominant role. The results raise important questions about how architectural experience accumulates over time, how context may modulate neural responses, and how design principles might move beyond visual naturalness toward more comprehensive models of environmental experience.

Key Reference

Conclusion

At the bottom level of the funnel, CEN studies the mechanisms behind the causal effects of immediate and momentary environmental exposures, using methods that aim to capture participants' lived experience as robustly as possible. Through laboratory and quasi‑naturalistic experiments, CEN researches the proximal neural pathways by which environments alter affect, cognition, and pain. Multisensory manipulations show supra‑additive gains for affect and nature connectedness, while nature exposure can down‑regulate nociceptive processing—evidence for genuine analgesia rather than mere report bias. CEN's soundscape work links nature sounds to functional connectivity patterns that align with cognitive improvements—clarifying when and how environments "get under the skin." Architectural studies indicate that perceived naturalness in indoor scenes is represented predominantly in early visual systems, and that curvature can bias approach tendencies—mechanistic clues that refine biophilic claims. Advances in methodology (VR/fNIRS/fMRI, DES chat‑interviewing) increase ecological fidelity while preserving experimental control.