ABSTRACT. Background

Social media surveillance behaviours are increasingly prevalent yet poorly operationalised in psychological research. Existing measures often fail to differentiate normative engagement from problematic monitoring. The Social Media Surveillance Scale (SMSS) was developed to address this gap by capturing behaviours along a continuum from everyday “creeping” to more intrusive behaviours such as “cyberstalking.” Previous stages involved systematic item generation and expert content validity review to ensure theoretical coverage and face validity. Building on these foundations, the current phase aimed to refine the initial 113-item pool and validate its structure, with construct validity testing now underway.

Method

A large and diverse sample of social media users (N = 1,641; Mean age = 42.65) were recruited via Prolific. Inclusion criteria required participants to be aged 18+ and active social media users. The refined 113-item SMSS was administered online via Qualtrics. Items were rated on 7-point Likert scale assessing behavioural likelihood and motivational agreement.

Data were screened for missingness, coding errors, and assumption violations. Given ordinal data and non-normality (Mardia’s test, p < .001), polychoric correlations and robust estimation methods were applied. To minimise capitalization on chance, the dataset was randomly split using the Solomon method: Subsample A (n = 821) for Exploratory Factor Analysis (EFA) and Subsample B (n = 820) for Confirmatory Factor Analysis (CFA). EFA employed Principal Axis Factoring with oblimin rotation to allow correlated factors. Factor retention decisions were guided by parallel analysis, scree plot inspection, eigenvalues > 1, and theoretical interpretability. Items were retained if they demonstrated primary loadings ≥ .40, cross-loadings < .30, and communalities ≥ .30. Internal consistency was assessed using Cronbach’s α and McDonald’s ω.

All participants provided informed consent, and the study was conducted in accordance with the British Psychological Society (BPS) Code of Ethics and approved by the University of Sunderland Ethics Committee.

Results

Initial EFA suggested an eight-factor solution, which was iteratively refined to achieve improved fit and interpretability. Parallel analysis and scree plot inspection supported a six-factor structure comprising 48 items. The six dimensions were labelled based on thematic inspection: Preoccupied Engagement, Investigative Searching, Deceptive Identity Monitoring, Privacy Intrusion, Affiliative Social Interaction, and Relational Following. Internal consistency was excellent across factors (Cronbach’s α = .82–.95; ω = .84–.95).

CFA on Subsample B confirmed the six-factor model as the best-fitting representation of the data (χ²[1065] = 11,347.39, CFI = .98, TLI = .98, RMSEA = .11), supporting the multidimensional nature of social media surveillance behaviours. Item loadings demonstrated conceptual clarity with minimal cross-loadings, and reliability indices exceeded conventional thresholds. The final 48-item SMSS therefore provides a parsimonious yet comprehensive measure of online surveillance behaviours.

Current Work and Future Directions

Construct validity testing is currently in progress and will be completed before the conference. This phase examines convergent validity with theoretically related constructs (e.g., anxiety, dark personality traits, problematic social media use) and divergent validity with opposing constructs (e.g., empathy, prosocial tendencies). Preliminary hypotheses predict positive associations with GAD-7, Short Dark Triad, and Bergen Social Media Addiction Scale, and negative associations with the Toronto Empathy Questionnaire and Prosocial Tendencies Measure. These analyses will aim to establish the SMSS within a broader nomological network, providing evidence of its theoretical and practical utility.

ABSTRACT. Introduction: As smartphones become inseparable from daily life, the phenomenon of "phubbing"—ignoring physically present individuals in favor of a mobile device—has emerged as a significant threat to the quality of face-to-face social interactions. While traditional explanations often rely on technological addiction, this study utilizes the Uses and Gratifications Theory to argue that phubbing is a goal-oriented behavior driven by fundamental human motives. We specifically examined how the need to belong (the drive for social connection) and the need for popularity (the drive for social status) motivate phubbing behavior. Central to our investigation is the role of the fear of missing out (FoMO) as the emotional mechanism that translates these broad social needs into the specific act of phubbing. By investigating these pathways, we aim to demonstrate that phubbing is not merely a habit, but a strategic—albeit often counterproductive—attempt to satisfy social needs in the digital age. Methods: The study utilized a sample of 242 adult participants (Mage = 36.67, range 18–71; 55.8% female). Data were collected using the Phubbing Scale, the Fear of Missing Out Scale, the Need for Popularity Scale, and the Need to Belong Scale. The focal analysis employed a series of mediation models to test whether FoMO (mediator) explains the relationship between social needs (predictors) and two dimensions of phubbing: Communication Disturbance and Phone Obsession (criterion variables). Results: The results showed that both the need to belong and the need for popularity are significant predictors of both dimensions of phubbing. Furthermore, FoMO was positively correlated with both social needs and served as a strong predictor of phubbing dimensions. Mediation analyses revealed distinct psychological pathways: FoMO was found to fully mediate the relationship between the need to belong and Communication Disturbance, while it partially mediated the relationship between need to belong and Phone Obsession. This suggests that individuals with a high need to belong do not phub others simply by default; rather, they do so because the anxiety of being excluded (FoMO) "overrides" the immediate social cues of the face-to-face environment. Regarding the need for popularity, FoMO was found to partially mediate its relationship with both dimensions of phubbing. These results suggest that while the pursuit of social status leads to anxiety about missing opportunities for digital validation, the need for popularity also exerts a direct influence on phubbing, likely due to the perceived efficiency of digital platforms for maintaining a broad social audience. Discussion: Our findings provide compelling evidence that phubbing is a manifestation of underlying social motives rather than a simple lack of self-control or technological addiction. The study highlights a modern paradox: the very needs that drive us toward others (belonging and popularity) are the same drivers that lead us to ignore the people physically present before us. Phubbing represents a shift where individuals prioritize the "attention economy" of the digital world—with its rapid rewards of likes and comments—over the more nuanced but demanding nature of offline communication. Practically, these results underscore the need for "digital hygiene" programs that help individuals recognize how their core social needs are being exploited by high-engagement digital content, potentially leading to a cycle of reduced social competence and increased loneliness.

ABSTRACT. Internet Addiction Disorder (IAD) has been increasingly conceptualized as a maladaptive behavioral pattern supported by specific cognitive underpinnings, including attentional mechanisms. Neurocognitive models suggest that altered selective attention, reward sensitivity, and dissociative processing may contribute to the persistence of problematic Internet use. The present study examined the association between Internet addiction and dissociative experiences by focusing on attentional bias toward Internet-related stimuli and temporal dissociation due to cognitive absorption during Internet usage as potential neurocognitive markers. Seventy young adults (69% female; M age = 23.8 years, SD age = 6.2) completed self-report measures assessing the risk for Internet addiction (IAT), temporal dissociation due to cognitive absorption during Internet usage (CABS), and problematic and normal dissociative experiences in everyday life (DES-II). Attentional bias toward social network–related stimuli (AB) was assessed using a visual dot-probe task during a laboratory session. Negative AB score indicates the presence of attentional bias toward social network–related stimuli. IAT score allowed participants to be categorized into a risk group (n = 36, IAT score ≥ 50) and a control group (n = 34, IAT score ≤ 30). In line with literature, participants in the risk group showed significantly lower AB, thus exhibiting a stronger attentional capture by social network stimuli. Moreover, the risk group showed significantly greater temporal dissociation due to cognitive absorption during internet use, reflecting reduced metacognitive monitoring of time during Internet engagement for these participants and suggesting a possible association between attentional capture by addictive stimuli and altered states of consciousness. Also, the risk group reported significantly higher dissociative experiences in general life compared to those without bias. Of relevance, dissociative scores in this group exceeded the established clinical cutoff employed for dissociative disorder screening. These findings support neurocognitive accounts of Internet addiction by highlighting how both attentional bias and temporal dissociation due to cognitive absorption work as interconnected mechanisms underlying excessive Internet use. Attentional bias may reflect enhanced stimulus-driven processing and dysregulated top-down control, contributing to compulsive engagement. The results underscore the relevance of also targeting dissociative mechanisms as possible influencing individual difference and dissociation risk as possible consequence, in cognitive and neurobehavioral interventions for Internet addiction.

ABSTRACT. Excessive time on social media is consistently linked to more problematic and addictive patterns of use (Amirthalingam & Khera, 2024; Leite et al., 2023; Zhao, 2021), yet time online alone does not explain why only some individuals develop maladaptive engagement (Matthews et al., 2025). Emotional processes such as shame, a self-conscious emotion associated with perceived social threat and feelings of inferiority or unworthiness (Gilbert, 1998, 2003), have also been related to more addictive social media use (Farkush et al., 2022). How people respond to these aversive internal experiences may depend on psychological flexibility, a core process in Acceptance and Commitment Therapy (ACT; Hayes et al., 1999) referring to the capacity to remain present and act according to personal values despite difficult thoughts and emotions. Lower psychological flexibility has been linked to maladaptive coping with shame (Gul & Aqeel, 2020) and to higher social media addiction (Güldal et al., 2022). However, it is unclear whether these processes interact and whether their interplay depends on daily time spent online. The present study aimed to examine whether psychological flexibility mediates the association between shame and social media addiction, and whether this indirect effect varies according to daily time spent on social media. Participants were 352 adults aged 18-89 years (M = 31.48, SD = 13.05) who completed an online survey including the Social Media Addiction Scale (SMAS), the External and Internal Shame Scale (EVEI), and the Comprehensive Assessment of Acceptance and Commitment Therapy Processes (CompACT). Analyses were conducted in SPSS 29 using the PROCESS macro (Model 15) in which daily hours of use (<1h; 1-4h; >4h) were specified as a categorical moderator of the paths from psychological flexibility to social media addiction and from shame to social media addiction, with age included as a covariate. Age showed significant associations with the main variables and was therefore controlled. Higher shame was associated with lower psychological flexibility and higher social media addiction, while lower flexibility and greater daily use were both related to higher social media addiction. In the moderated mediation model, shame significantly predicted lower psychological flexibility (b = -1.14, p < .001). The final model explained 32% of the variance in social media addiction (R² = .32, p < .001). Compared to very low users (<1h/day), both the 1-4h and >4h groups reported higher addiction levels (ps < .001). There was no significant direct effect of shame on addiction at any level of use (ps = .14-.91). The association between psychological flexibility and addiction depended on hours of use. Psychological flexibility was unrelated to addiction among very low users (<1h; p = .72), but was negatively associated with addiction among those using social media 1-4h/day (b = -0.21, p < .001) and >4h/day (b = -0.21, p < .001). Conditional indirect effects showed that shame was linked to higher addiction through reduced psychological flexibility in the 1-4h group (b = 0.24, 95% CI [0.15, 0.34]) and the >4h group (b = 0.24, 95% CI [0.11, 0.39]), but not in the <1h group (b = -0.04, 95% CI [-.26, .21]). Indices of moderated mediation confirmed that this indirect effect was significantly stronger in both regular-use groups than in very low users. These findings indicate that the pathway from shame to social media addiction through reduced psychological flexibility is present only among individuals who use social media for at least about one hour per day. Results suggest that interventions for problematic social media use may benefit from not only reducing time online but also strengthening psychological flexibility, particularly among regular users.

ABSTRACT. Introduction: The person-oriented approach sought to examine individual behaviors and patterns of functioning in a promising, holistic way, not only in traditional offline contexts but also in online settings. Although the usefulness of this approach is evident, research on temperamental patterns in adolescence is even more limited in examining how these patterns predict different online behaviors. Given that digital devices and platforms are deeply embedded in adolescents’ daily routines, it is crucial to identify individual-level predictors of adaptive and maladaptive online behaviors by integrating established temperamental theories with recent research on new technologies. This integration could help identify both vulnerabilities and protective patterns in online use. Aim: We examined patterns of temperamental profiles and their associations with online use among adolescents. Specifically, (1) we firstly identified typological temperamental profiles, and (2) we connected profiles with several indicators of online positive and negative use, such as social media engagement, problematic and excessive smartphone and social network use, and specific online behaviors (i.e., active social network use; offering and searching for online social support). Method: 339 Italian adolescents (Mage = 15.35; SD = .64; 36% females) completed the following measures: Early-Adolescent Temperament Questionnaire (anger/frustration, sadness, attentional control, activation control, and inhibitory control subscales); Social Media Engagement Scale; Smartphone Addiction Scale; Bergen Social Media Addiction Scale, Active and Passive Social Network Scale. Results: (1) Latent Profile Analysis identified four distinct patterns: (a) Adjusted (10% of the sample), low anger/frustration and sadness, high attentional control, and average-high activation and inhibitory control. (b) Average (28% of the sample), average scores in all the temperamental dimensions. (c) Frustrated/regulated (21% of the sample), high anger/frustration, low sadness, average-high attentional and activation control, and average inhibitory control. (d) Emotional/dysregulated (31% of the sample), average-high anger/frustration and sadness, low attentional, activation, and inhibitory control. (2) Post-hoc analyses of variance showed that the four profiles differed in online behaviors, revealing the following trends: a) for social media engagement [F (3,324) = 10.52; p < .001; η = .08], the Adjusted significantly showed lower mean scores than the other three profiles. b) for problematic smartphone use [F (3,319) = 21.30; p < .001; η = .16] and problematic social network use [F (3,316) = 23.07; p < .001; η = .18], the Adjusted significantly showed the lowest mean score, the Average and the Frustrated/regulated showed mean scores, and the Emotional/dysregulated showed the highest mean score. c) for active social network use [F (3,318) = 3.82; p < .05; η = .03], the Adjusted significantly showed the lowest mean score, the Average did not significantly differ in their mean scores, while the Frustrated/regulated and the Emotional/dysregulated showed the highest means. d) for seeking online social support [F (3,315) = 2.27; p = n.s.; η = .02] and offering online social support [F (3,317) = 1.89; p = n.s.; η = .02], despite profiles did not significantly differ in mean scores, we evidenced a tendency for seeking social support from the Frustrated/regulated and the Emotional/dysregulated, that showed high scores. Conclusion: Our preliminary findings can provide a useful framework for understanding adolescents’ engagement with online contexts. Specifically, adopting a person-oriented perspective enables the identification of distinct patterns of functioning based on key individual differences. This approach can help analyze how new generations approach and use technology and online platforms, identify potential vulnerabilities, and implement timely, preventive interventions to counter maladaptive online behaviors.

ABSTRACT. Technology facilitated sexual violence (TFSV) encompasses a broad spectrum of behaviours where digital technologies enable or amplify sexually-based harms. Within the landscape of digital relationships, sexting has become a prevalent and normative behaviour among young adults, but exists along a continuum from consensual to non consensual interactions. This presentation synthesises a large body of research examining sexting practices, the dissemination of intimate images, associated harms, and the complex and frequently blurred or miscommunicated role of consent. Multiple studies by the authors and others indicate that rates of receiving, sending, and disseminating sexts are high across Western contexts. While much of this is consensual, substantial proportions of participants report receiving unwanted sexts, sending under pressure, or having their images shared more broadly. Importantly, consensual sexting itself is not associated with negative mental health outcomes; however, non consensual experiences as outlined above, are associated with elevated levels of depression, anxiety, stress, and reduced self esteem. Some of these experiences are gendered, with women more likely to report receiving unwanted images in particular. A key focus of this paper will be sext dissemination, a behaviour often poorly assigned the moniker “revenge porn”. Evidence consistently demonstrates that revenge based motivations are uncommon; instead, dissemination is typically driven by more mundane or socially orientated reasons such as perceiving the person as attractive, sharing as a joke, gaining social status, or responding to peer request or pressure. This reframing is crucial for prevention work, as although common motivations may appear innocuous, non-consensual dissemination produces significant emotional, social, and legal consequences for victims and, at times, for perpetrators. Consent emerges as a central yet inconsistently understood or communicated component of digital sexual interactions. Across quantitative and qualitative data, marked discrepancies appear between individuals who disseminate and those whose images are shared regarding whether permission was granted. While around one fifth of disseminators report having consent, only a small proportion of victims believe they gave such permission. Qualitative findings reveal that discussions regarding privacy, saving or deleting images, and expectations for use are often implicit or assumed rather than explicit. Many participants rely on technological features (e.g., self deleting messages) rather than interpersonal communication. Only a minority explicitly negotiate consent for sending or disseminating images, and time bound or context specific consent is rarely addressed. We will present a synthesis of both quantitative and qualitative data over four years addressing trends in consent communication, which underscore the need for clearer conceptual and practical understandings of digital consent. As with in person sexual interactions, active, informed, and ongoing consent is essential. Interventions should avoid harmful or inaccurate terminology, emphasise the legal and emotional repercussions of dissemination, and equip individuals with the language and strategies needed to engage in explicit consent practices online. We will also address future directions for research including standardisation of tools and definitions, opportunities for ethical experimental designs, diverse populations, and ways to integrate the increasing challenges of AI and its association with digital intimacy. As digital intimacy continues to evolve, evidence based approaches are vital to informing policy, education, and prevention initiatives that support safer online sexual interactions and reduce TFSV-related harms.

ABSTRACT. Cognitive and motor impairments are highly prevalent among individuals affected by traumatic brain injury (TBI), stroke, and brain tumors, often leading to long-term deficits in autonomy and quality of life. These impairments commonly interfere with activities of daily living (ADL), such as shopping, navigation, and object manipulation. While conventional rehabilitation approaches remain effective, they often lack ecological validity and adaptability to diverse impairment profiles, limiting the transfer of learned skills to real-world contexts. Virtual reality (VR) technologies have emerged as a promising solution for rehabilitation, offering immersive environments that support controlled and engaging training scenarios. Increased immersion and ecological validity have been shown to contribute positively to motivation and long-term rehabilitation outcomes. A notable example is Reh@City v2.0, developed by Paulino et al., which leveraged VR-based simulations of everyday environments, including supermarkets and clothing stores, to support cognitive and motor rehabilitation. To accommodate users with severe motor impairments, Reh@City v2.0 implemented a simplified interaction system based on a camera-based tracking software, enabling participation by users with limited hand function or impaired hand–eye coordination. Even though Reh@City v2.0 made important contributions, advances in head-mounted display (HMD) technologies offered a new opportunity to overcome its limitations, regarding ecological validity and immersion. Novel HMDs provide substantially improved visual acuity, enhancing virtual environments with clearer and more detailed visuals. Integrated hand tracking allows for more natural and intuitive interactions, reducing the need for complex controllers while improving accessibility for users with motor or cognitive challenges. In addition, eye-tracking capabilities benefit users by supporting gaze-based navigation and personalized feedback. Both integrated tools enable the creation of more realistic interactions, therefore increasing ecological validity and promoting more flexible and accessible rehabilitation platforms. With this in mind, we propose Reh@City XR, an extended-reality rehabilitation platform currently under development. Reh@City XR introduces a modular accessibility framework based on multiple layers of abstraction, allowing therapists to tailor interaction, feedback, and task complexity to individual user needs. The system continues to employ activities of daily living as its core rehabilitative paradigm, thereby reinforcing real-world relevance and skill transfer. Advanced technologies, such as VR headsets, eye-tracking systems, and external controllers, are to be integrated into accessibility modules to support alternative interaction methods, such as gaze-based selection and multimodal feedback. Building upon previous work, Reh@City XR expands its focus to include more complex cognitive abilities, particularly social cognition and executive functioning. The platform implements ecologically valid scenarios targeting multiple rehabilitation domains, including activities of daily life (e.g., shaving, eating and washing), social cognition through interaction with intelligent virtual avatars, and visual spatial navigation through immersive wayfinding tasks (e.g., crossing the street). Additional cognitive processes such as attention, working memory, and problem-solving are addressed within structured, goal-oriented activities. Usability and accessibility evaluations will be conducted with healthy participants prior to clinical deployment, followed by case studies involving individuals with diverse motor and cognitive impairments. Ultimately, this work presents an ecologically valid XR rehabilitation platform integrating cognitive, motor, social, and visuo-spatial training within realistic daily life scenarios. It introduces a modular accessibility framework for personalized adaptation, novel avatar-based interventions targeting social cognition, and validated ADL tasks to support real-world skill transfer. The study also provides empirical evidence on the impact of advanced HMD technologies on accessibility, immersion, and rehabilitation outcomes, alongside design guidelines for inclusive and scalable neurorehabilitation.

ABSTRACT. Cognitive decline, functional frailty and sensory loss frequently co-occur in late life, especially in institutionalized or socially supported older adults. There is a need to promote scalable, engaging and ecologically valid interventions to support autonomy and quality of life at this age. The Sénior+Ativo promoted by the Municipality of Porto, Portugal is a project implemented in the city of Porto and consists of a multidomain model integrating neuropsychology, physical activity, and audiology. A technology-based cognitive stimulation program using the Systemic Lisbon Battery was conducted to stimulate cognition through simulated instrumental activities of daily living (IADLs) using tablet devices. The project involved 11 social institutions across Porto and enrolled 259 older adults. Neuropsychological assessment was conducted at baseline and at post-intervention. Measures included global cognition (M-ACE), executive functioning (FAB), memory (WMS-R subtests), cognitive reserve, depressive symptoms (GDS-15), and quality of life (IAQdV). The intervention consisted of 8 small-group sessions (5 participants; biweekly, 60 minutes) over 8 weeks, focusing executive functions, attention, and memory via ecologically valid tasks (e.g., shopping, pharmacy, meal preparation, medication-related activities, visual search, and planning). In parallel, participants engaged in structured physical exercise sessions aligned with weekly cognitive themes. Finally, a hearing screening component addressed presbycusis as a modifiable risk factor for cognitive decline. Neuropsychological data at baseline assessment were available for 204 seniors (mean age 82, range 60–99; 73% women). Initial screening suggested substantial vulnerability: mean M-ACE was 14.88 (SD 7.36; 61.3% below cut-off), and 56.9% scored below the executive function cut-off on the FAB. Post-intervention reassessment (currently n=61) showed higher mean M-ACE (19.21, SD 5.49) and FAB (10.51, SD 3.70) suggesting a positive impact of the intervention on the cognitive performance of this group. Additionally, qualitative satisfaction data indicated high acceptability, highlighting enjoyment, perceived memory benefits, social engagement, and digital inclusion (many participants reported first-time tablet use). A tablet-based, VR/serious-games cognitive stimulation program based in functional IADLs is feasible within community and residential care settings and is well received by very old adults with low digital literacy. Early data suggest potential cognitive benefits, supporting further controlled evaluation with improved follow-up completion and harmonized pre–post samples.

ABSTRACT. This abstract outlines a developing programme of research investigating how adults with acquired brain injury (ABI) engage with digital gaming communities, and how these socio-technical environments may support social capital formation, wellbeing, and engagement. While research on disability and gaming has traditionally emphasised accessibility, impairment, and exclusion (Hassan & Baltzar, 2022), considerably less attention has been directed toward how disabled players themselves co-create meaningful, pleasurable, and socially supportive experiences within networked play. This emerging work contributes to cyberpsychology by positioning gaming as a potential digital health resource that intersects with identity, community, and post-injury adjustment. Contemporary social capital frameworks, bonding, bridging, and linking (Putnam, 2015), offer a useful behavioural-science lens for understanding how adults with ABI participate in online games and gaming communities. Empirical work indicates that online and collaborative play may enable low-barrier participation, flexible communication, and reduced social isolation for individuals with cognitive, sensory, or physical limitations (Nilsen et al., 2024; Baltzar, Hassan & Turunen, 2024). Yet, the mechanisms underpinning these relational benefits remain underexplored. For people adjusting to ABI, digital play may offer structured opportunities to rebuild confidence, renegotiate social identity, and participate in communities where competence, reciprocity, and shared goals organise social interaction. The programme adopts an intersectional framework that considers disability alongside gender, class, age, and other identity dimensions that shape access to engagement and participation online. Within cyberpsychology, this approach emphasises engagement as a relational construct, linked to agency and competence. For adults with ABI who frequently report fatigue, mobility limitations, stigma, and reduced offline social opportunities (Åkerlund et al., 2021; Williams & Willmott, 2012; Bracho & Salas, 2024), digital play may offer alternative modes of engagement that align with energy constraints, fluctuating abilities, or the desire for socially safer environments. In this context, gaming communities are conceptualised as socio-technical systems in which platform design, communication infrastructures, moderation practices, and cultural norms shape participation and the forms of pleasure made available to disabled players (Judin, 2025). Understanding these structures has implications for digital health interventions, and potentially AI-driven systems aimed at supporting inclusive online participation. The developing project therefore aims to examine both the lived experiences of adults with ABI and the technological architectures that influence access, inclusion, and wellbeing. Preliminary qualitative insights suggest that some adults with ABI use gaming platforms to cultivate micro-communities, maintain long-standing relationships, and interact in environments where disability is less visible. These emerging patterns indicate potential pathways through which digital play may scaffold wellbeing and social connection. The planned research will use mixed methods, co-design, and participatory analysis with people with lived experience to investigate these processes more systematically and ethically. Through this orientation, the work contributes to growing conversations within cyberpsychology, cybertherapy, and digital health about how emerging technologies can be designed to support behavioural wellbeing and human flourishing. By centring the perspectives of disabled players, this programme aims to advance inclusive design principles, inform ethical technological development, and broaden understanding of how digital play functions as a site of relationality and social participation for adults with acquired brain injury.

ABSTRACT. Introduction: Visual memory impairments are among the most persistent and functionally disabling cognitive sequelae following acquired brain injury (ABI), compromising object recognition, learning, and everyday autonomy. Traditional neuropsychological assessment and rehabilitation protocols predominantly rely on bidimensional (2D) visual stimuli, which fail to preserve critical depth and spatial cues inherent to real-world perception. Emerging holographic 3D visualization technologies offer a promising alternative by providing richer spatial information while potentially reducing the cognitive load associated with mental reconstruction of object structure. Despite this potential, empirical evidence regarding the clinical utility of holographic displays in ABI populations remains scarce. Objectives: Grounded in Dual Coding Theory, spatial encoding models, perceptual saliency mechanisms (Von Restorff effect), and structure-from-motion principles, this ongoing study investigates whether holographic 3D visualization improves visual memory encoding, recognition accuracy, and retrieval efficiency compared to standard 2D presentations in individuals with ABI. Methodology: The study employs a single-session, within-subject cross-over design, enabling direct comparison of memory and processing speed across visualization modalities while controlling interindividual variability. Neutral everyday objects are selected from the Emotional Daily Life Library (E-DLL), a validated database of 3D stimuli with controlled emotional valence and arousal. During the encoding phase, participants are sequentially presented with 20 rotating objects (8 seconds per stimulus) under each condition: (a) 2D rotation displayed on a high-resolution flat-panel monitor, and (b) volumetric 3D rotation presented via a holographic curved half-dome Pepper’s Ghost display, similar to that presented by Luo and colleagues. Condition order is counterbalanced. A brief distractor task (3 minutes) follows to minimize short-term rehearsal effects. During the recognition phase, participants are shown 30 objects, comprising 20 previously encoded targets and 10 novel distractors, presented in randomized order. Participants indicate as quickly and accurately as possible whether each object was previously seen. Primary outcomes include recognition accuracy and reaction times, while secondary outcomes assess perceived cognitive load using the NASA-TLX. Expected Results: It is hypothesized that holographic visualization will facilitate more efficient memory encoding by preserving natural depth cues, thereby necessitating less cognitive effort for object structure interpretation. We expect this to manifest as significantly higher recognition accuracy and reduced reaction times in the 3D condition compared to 2D. Furthermore, the study explores: 1) the balance between the increased perceptual saliency provided by the E-DLL stimuli and their potential differential impact on cognitive processing in the ABI population; and 2) the balance between the potential effects of increased perceptual saliency on memory encoding and cognitive overload, which may differentially impact attention and cognitive processing in the ABI population. Conclusion: This work addresses a significant gap in cyberpsychology literature by empirically examining holographic visualization, distinct from stereoscopic or immersive virtual reality, in a clinical population. Unlike immersive VR, holographic displays allow naturalistic viewing without the discomfort or sensory-motor conflicts often associated with head-mounted displays in patients with ABI, and they represent a more affordable and accessible technological solution that can be implemented using standard flat-panel displays. By isolating the cognitive mechanisms tied to volumetric representation using a validated stimulus set, this study aims to establish evidence-based design principles for the next generation of neuropsychological tools. These findings will advance the integration of emerging technologies into clinical practice, moving beyond technological novelty toward functional recovery.

ABSTRACT. Introduction: Cognitive deficits can result from various causes, including neurological disorders, traumatic brain injuries, and neurodegenerative conditions. These deficits manifest differently and significantly impact an individual’s independence and quality of life. Technological solutions such as virtual reality, tablets, and software have shown considerable promise in cognitive rehabilitation; however, their effectiveness is often limited by a lack of personalization, with standardized content and technological interactions that may not be adequate to the patient’s needs. Tailoring interventions to meet patients' specific functional needs and interests can optimize engagement and improve clinical outcomes. Musiquence emerges as a versatile cognitive stimulation platform that addresses the need for personalization, allowing for total customization of content, interaction modes (e.g., touch, full-body motion), and unique musical feedback mechanisms. Nonetheless, the extent to which healthcare professionals can effectively utilize such platforms to develop diverse, clinically relevant interventions remains a crucial area for investigation. Objective: The primary goal of this study was to evaluate the customization potential of the Musiquence platform and assess the effectiveness of a co-creation methodology through a “Gaming Hackathon.” By bringing together healthcare professionals from several backgrounds, designers, and developers, this study aimed to determine how multidisciplinary collaboration could produce diverse, clinically feasible Sets of Activities (SOAs). Methods: This study employed a Gaming Hackathon held over 1.5 days in the Autonomous Region of Madeira. A total of 23 healthcare professionals (mean age 36.96; 22 women) from various fields, including psychology, nursing, physiotherapy, and speech therapy, participated. These professionals were organized into seven multidisciplinary teams, each supported by a professional designer and a Musiquence expert. The co-creation method followed a structured seven-step workflow: (1) material provision, (2) ideation and persona creation, (3) platform exploration, (4) initial feedback sessions, (5) stimulus selection, (6) technical implementation, and (7) final presentation of the functional SOA. Data collection occurred at two moments: (1) immediately after the event and (2) again six months later, to assess the long-term implementation of the solutions developed in clinical practice. Results: The hackathon successfully generated seven diverse SOAs targeting various clinical and non-clinical populations, including: a) stroke survivors, b) individuals with dementia (both institutionalized and community-dwelling), c) college students, and d) adults with intellectual disabilities. The resulting interventions exhibited high technical variability: three incorporated virtual reality with full-body interaction, two were tablet-based (touch-based interaction), one was computer-based (keyboard-based interaction), and one featured cross-platform compatibility. The specific SOAs demonstrated the platform's ability to support diverse therapeutic goals through activities ranging from semantic memory training to simulation of activities of daily living (ADL). Qualitative feedback from professionals was positive, highlighting the platform's versatility and offering important technical suggestions to enhance its performance. However, at the six-month follow-up, only two teams (28%) had successfully integrated their SOAs into daily clinical practice, reporting high patient engagement and session dynamism. Discussion and Conclusion: The findings validate the Musiquence platform as a highly customizable tool that meets the nuanced requirements of various rehabilitation contexts. The hackathon methodology proved to be an effective and rapid prototyping environment that can help overcome traditional barriers to innovation in healthcare by fostering interdisciplinary collaboration in user-centric design. While the six-month follow-up identified common barriers to adopting technological solutions—such as staff shortages and a lack of hardware—successful implementations confirmed the clinical viability of the co-created tools.

ABSTRACT. Background and Objectives: Cyberscams present a global threat with serious financial and psychological impacts. With the evolving sophistication of scammer manipulation and social engineering techniques and novel scams accelerated through artificial intelligence, anyone can become a victim. Characteristics relating to emotional, behavioural and social functioning commonly increase susceptibility. People with disability are an often-overlooked cohort in scam prevention efforts. Due to injury-related cognitive impairments and psychosocial changes, people with acquired brain injury (ABI) may be particularly vulnerable to cyberscam victimisation and require tailored vulnerability assessment tools and prevention supports. Identifying risk profiles and individual cyberscam vulnerability is important for early scam prevention, particularly as shame and denial often delay scam detection. To enable such evaluation, a new measure of cyberscam risk, The CyberAbility Scale was co-designed with scam survivors and validated for people with and without ABI. This study aimed to 1) determine whether people with ABI have greater risk of being scammed than people without ABI, and 2) explore demographic, injury, and psychosocial factors associated with cyberscam risk for people with and without ABI. Method: Using a cross-sectional design, participants with (n=149) and without (n=153) ABI completed an online questionnaire. Survey questions collected demographic information, prior scam experiences and technology use, and measures of cybersafety (The CyberAbility Scale, comprising a practical scam identification task and self-reported cybersafety questions), community integration (The Community Integration Questionnaire-Revised), loneliness (The Revised University of California Los Angeles Loneliness Scale), trust (trustworthiness items from The International Personality item pool), impulsivity (Short UPPS-P Impulsive Behaviour Scale), and mood (The Hospital Anxiety and Depression Scale). Descriptive statistics were summarised. Bivariate analyses and multiple regressions were conducted to examine demographic, injury, and psychosocial factors with cyberscam risk based on prior scam experiences and The CyberAbility Scale scores. Results: Participants with ABI reported approximately 50% higher rates of prior cyberscam experiences than participants without ABI. Participants with ABI performed worse on the scam identification items of The CyberAbility Scale than those without ABI. Older age was also associated with lower ability to identify scams for all participants. There were no significant differences in self-rated cyberscam risk score between ABI and non-ABI groups. Lower self-rated cyberscam safety on The CyberAbility Scale was significantly associated with higher loneliness, impulsivity, and mood symptoms, and lower trust and community integration. Higher loneliness was the strongest risk factor associated with self-reported cyberscam risk on multiple regression analyses. The relationship between loneliness and overall cybersafety was strongest for individuals with ABI compared with participants without ABI. Conclusions: Findings delineate a psychosocial risk profile for cyberscam vulnerability, with potentially modifiable psychosocial factors including loneliness, impulsivity, mood, trust, and community integration indicating targets for prevention efforts. Older adults and individuals with acquired brain injury (ABI) demonstrated greater difficulty detecting scam warning signs, underscoring the need for targeted cybersafety education for these vulnerable cohorts. Further, these results contribute to the evidence supporting the utility of The CyberAbility Scale. Overall, the results highlight the importance of accessible, cybersafety-specific education combined with broader social interventions that address underlying vulnerabilities exploited by scammers, particularly loneliness. Although cyberscam susceptibility is multifactorial, this study identifies a vulnerability profile that warrants monitoring and intervention, especially among individuals with ABI. These findings will inform the development of urgently needed, tailored prevention programs to better protect at-risk populations from increasingly sophisticated scams.

ABSTRACT. Introduction. Prolonged grief disorder (PGD) is a debilitating condition characterized by persistent separation distress, emotional dysregulation and functional impairment that extends beyond culturally expected mourning periods. While evidence suggests that internet and mobile-based interventions grounded in cognitive-behavioural principles can effectively reduce symptoms, there remains a gap in real-time monitoring and support. Although ecological momentary assessment (EMA) and ecological momentary intervention (EMI) have emerged as powerful strategies for monitoring affective states and deliver just-in-time adaptative support, there is limited applications of these methods specifically targeting PGD. Furthermore, few systems currently integrate continuous monitoring with multimodal, momentary grief-focused interventions. Objective. This study describes the development and usability testing of a mobile application (iGROw) specifically designed to support a self-administered online intervention (GROw) for individuals with PGD by integrating EMA and EMI. Method. A pilot study was conducted with a non-clinical sample of 7 women with an average age of 32.71, most of whom had higher education. Three participants used iOS devices (iPhone), while four used Android devices. The application was developed by mental health professionals specializing in PGD, adhering to best practices for EMA/EMI design. Key design decisions included minimizing questions to enhance adherence, providing clear rationales for assessments and utilizing multimodal delivery to accommodate varying literacy levels and preferences. The application has two main functionalities: brief daily questions (levels of grief intensity, sadness, anxiety, motivation and memories) and on-demand support. Based on their responses, the app offers short, tailored activities, short videos, images and text using approaches such as psychoeducation, expressive writing, cognitive reappraisal and behavioural activation. In addition, users can request support whenever they feel distressed by completing a quick check-in, which provides immediate content adapted to their current emotional state and helps offer timely relief. Participants utilized the application over a two-week period, completing daily EMAs and accessing the on-demand questions. This pilot usability study was conducted prior to testing the application in a clinical PGD population. Usability was assessed using the System Usability Scale (SUS) and specific questions regarding application features. Results. The results indicated very high perceived usability, with SUS scores ranging from 90 to 100 (M = 96.07). Most participants rated the app’s usefulness highly and indicated they would recommend it to people who have experienced the loss of a loved one. Navigation and instructions were described as clear and intuitive. While text-based instructions were easy to understand, two participants suggested incorporating audio or video formats to improve accessibility. Usage was reported as comfortable, with no participants finding the protocol burdensome. Qualitative feedback highlighted the quality and appropriate length of the videos, although some technical issues related to loading times were noted. Suggested improvements included optimizing video performance, increasing interactivity, and expanding resources to include recommended books, podcasts, and external links on grief processing. Conclusions. These preliminary findings suggest high perceived usability, clarity and acceptability of iGROw. This app is intended to be used in combination with a self-administered online intervention for PGD, within a research setting and under professional supervision, rather than as a standalone unguided tool. However, results should be interpreted with caution due to the small sample size and limited variability in participant characteristics, which may restrict the generalization of the findings.

ABSTRACT. Elite athletes experience mental health challenges at rates comparable to general populations, yet face unique vulnerabilities absent from standard clinical contexts. Mental health data in sport can influence team selection, contract negotiations, sponsorship decisions, and media narratives—creating institutional surveillance risks and coercion dynamics. AI-powered applications for psychological support increasingly address mental health and wellbeing concerns alongside performance enhancement, including stress management, anxiety reduction, and emotional regulation. This convergence blurs boundaries between sport psychology and clinical intervention, yet no validated instrument exists to evaluate ethical adequacy of these technologies for sport populations. Existing AI ethics frameworks developed for general healthcare or consumer contexts inadequately address athletes' dual roles as patients requiring confidential care and employees whose data informs high-stakes institutional decisions. Recent systematic evaluation identified 28 sport-specific psychological skills training apps, yet none have undergone ethical assessment for athlete data protection. This work-in-progress study addresses this gap by developing and preliminarily validating a sport-specific ethical assessment framework for AI-powered psychological support technologies. We conducted a conceptual synthesis of four international AI ethics frameworks to develop a 40-point assessment instrument with sport-contextualized criteria. The synthesis integrated principles from global health AI governance guidelines, computational psychiatry ethical frameworks, professional psychology guidance, and public sector AI implementation standards. From this synthesis, we operationalized six ethical principles into 40 discrete evaluation criteria: (1) Privacy & Confidentiality (8 points): Data minimization; third-party disclosure; institutional access prohibitions; athlete-controlled deletion. (2) Informed Consent & Autonomy (7 points): Plain-language consent; granular opt-out provisions; voluntary vs. mandatory data distinctions; consent revocation mechanisms. (3) Transparency & Explainability (7 points): Algorithm disclosure; AI-generated vs. human-curated content labeling; clinical validation evidence; limitation disclosures. (4) Accountability & Scientific Integrity (6 points): Liability specification; complaint and redress mechanisms; ethics review evidence; data export capabilities. (5) Safety & Crisis Response (7 points): Suicide/crisis detection with human referral protocols; prevention of harmful gamification; inclusive content addressing diverse athlete populations. (6) Fairness & Non-Discrimination (5 points): Bias testing and mitigation; equitable access; uniform algorithmic processing. Each criterion receives a binary or scaled score. Critical violations override numerical scores, classifying applications as "high risk" regardless of total points. Three sport-specific applications underwent evaluation, selected from recent systematic quality assessment: Perform - AI Mindset Coach (highest-rated app, MARS 4.19/5, AI chatbot with CBT components), Champion's Mind (MARS 3.77/5, mindfulness and stress management), and Nui Sport (MARS 2.11/5, AI chatbot for stress management). This sample represents apps where performance psychology and mental health support intersect. Evaluation involved Terms of Service and Privacy Policy review, interface examination, feature testing, and documentation. Preliminary assessment reveals systematic ethical deficiencies across quality tiers. High-quality apps demonstrated marginally better privacy practices but exhibited similar critical violations as low-quality apps. Privacy failures: all apps used vague third-party language enabling undisclosed institutional access; none explicitly prohibited coach/team/federation access; most retained data indefinitely or required contacting support for deletion. Consent deficiencies: all employed bundled agreements obscuring psychological data implications; none offered meaningful opt-out without substantial feature loss; most used dark patterns. Transparency gaps: all provided zero algorithm disclosure beyond marketing terms; most conflated AI-generated content with human expertise; none cited peer-reviewed validation studies. Safety concerns: most relied on passive crisis disclaimers despite addressing stress, anxiety, and emotional struggles; most incorporated gamification potentially reinforcing performance anxiety. Critical violations include unnamed data-sharing partners, institutional access provisions, and liability disclaimers for psychological outcomes. Preliminary validation demonstrates that sport-specific psychological support applications systematically fail to address athlete-protective ethical requirements, with app quality not predicting ethical compliance. The developed framework provides sport psychologists and athlete welfare officers with evidence-based evaluation criteria. Full validation with expanded sample will be completed prior to conference presentation.

ABSTRACT. Background: There has been rapid growth in the area of digital mental health with between 10,000 and 20,000 digital resources available, but a majority of these are not informed by evidence or have not been empirically validated. In addition, most mental health resources are targeted primarily at depression and anxiety, with few other mental health conditions represented. Body image refers to one’s evaluation, perception, and behaviours related to their body and negative body image is a critical global health concern associated with the onset of mental health conditions such as depression and eating disorders across the lifespan. Body dissatisfaction is highly common, particularly in young adults, with reports suggesting that over 70% of emerging adults globally report dissatisfaction with their weight, shape, and appearance. Currently, while some digital interventions have been developed in research contexts, there are no apps available to the public which aim to reduce body dissatisfaction or improve positive body image and there are a limited number of body image interventions which have digital adaptations. Furthermore, while digital interventions are shown to be effective, they report large rates of participant attrition. Recognising the need for a multipronged approach to support young adult body image, our resource – Attune – will combine, for the first time, complimentary, evidence-based approaches and explore the potential utility and perceived acceptability of incorporating artificial intelligence (AI) to enhance personalisation, engagement and facilitate progress monitoring (i.e., track their progress in improving their body image). In order to understand how to best implement these innovative features, as well as to investigate stakeholder perspectives on how to optimise engagement, inclusivity, acceptability, reach, and accessibility, we will run a series of co-design workshops with key stakeholders (i.e. young adults, researchers, clinicians). Methods: Four co-design workshops will be run with key stakeholders: our lived experience group comprised of young adults with experiences of positive and negative body image (N=12) and our learned experience group comprising researchers, practitioners and clinicians in relevant fields such as body image, digital intervention, and human-computer interaction (N=6). Workshops one and two will be conducted with young adults and researchers/clinicians independently in order to co-design key content, delivery modalities and features to be included in the resource with a particular focus on how we can feasibly incorporate novel techniques such as AI and progress tracking to enhance engagement. In workshop three, we will bring all stakeholders together to engage in iterative prototype design and development. Workshop four will allow stakeholders to engage with a working prototype of the resource and provide feedback on its design, usability, and content. Progress: We are currently recruiting participants for the workshops which are due to take place from March 2026. By the end of June 2026 we aim to have completed three of the four workshops. We will be able to present on discussions from completed workshops surrounding development of a digital intervention resource aimed at young adults from the perspective of young adults and professionals in body image, digital intervention, and human-computer interaction fields. Expected impact: The results of this study will not only inform the development of a digital body image resource, but will also provide valuable insights into the expectations and desires of the end user for digital health resources, particularly in relation to novel technological advancements such as AI, and innovative features in this field such as personalisation and progress tracking. These workshops will advance our understanding of designing ethical, human-centred digital resources which facilitate real-world behavioural, cognitive, and affective change, as well as inform how innovative features such as AI can be incorporated to improve engagement and accessibility.

ABSTRACT. Vietnam is undergoing one of the world’s most rapid demographic shifts toward an aging society, creating urgent needs for sustainable healthcare innovations. This project centered on the ethical design and integration into daily life of a digital health app, promoting the well-being and autonomy of a vulnerable population in an under-served demographic group. Medication non-adherence among older Vietnamese adults—driven by polypharmacy, cognitive decline, and fragmented support systems—poses significant risks to health outcomes and healthcare costs. While digital health tools such as medication management apps show promise in Western contexts, their applicability in non-Western, middle-income settings remains understudied. This project aimed to identify the usability and acceptability of existing medication apps for elderly Vietnamese and co-design a culturally appropriate mobile application to improve adherence.

Using a mixed-methods, community-engaged approach, we conducted: (1) a systematic review of global and local medication apps; (2) a quantitative survey of 300 elderly patients at a national geriatric hospital, incorporating the Morisky Medication Adherence Scale (MMAS-8) and measures of health status and digital readiness; and (3) in-depth qualitative interviews with 16 patients and 7 healthcare professionals to explore lived experiences, barriers, and feature preferences.

Survey results indicated a higher-than-expected adherence rate (70%) within the hospital sample, though significant challenges persisted regarding multi-dose regimens and drug interactions. Most participants expressed openness to using a mobile app, provided it was simple, voice-enabled, and available in Vietnamese. Qualitative themes highlighted desires for audible reminders, family notification features, clear medication schedules, and interaction warnings—all within an intuitive, low-literacy interface.

Guided by user-centered design principles, we developed a high-fidelity prototype prioritizing localization, accessibility, and emotional reassurance. Key design innovations include spoken-time reminders, visual medication logs, caregiver alert systems, and culturally resonant visual cues. The project underscores the critical role of co-design and contextual adaptation in digital health interventions for aging populations in non-Western settings.

This research contributes to the "Human 5.0" vision by demonstrating how behaviorally-informed, ethical technology, developed through deep community partnership, can promote well-being and independence. By illuminating how elderly users in a transitioning economy perceive, trust, and engage with mobile health technologies, it serves as a case study in creating digital tools that are truly integrated into the social fabric, respecting cultural norms, ethical data use, and the human need for connection and support. Future work will involve usability testing of the prototype and pursuit of implementation funding to scale the solution in community and clinical settings. Our approach offers a critical blueprint for developing ethical, human-centered cyberpsychology interventions in globally aging societies.

ABSTRACT. A growing body of research in positive and cyber psychology shows that everyday behaviors, especially self-care and prosocial kindness, are strongly associated with well-being, resilience, and mental health. Self-care practices (e.g., sleep hygiene, mindful breaks, boundary setting, restorative routines) help regulate affect and reduce stress reactivity, while random acts of kindness foster meaning, belonging, and positive emotions that broaden cognitive and social resources. Despite this evidence, many people struggle to translate these insights into consistent daily habits. Digital environments, however, offer scalable opportunities to nudge, track, and reinforce such micro-behaviors in real time. This study examines whether a 30-day web-app–based behavioral tracking intervention focused on self-care and kindness can produce measurable improvements in mental health, indexed by changes in flourishing.

We developed a web based application that prompted participants to record two daily actions for 28 consecutive days: (1) at least one self-care behavior and (2) at least one random act of kindness toward another person. Participants (N = 100) were recruited online and completed a baseline survey before beginning the program. The app provided daily reminders, simple logging interfaces, and brief reflective prompts to enhance intentionality. No therapeutic content was delivered; instead, the intervention relied on behavioral activation and self-monitoring supported by digital prompts.

Mental health was assessed using the Flourishing Scale at pre-test (Day 1) and post-test (Day 30). Additional exploratory measures included perceived stress, positive affect, and self-compassion. Compliance data (daily logs, completion rates) were automatically captured by the app. A pre–post design was used to evaluate changes in flourishing. Paired-samples t-tests and effect size estimates (Cohen’s d) were computed. We also explored whether frequency of logged behaviors predicted magnitude of improvement.

As data collection is currently underway, the planned analyses will examine whether there is a statistically significant increase in Flourishing Scale scores from pre-test to post-test, with an expected moderate effect size. We will also test whether higher frequencies of logged self-care and kindness behaviors are associated with greater improvements, indicating a possible dose–response relationship. In addition, participants’ written logs describing their self-care and kindness activities will be subjected to text analysis to identify meaningful thematic categories and linguistic patterns. This analysis will explore whether certain types of activities or expressions are more strongly associated with improvements in flourishing and related psychological indicators.

In terms of practical implications, this study demonstrates how simple, low-cost digital tools can translate psychological theory into daily practice. Rather than delivering psychoeducation or therapy, the intervention leveraged behavioral tracking, prompts, and reflection to support sustainable habits. Such web-based micro-interventions can be easily adapted for schools, workplaces, and community settings to promote preventive mental health at scale.

The findings suggest consistent engagement in small self-care and kindness behaviors, when scaffolded by digital tools, can meaningfully enhance flourishing. This research contributes to cyber psychology by showing how web applications can function as behavioral ecosystems that reinforce prosocial and self-regulatory habits. It extends digital well-being research beyond screen-time concerns to examine how technology can actively cultivate human flourishing. By integrating self-monitoring, behavioral activation, and prosocial action within a digital platform, this study offers a replicable model for technology-assisted mental health promotion. In an era where digital technologies often contribute to distraction and fatigue, this work highlights their potential to reconnect users with intentional, meaningful daily behaviors that support both personal and collective well-being.

ABSTRACT. Introduction: Digital mental health technologies offer new opportunities to design human-centred, behaviourally informed interventions that integrate real-time data and personalization to promote wellbeing. Borderline Personality Disorder (BPD) represents a particularly relevant context in which to explore the potential of ethical and adaptive digital interventions, given its core features of emotional dysregulation, impulsivity, and interpersonal difficulties, which often result in persistent psychosocial impairment. Despite recent technological advances, most digital tools developed for BPD remain primarily symptom-focused and are closely aligned with deficit-based models, leaving positive psychological processes largely underexplored. In contrast, Positive Psychology Interventions (PPIs) provide a promising framework for complementing problem-focused models by targeting strengths, positive affect, and personal values. Within this perspective, brief, low-intensity digital microinterventions -defined as short, context-sensitive exercises designed to promote adaptive psychological processes in daily life- may represent a promising adjunct to traditional treatments by fostering both hedonic (positive affect, life satisfaction) and eudaimonic (meaning, values, personal growth) wellbeing.

Objective: This study aimed to examine the feasibility, acceptability, usability, and preliminary effectiveness of a personalized smartphone-based PPI in individuals with BPD, using a mixed-methods design to assess both symptom reduction and enhancement of wellbeing.

Methodology: A pilot randomized controlled trial was conducted in which participants with BPD were randomly assigned to an experimental group (n=50) or a waiting-list control group (n=94). The experimental group received access to IMPULSA, a six-week smartphone-based PPI consisting of more than 40 brief microinterventions. The program was structured into three thematic modules -strengths, values, and positive affect- with a personalized sequence of components tailored to individual profiles and behavioural data. Participants engaged with the microinterventions in two ways: through daily recommended exercises delivered via notifications and through self-directed, on-demand access to the content. Quantitative outcomes were assessed using repeated-measures designs, including distal outcomes (pre-post intervention) and proximal effects (after each microintervention). Psychopathology was measured with the Patient Health Questionnaire-9, the Generalized Anxiety Disorder Scale-7, and the Borderline Symptom List-23; Wellbeing was measured with the Hedonic and Eudaimonic Wellbeing Scale and the Pemberton Happiness Index. Strengths were measured using the Strengths Knowledge Scale and the Strengths Use Scale, while values were measured with the Engagement in Life Scale and the values dimensions of the Multidimensional Psychological Flexibility Inventory. Proximal data were collected on wellbeing, symptomatology, values, strengths, and positive affect weekly and after each microintervention. Qualitative data were collected through semi-structured interviews exploring user experience.

Results: Significant group x time interactions were observed across multiple outcomes, with the experimental group showing greater pre-post improvements than the control group in depressive symptoms (ηp²=.042), anxiety (ηp²=.170), BPD symptomatology (ηp²=.070), wellbeing (ηp²=.151), positive affect (ηp²=.031-.056), strengths aspects (ηp²=.054–.068), and values aspects (ηp²=.07–.105). Qualitative findings converged with quantitative results, with participants generally emphasizing the usefulness of the intervention, its positive emotional impact, and the successful integration of the microinterventions into daily life. Proximal data are currently being analysed.

Conclusions: Results suggest that IMPULSA platform is a feasible, acceptable and potentially effective complement to traditional treatments for individuals with BPD. Beyond symptom reduction, the intervention appears to facilitate the activation of positive psychological processes related to wellbeing, strengths, and values. The mixed-methods approach further underscores the value of integrating quantitative outcomes with user experience data to inform the design of ethical and effective digital interventions. Overall, these findings highlight the potential of personalized smartphone-based PPI microinterventions as a human-centred digital approach to mental health care, bridging positive psychology, clinical practice, and behavioural technology.

ABSTRACT. Introduction:

Virtual reality (VR) has increasingly been explored as a supportive tool in palliative and end-of-life care across a range of clinical, psychosocial, and educational contexts. Given the growing use of virtual reality (VR) in these contexts, it is important to explore not only the usability of VR interventions but also their impact on feelings of safety, satisfaction, and overall user experience with VR content specifically designed for this population.

Purpose:

The aim of this study is to explore the usability, feelings of safety, and satisfaction associated with the Canadian VR platform “Come With Me” and the Czech-developed VR application BreezeTerraVR. The two platforms differ in their content, with one focusing on real-life scenarios and the other using animations to guide relaxation.

Method:

Ten French-speaking family caregivers of individuals in palliative care (mean age 56.8 years, SD = 22.8, range 22–78, 70% males) tested the Canadian VR platform “Come With Me” in an end-of-life unit in Montreal using a Pico 3 headset. Sessions lasted an average of 8 minutes and 42 seconds (SD = 0.5 minutes). In addition, four French-speaking family caregivers (mean age 56.2 years, SD = 16.8, range 37–78, 50% females) tested the Czech-developed VR application BreezeTerraVR in the same setting using Pico 3 goggles, with sessions lasting approximately five minutes. The two groups were equivalent in terms of age. Following the VR experiences, participants completed the 10-item System Usability Scale (SUS) and 11-point visual analog scales to assess their perceived satisfaction and sense of security with the technology.

Results:

The Canadian VR content “Come With Me” had a statistically significantly longer session duration than the Czech-developed VR application BreezeTerraVR (p < .05). A Mann-Whitney U Test revealed no significant differences in usability levels between family caregivers who used The Canadian VR content “Come With Me” (Md = 93.7, n = 10) and those who used the Czech-developed VR application BreezeTerraVR (Md = 92.5, n = 4), U = 18, z = -2.86, p = .78, r = 0.8. Similarly, a Mann-Whitney U Test revealed no significant differences in safety levels between the “Come With Me” Group (Md = 0, n = 10) and the BreezeTerraVR Group (Md = 0, n = 4), U = 12, z = -1.43, p = .3, r = 0.4. Finally, a Mann-Whitney U Test revealed no significant differences in satisfaction levels between family caregivers using “Come With Me” (Md = 10, n = 10) and those using BreezeTerraVR (Md = 10, n = 4), U = 22, z = .35, p = .84, r = .09.

Conclusions:

This pilot study suggests that both types of VR content exhibit excellent feasibility, along with high levels of perceived safety and satisfaction, among family caregivers of individuals in palliative care following the VR sessions. Both platforms appear to be safe, user-friendly, and promising tools to be used among family caregivers in palliative care settings. Ethically, the use of VR in this context highlights the need to ensure informed consent, safety, and respect for caregivers’ vulnerability in palliative care settings.

Funding:

With financial assistance provided by the Natural Sciences and Engineering Research Council of Canada (NSERC), and Gouvernement du Québec (Ministère de l’Économie, de l’Innovation et de l’Énergie) in Quebec, Canada. Dr. Moreno is supported by Fonds de recherche du Québec (https://doi.org/10.69777/376160), by an AGE-WELL-EPIC-AT Fellowship, and the Réseau Québécois de Recherche sur le Vieillissement (RQRV), a Research Network financed by Fonds de recherche du Québec. Dr. Fajnerova is partially supported by the project Cooperatio Neurosciences, Charles University, Third Faculty of Medicine. BreezeTerraVR development was funded by DigiWELL CZ.02.01.01/00/22_008/0004583, co-financed by the European Union.

ABSTRACT. A cancer diagnosis affects not only the body but also the psychological well-being of patients. In recent years, the number of new diagnoses has increased and survivorship has lengthened, making psycho-oncological support increasingly essential. This makes it difficult for patients to access services and cover costs. Digital platforms dedicated to psycho-oncology therefore represent a promising solution to expand access to care. This pilot study aimed to examine patients’ intention to use, and perceived quality of a digital psycho-oncology platform, as well as to explore their preferences regarding the desirable characteristics of therapists working on it. Participants (n = 33), women in diverse disease phases, had largely received prior psycho-oncological support. Findings indicate good preliminary acceptance of the platform. Perceived quality measured via The Mobile Application Rating Scale (MARS) was moderate (M=10.03, SD=2.35). The Technology Acceptance Model (TAM) components showed medium-to-high scores: perceived usefulness (M=14.42, SD=4.10), perceived ease of use (M=15.36, SD=3.57), and intention to use (M=7.12, SD=2.27). The total TAM score was also high (M=36.99, SD=8.63), suggesting that the platform is perceived as useful, simple, and potentially adoptable, but with significant individual differences. In addition, the study explored whether distinct psychological profiles could be identified among potential users. To this aim, a cluster analysis was performed on the main psychometric variables (BFI, Mini-MAC, MHLC, TAM), revealing a three-cluster solution. Cluster 1 represented a group of psychologically adaptive and proactive patients, characterized by low distress, higher openness and conscientiousness, flexible coping strategies, and a balanced health locus of control. This group also showed the highest acceptance of the platform, with elevated scores in perceived usefulness, ease of use, and intention to adopt the app. Cluster 2, in contrast, included patients with low distress and high emotional stability, displaying the lowest levels of anxious preoccupation and helplessness. Despite their psychological well-being, these participants showed the lowest perceived usefulness and intention to use the platform, possibly reflecting a reduced perceived need for additional support. Finally, Cluster 3 consisted of patients with heightened psychological vulnerability, marked by elevated anxiety, hopelessness, and avoidant coping, as well as a more external health locus of control. Despite their higher distress, these participants demonstrated a relatively strong intention to use the platform, suggesting that individuals facing greater emotional burden may regard digital psycho-oncology tools as particularly valuable resources. Together, these clusters highlight that technological acceptance is not uniform across patients but is closely intertwined with their psychological conditions. Digital platforms may therefore serve distinct functions for different subgroups. Participants also answered two open-ended questions about the desired characteristics of the ideal psycho-oncologist. Open-ended responses emphasize that despite the positive technological acceptance, the therapeutic relationship remains deeply rooted in human elements. Desired characteristics included empathy (14/33), experience in the field (8/33), clarity and understanding (4/33), socio-demographic or personal characteristics (6/33), and other traits (1/33). Undesirable characteristics included judgment (11/33), coldness (6/33), lack of experience (3/33), negative personality traits (10/33), distraction (2/33), and other aspects (1/33). The findings of this study show that digital technologies can support psychological well-being only when they remain firmly grounded in the human values they aim to enhance. An ethically designed, human-centred digital psycho-oncology ecosystem can expand access to support without compromising its relational core. The future of mental health care in oncology is therefore not a choice between the human and the technological, but their thoughtful integration: technologies that amplify, rather than replace, what makes care deeply human.

ABSTRACT. Background The growing adoption of digital health tools in clinical settings offers novel ways to support psychological screening and assessment, symptom monitoring, and patient engagement. Narrowing to the oncological context, different technologies can be used to promote patients’ well-being, who could experience psychological and cognitive difficulties extending the emotional strain to their family members. This mixed-method study, conducted within the ALTHEA project, aims to explore stakeholders’ opinions regarding a mental health digital platform to be integrated in cancer care pathways.

Methods A mixed-methods design employing surveys and focus groups was conducted across three stakeholder groups to identify barriers, needs, and preferences related to online psychological-cognitive screening and support. In total, 40 participants were recruited through convenience sampling: fourteen Italian cancer patients (12 F, 55.93 ± 13.18 years old), with and without prior psychological support, 7 caregivers (2 F, 55.86 ± 13.79 years old) and 19 healthcare professionals (HCPs) (14 F, 43.61 ± 11.23 years old) including oncologists, nurses, psycho-oncologists and allied health professionals. Six focus groups were performed online via Microsoft Teams with the same stakeholder groups and participants. Questionnaire data, including Likert-scale items, open-ended and closed-ended questions, were analysed using descriptive statistics. The interviews were audio recorded and transcripts of focus groups were analysed using inductive thematic analysis guided by predefined domains related to platform use, needs, and barriers. The integration of quantitative and qualitative findings was performed at the interpretation stage to identify convergences and complementarities.

Results Survey results reveal that the majority of cancer patients (CP) and caregivers (CG) (76.2%) expressed interest in engaging with a digital platform providing resources for psychological or cognitive support. Preferred features include: online psychotherapy (CP: 71.4%, CG: 66.7%), cognitive support (CP: 78.6%), symptom monitoring (CP: 85.7%) and educational resources (CP: 71.4%). However, only 38.1% of respondents reported potential usability concerns. HCPs’ surveys indicate that the introduction of new digital tools (94.7%) and workshops on mental health topics (84.2%) could represent potential opportunities to ease mental health support. Overall, the platform was perceived as a possible safe and reliable space for sharing experiences and information, emotional and cognitive support, and suggestions to dedicated services. Patients indicated the need for guidance in navigating resources, direct communication with the clinical team, reminders for pills or other treatments, cognitive exercises, monitoring tools, and self-help content. Regarding online psychological support, concerns related to technological barriers and a reduced perception of empathy emerged among patients and caregivers. Caregivers emphasised the importance of communicating with a common language, of emotional anchoring and of the role of institutional credibility in the use of digital tools. All the participants displayed minimal privacy concerns despite occasional initial resistance to the use of technology. HCPs stressed the need for the integration of agile screening tools into routine practice, patient pathway updates, tailored assessment items, and features such as professional networks, training opportunities, and clinical data availability.

Conclusions Current findings should be interpreted considering methodological limitations, including a small and not completely representative sample and the exploratory nature of assessing perceptions of a digital platform without an actual prototype. Indeed, current results are intended to inform early-stage co-design rather than predict actual utilisation or engagement. Additionally, the preferences for services such as online psychotherapy raises ethical considerations regarding quality, safety, and regulation of digitally delivered psychological care. Future research should focus on iterative prototyping and user testing to translate the identified preferences and needs into concrete functionalities and evaluate real implementation.

ABSTRACT. BACKGROUND: New technologies and digital solutions have become an integral part of oncological care. Informational videos are increasingly used to support patients throughout the care pathway by facilitating access to information. Recent advances in oncology promote the use of genomic testing (Gene Expression Profiling test, GEP) for tailoring therapeutic interventions for breast cancer patients. Effective communication of this test is essential to improve patients’ understanding and support shared decision-making in clinical consultation. However, communicating GEP results can be challenging due to uncertainty, poor understanding of risk and relevance of chemotherapy in reducing this. On these basis, this study aims to evaluate whether a short (8 minute) informational video about GEP testing developed by researchers and clinicians can enhance patients’ comprehension and satisfaction, and to examine the role of psychological variables such as decisional conflict and intolerance of uncertainty.

METHODS: A total of 53 Italian patients with early breast cancer who consented to undergo GEP testing were recruited and randomly assigned to either Group A (n = 28) who received standard GEP consultation, or Group B (n = 25) who received standard consultation plus an informational video about the GEP test. Patients’ satisfaction with the clarity, amount, and comprehensibility of the information was assessed post-consultation in both groups. Trait anxiety was measured at baseline, state anxiety before and after consultation, and intolerance of uncertainty pre-consultation, while knowledge of the GEP test and decisional conflict were assessed post-consultation. T-tests were conducted to compare satisfaction and comprehension between groups. Correlation analyses examined relationships between satisfaction, comprehension, and psychological variables, including anxiety, intolerance of uncertainty, and decisional conflict.

RESULTS: T-test analysis revealed that the satisfaction related to the clarity, amount, and comprehensibility of the information received during the consultation did not differ between participants in both Group A and B. However, correlation analysis conducted on participants who received usual consultation highlighted that patients who perceived receiving a higher quantity of information tended to report higher decisional conflict (r=.67, p<.001), higher levels of intolerance of uncertainty (r=.44, p<.05) and post-consultation state anxiety (r=.60, p<.01). Furthermore, post-consultation state anxiety was found to be positively correlated with decisional conflict (r=.68, p<.001). If patients perceived information to be easily comprehensible, this was negatively related to intolerance of uncertainty (r=-.46, p<.05) in participants who viewed the informational video. Furthermore, a positive correlation persisted between post-consultation state anxiety and decisional conflict (r=.43, p<.05). Regarding the GEP procedure comprehension, t-test results highlighted no statistically significant differences (p=.08) among the two groups, even if participants who viewed the video tend to show higher level of knowledge (M=4.96, SD=2.85) compared to patients who received usual consultation (M=3.60, SD=2.79). Interestingly, correlation analysis highlighted that, after viewing the informational video, the knowledge regarding the GEP procedure was positively related to the satisfaction of the quantity of information received (r=.427, p<.05).

CONCLUSIONS: Integration of structured audio-visual aids are a potentially important supplementary tool that can optimize the doctor-patient consultation regarding specific clinical procedures, such as the genomic testing consultation. Enhancing objective knowledge regarding GEP test mechanisms and the rationale for adjuvant chemotherapy appears to be a critical factor in supporting patient decision-making and improving overall emotional well-being.

ABSTRACT. #Background Entrainment refers to the process by which two or more individuals align their perceptions, attention, movements, and/or emotional states through shared rhythmic cues (Phillips-Silver et al., 2010). It plays a key role in joint actions such as clapping at concerts or dancing with others and is closely linked to enjoyment and social bonding. Entrainment is relevant in virtual reality (VR), where collaborative, interactive social activities are prevalent (Freeman & Acena, 2021). From a positive technology perspective, social VR dance is particularly noteworthy because it integrates physical movement, which may support social bonding, while also being influenced by factors such as latency and sensory mismatches. Current research has primarily focused on the VR dance experiences of younger adults (e.g., Piitulainen et al., 2022), while insights from older adults are lacking.

#Research Aim Against this background, the present study addresses the following research questions (RQ): RQ1: How do younger and older adults experience and enjoy social dancing in VR? RQ2: How do younger and older adults experience entrainment in social VR dance? RQ3: Which entrainment dimensions (sensory consistency, cognitive engagement, motor entrainment, emotional entrainment) influence the overall entrainment perception of younger and older adults in social VR dance?

#Methods ##Design and sample: Given the limited research on the topic, a qualitative study design was adopted to capture in-depth experiences (Bryman, 2016). Twenty participants living in Germany took part in the user experience laboratory study (11 younger and 9 older adults). Eligibility criteria included sufficient German or English language proficiency, physical ability to dance, and no prior VR dance experience. Younger adults were aged 21–36 years (M = 27.4, SD = 4.7), older adults 60–82 years (M = 72.6, SD = 6.9).

##Procedure and data collection: Participants were recruited in November 2025 and invited to the laboratory. After receiving study information, they provided informed consent and demographic information. They then received an introduction to the VR equipment, including fitting and calibration. Meta Quest 3 headsets were used with full-body tracking enabled via Vive trackers. The social VR dance experience took place in a dedicated virtual room in VRChat.com with preselected humanoid animal avatars, where each participant danced together with the researcher to three songs, with the first song self-selected. Participants engaged in a non-contact partner dance session in which the researcher guided predefined choreographed movements and the participant was instructed to follow in synchrony. After the dance session, a semi-structured interview was conducted.

##Data analysis: Interview transcripts were analyzed using qualitative content analysis with MAXQDA (Rädiker & Kuckartz, 2020). The study was approved by the university’s ethics committee and conducted as part of a larger funded research project (both blinded for peer review).

#Preliminary Results Regarding RQ1, older participants perceived the VR dance experience as more enjoyable than the younger participants, describing it as “fun” and “a new experience.” Younger adults reported experiencing less social pressure in the VR dance scenario compared to dancing in a club setting (“with avatars, there is less social awkwardness”). Regarding the entrainment experience (RQ2), most participants reported that synchrony was achieved through conscious effort to follow their dance partner’s movements (“I think it was kind of conscious, like, she’s raising her arms now—so I should raise mine, too. And then I started to fall into the rhythm”). Both younger and older participants stressed emotional entrainment as a core element of their entrainment experience supported by the perception of mutual fun, eye contact, and smiles of avatars (RQ3).

#Conclusion Social dancing in VR may foster social connectedness and well-being across generations and distances. Better understanding these experiences can inform both users and system developers.

ABSTRACT. This text advances a theoretical framework in which motor control is understood as a probabilistic **“sensorimotor bet”** continuously placed by the central nervous system (CNS) on the sensory consequences of its actions. Actions are not specified as deterministic trajectories but emerge from an inferential process that combines prior beliefs with incoming sensory evidence to estimate the current and future state of the body and the world. Building on Bayesian brain models, each motor command is treated as an explicit probabilistic forecast under uncertainty, reshaping motor control from a descending chain of commands into an inherently predictive system. In this view, uncertainty becomes the central organizing parameter of motor behavior, as the CNS evaluates distributions over possible states rather than single-point solutions.

The “bet” metaphor is proposed as structurally accurate: like a gambler evaluating expected risks and gains, the motor system computes expected cost functions that integrate performance error, variance, and energetic expenditure. Motor style is defined as the idiosyncratic way each individual “bets with their body,” reflecting specific priors, cost functions, and histories of interaction. At the same time, the framework foregrounds communal constraints that all humans satisfy to govern motricity, such as maintaining the center of mass within the base of support during upright stance and compensating for gravitational forces. This duality between style and communality offers a principled way to relate inter-individual variability to species-wide invariants in motor control.

Efference copy is a central mechanism for implementing and updating these sensorimotor bets, functioning as the neural vector through which the motor bet is rendered explicit and compared with sensory reality. Whenever a motor command is issued, an internal copy is generated that enables the system to predict sensory consequences before the movement unfolds. This prediction is characterized as a forward model integrating the dynamic properties of the effector, relevant environmental characteristics, the history of previous bets and their outcomes, and abstract knowledge about the laws governing the environment and the intended act. Through this mechanism, the system can pre-emptively attenuate predictable sensory consequences of self-generated actions while remaining sensitive to unexpected events.

On this basis, efference copy supports the distinction between reafference (sensory consequences of one’s own actions) and exafference (externally generated stimuli). Within a Bayesian filter framework, predictions derived from efference copy act as strong priors that reduce sensory surprise when precise, yet amplify the salience of external stimuli when prediction errors exceed an expected precision threshold. This selective modulation of sensory precision allows the organism to preserve perceptual stability in the face of self-generated perturbations while rapidly detecting external events that may demand corrective action or reorientation. The sensorimotor bet is thereby continuously updated as prediction errors are evaluated and incorporated into the internal models guiding future actions.

Internal representations are the cognitive infrastructure that provides the reference frame for these bets. Rather than static copies of the environment, internal models are dynamic, generative structures encoding probabilistic relationships between states, actions, and sensory consequences. They organize prior beliefs about bodily dynamics and environmental regularities, regulate how new evidence is weighted, and constrain the actions considered plausible or efficient in a given context. Through learning, the outcomes of past bets reshape these representations, progressively refining the mapping between motor commands and expected consequences.

Taken together, this framework reframes motor control as the ongoing resolution of a sensorimotor wager under uncertainty, implemented through predictive mechanisms such as efference copy and grounded in richly structured internal representations. It proposes a unifying language for describing how universal biomechanical and physical constraints coexist with individual motor styles, and how perception and action are coupled through prediction, error evaluation, and model updating.

ABSTRACT. Neural electrodes are key component of advanced brain–computer interface (BCI) systems and responsible for establishing connections between the biological nervous system and external devices. Compared to traditional electrodes, stretchable and tissue-compliant neural electrodes can significantly reduce potential damage to surrounding brain tissue during movement and achieve stable neural signal acquisition under mechanical deformation and long-term operation. We developed highly stretchable neural electrode arrays based on liquid metal–polymer conductors by embedding liquid metal within soft polymer substrates. These electrodes are capable of maintaining stable electrical performance under large mechanical deformation, with stretchability reaching approximately 100% strain, and supports high-quality electrophysiological recording in vitro and in vivo. The platform enabled consistent detection of neural activity across a wide range of signal amplitudes and frequency bands, including the recording of pathological neural patterns such as epileptiform discharges. We further integrated biodegradable substrate materials with liquid metal to obtain a biodegradable and flexible neural electrodes. These electrodes exhibit several characteristics including low mechanical constraint, stable signal transmission under deformation, and suitability for both surface and depth-related neural recordings. Moreover, they can be degraded in controllable biodegradation rate. The stretchable electrode arrays are further combined with miniaturized neural signal acquisition electronics to form an integrated BCI platform. The resulting BCI platform enables low-noise, multichannel electrophysiological recording, and minimize the impact of rigid interconnects and bulky hardware, achieving continuous neural monitoring during natural movement and extended operation.

ABSTRACT. Quantitative assessment of muscle oxygen metabolism requires simultaneous measurement of tissue absorption and scattering coefficients to derive absolute hemoglobin and myoglobin concentrations. Traditional continuous-wave near-infrared spectroscopy (CW-NIRS) is fundamentally limited to measuring only relative changes in oxygenation signals and lacks spatial depth discrimination capability. Frequency-domain NIRS (FD-NIRS) overcomes these limitations by measuring both phase delay and modulated amplitude attenuation of light, enabling absolute quantification of tissue optical properties and multi-depth spatial resolution. This study presents a comprehensive wearable high-density FD-NIRS system integrating optimized hardware design, probe arrangement, and physiological modeling. Hardware System Development: A self-developed dual-wavelength FD-NIRS system (650 nm and 830 nm) was engineered featuring: (1) Laser modulation driver using OPA690 operational amplifier (230 MHz bandwidth) for stable frequency-modulated output; (2) High-sensitivity photodetection combining avalanche photodiode (LSSAPD9-500) and low-noise transimpedance amplifier (TIA, 200 MHz bandwidth); (3) Signal conditioning chain with 75 MHz bandpass filter (SXBP-75+) and frequency down-conversion mixer (ADE-1L+); (4) 150 V stable DC power supply for APD biasing; (5) Embedded control platform (STM32F4 microcontroller) for real-time synchronous signal acquisition, digital phase demodulation, and wireless data transmission. The system achieves wearability through miniaturized, high-integration design optimized for continuous on-body monitoring (SNR >80 dB at 30 mm; >42 dB at 50 mm tissue path). Probe Optimization and Experimental Validation: Six probe configurations (4–188 measurement channels) were systematically evaluated using Monte Carlo photon transport simulations (MCX) and validated experimentally on tissue-equivalent phantoms (AB epoxy with carbon powder and TiO₂). Multi-distance measurements (1.0–2.7 mm) quantified optical parameter inversion capability via the slope method. Results revealed the medium-density cross-pattern layout (G-III) achieved optimal comprehensive performance with area ratio (AR) = 1.50 and contrast-to-noise ratio (CNR) = 4.65, demonstrating a "diminishing returns" effect where high-density configurations showed minimal performance gains despite increased channel count. Physiological Modeling and Results: A mathematical model combining mass balance equations and Hill-Langmuir oxygen binding kinetics was constructed to quantify individual Hb and Mb contributions to NIRS signals across varying blood flow conditions. The model successfully predicted maximum oxygen uptake trends consistent with experimental data, validating the framework for separating microvascular oxygen release from myocyte oxygen uptake. Conclusion: This work establishes a wearable high-density FD-NIRS platform advancing absolute tissue characterization and spatial depth resolution capabilities. The research provides probe design guidelines and signal processing protocols for developing portable muscle function monitoring devices suitable for sports physiology and metabolic assessment applications.

ABSTRACT. Ischemic stroke is the primary cause of global long term adult disability and leads to persistent multidimensional behavioral and psychological impairments including learning and memory decline, cognitive deficits, motor coordination disorders, and anxiety like behaviors. To synchronize acute neural microenvironment regulation with long term neurological repair, we developed an integrated all in one nanodelivery platform by combining synthetic biology and supramolecular engineering. This platform utilizes genetically engineered mesenchymal stem cell derived exosomes that effectively penetrate the blood brain barrier and constitutively express a neurotrophic cocktail consisting of brain derived neurotrophic factor, nerve growth factor, neurotrophin 3, and neurotrophin 4. These exosomes encapsulate a supramolecular core of curcumin and L arginine to enable a spatiotemporal sequential therapeutic strategy. The core provides acute neuroprotection by scavenging reactive oxygen and nitrogen species and restoring mitochondrial function, while the sustained release of neurotrophic factors supports neuronal survival and neural circuit reconstruction for long term functional recovery.

In a murine model of ischemic stroke, the platform exhibited exceptional targeting capabilities and significantly ameliorated multidimensional behavioral deficits. Specifically, in the rotarod test evaluating motor coordination and balance, the latency to fall increased from 73 s in the PBS group to 286 s in the treatment group, achieving a 98.6% recovery relative to the 282 s observed in the normal group. In the novel object recognition test assessing cognitive function, the discrimination index rose from 35% to 63%, representing a 96.9% recovery compared to the 65% in the normal group. For the open field test measuring spontaneous activity and anxiety like behavior, the total distance traveled reached 1695 cm in the treatment group versus 816 cm in the PBS group, a 93.1% recovery relative to the normal level of 1820 cm. In the Morris water maze test quantifying spatial learning and memory, the escape latency was reduced to 12 s in the treatment group, achieving a 97.4% recovery compared to the 12.32 s in the normal group and 47 s in the PBS group. Combined single cell RNA sequencing and proteomics revealed a dual regulatory mechanism characterized by the rescue of vulnerable neuronal subtypes and the reprogramming of M1 microglia into the M2 phenotype to remodel the cerebral microenvironment and promote neural circuit repair. This study establishes a promising translational paradigm for post ischemic stroke behavioral and psychological rehabilitation and cognitive reconstruction while providing robust experimental evidence for the underlying neuropsychological mechanisms.

ABSTRACT. The study and evaluation of human behavior in psychology very often rely on self-reported data, collected either through observational or interview methods or through psychometric tools such as questionnaires. These methods have greatly contributed to psychological research and the "quantified" assessment of human behavior. However, they also have limitations in terms of objectivity, as they remain inherently subjective by nature. Nowadays, currently available technological tools such as IoT enable the easy acquisition of measurement instruments for certain physiological behavioral characteristics, such as body movement, eye movement, facial expressions, or heart rate variability. Algorithmic data analysis methods, such as predictive AI and machine learning, open up innovative prospects for processing heterogeneous behavioral data: both subjective and objective, physical and psychological. But one of the challenges that remains predominant is being able to appropriately combine data of very heterogeneous types to better understand human behavior, including its inter-individual particularities. Our study presents an illustration of these methodological challenges. Our objective was to assess the sense of presence and anxiety VR immersion through psychometric scales, while simultaneously collecting physiological data to identify correspondences between self-reported and physiological measures. Our sample was composed of twenty-two participants, aged from 21 to 32 years old, without any psychiatric history. In our protocol, we evaluated immersive VR experiences using psychometric questionnaires alongside biometric sensor measurements. All participants performed motor tasks in the VR game Richie's Plank, which simulates walking on a narrow plank at a great height. Their trajectories, muscle activity, and heart activity (ECG) were recorded during the immersion in VR, while anxiety (HADS) and presence (PQ) were assessed before and after immersion. Our preliminary results suggest that, overall, there is no significant group correlation between HADS scores, presence questionnaire and ECG scores in the participants group. However, ECG correlated well with HADS for all participants during some movements of the VR task, especially while participants had to walk back to the elevator to return to the ground (r=0.72). Moreover, the global presence score is negatively correlated with the HADS score (r = -0.60). The interpretation of our results raises several questions, particularly on a methodological level. Regarding the use of IoT, several limitations should be highlighted. First, we relied solely on ECG data, although adding EMG data could have been valuable for a better understanding of the correlations between movement and anxiety, presence, and cybersickness scores. Next, the data collected via IoT were interpreted using cohort averages rather than individual results. However, significant inter-individual differences emerged in participants' reactions during VR immersions. Algorithmic modeling of intra-individual physiological variations (ECG, EMG) during the immersive task could offer valuable insights to address this limitation. Regarding the psychometric scales used, other limitations have emerged. For example, the presence results are unusual according to previous works about correlation between presence and anxiety disorders. But the presence scale used does not clearly distinguish the sense of presence from the immersive characteristics of the VR environments. This complicates the interpretation of the obtained scores when examining links between sense of presence and physiological variables. Other presence scales could be explored to better integrate users' psychological experiences with their physiological responses. Our results raise new challenges related to multimodality in the study of normal and pathological human behavior. Therefore, both subjective evaluations of psychological states and complementary physiological data on participants' behavior should be considered. The key challenge is integrating and analyzing these heterogeneous data sources to identify meaningful correlations. Nevertheless, this offers promising opportunities and innovative directions for future research.

ABSTRACT. Humans exhibit diverse perceptual–motor styles that reflect both intra- and inter-individual variability in sensorimotor transformations. A key question, however, is when such styles should be readjusted to maintain optimal motor control. Do changes in perceptual–motor style reveal basic emotions? Could they indicate the onset of a pathology? Or might they serve as markers for monitoring rehabilitation and recovery?

In the first study (Mantilla et al., 2020), we proposed a method to characterize the perceptual–motor style of healthy individuals during both quiet standing and locomotion. Our findings suggested that this style can be decomposed into a static component—a stable configuration of body segments with respect to the gravitational vertical—and dynamic components corresponding to the coordinated movements of the head, trunk, and limbs. We identified a set of static and dynamic markers that defined an individual’s perceptual–motor style, characterized by high inter-individual variability but low intra-individual variability.

In the second study (Wang et al., 2024), our goal was to investigate how one basic emotion, fear, modulates the perceptual motor style during locomotion in 16 individuals with no antecedent fear of height or acrophobia. To explore the emotional modulation of these markers, we employed a low-cost virtual reality (VR) video game, Richie’s Plank, designed to induce anxiety during locomotion. In this immersive scenario, participants walked along a narrow plank suspended between two skyscrapers at a height equivalent to the 30th floor. Our first finding was that both static markers (postural configurations of the head, trunk, and limbs relative to gravity) and several dynamic markers (jerk, root mean square, sample entropy, and adherence to the Two-Thirds Power Law) captured modulation induced by fear. Our second finding revealed a surprisingly large heterogeneity among young healthy participants: individuals exhibited markedly different responses to height exposure, suggesting that fear sensitivity and anxiety traits substantially influence perceptual–motor style, especially when locomotion control becomes challenging. Notably, 61% of participants showed alterations in at least one variable related to dynamic locomotion control on the ground after height exposure.

Based on these data, we will present a third study exploring the relationship, if any, between an individual's perceptual motor style in everyday life and their sensorimotor performance in risky situations, such as walking at height.

ABSTRACT. Immersive virtual reality (VR) has been increasingly investigated as a medium capable of modulating psychophysiological states through heightened sensory engagement and embodied interaction. Previous research on interactive and immersive media indicates that VR can influence autonomic activity, with heart rate (HR) emerging as a sensitive marker of emotional arousal and physiological engagement during mediated experiences. Importantly, immersive environments do not uniformly induce relaxation: while VR can initially elicit increased arousal due to novelty and sensorimotor involvement, repeated or sustained exposure has been associated with reductions in cardiac activation compared to less immersive formats, suggesting a dynamic regulation of sympathetic activity rather than a straightforward parasympathetic shift. Within this framework, the present study examined whether an immersive VR experience inspired by the overview effect—characterized by a first-person perspective of Earth observed from space—elicits a different cardiac response than a content-matched, non-immersive verbal condition. The focus of this contribution is exclusively on heart rate as an index of autonomic arousal, given its extensive use in VR and media research and its sensitivity to changes in sympathetic activation during emotionally salient and embodied experiences. A sample of healthy adults (N = 57) was randomly assigned to either a VR condition or a non-immersive condition. Participants in the VR group experienced a 360° multisensory space-travel scenario culminating in a distant view of Earth, while participants in the control group listened to the same narrative without visual immersion. Electrocardiogram (ECG) data were collected before and during the experience. Heart rate data were analyzed using a mixed repeated-measures ANOVA with Time (pre, post) as a within-subject factor and Condition (VR, noVR) as a between-subject factor. Results revealed a significant Time × Condition interaction, F(1,55) = 5.16, p = .027, alongside a significant main effect of Condition, F(1,55) = 4.21, p = .045. The main effect of Time did not reach statistical significance. Estimated marginal means indicated a decrease in HR from pre- to post-exposure in the VR group (81.2 to 78.3 bpm), whereas the non-immersive group showed no reduction and a slight increase in HR (86.3 to 86.6 bpm). These findings indicate that immersive delivery, rather than narrative content alone, was associated with a short-term reduction in cardiac activation. Consistent with prior literature, the observed HR modulation appears to reflect a reduction in sympathetic arousal rather than an increase in parasympathetic activity. This pattern aligns with evidence suggesting that brief immersive VR experiences may primarily affect arousal regulation and emotional engagement, whereas measurable parasympathetic changes typically require longer or explicitly biofeedback-oriented interventions. Overall, these results support the use of HR as an outcome measure for investigating embodied and emotionally evocative VR experiences and suggest that overview-effect–inspired immersion may facilitate cardiac downregulation compared to non-immersive formats, even within a single-session exposure.

ABSTRACT. Introduction Virtual reality (VR) is increasingly used in social neuroscience research, offering new opportunities to connect physically distant individuals within shared, interactive environments. Recent advances enable users to meet and cooperate in shared virtual spaces through personalized, hyper-realistic avatars that closely approximate real-world appearance and behavior. In this preliminary study, we describe a newly developed hyper-realistic VR system in which two users engage in collaborative and learning activities. We focus on how dyads’ sense of presence, immersion, and related experiential constructs are associated with task performance within the system. Method Participants will be paired into dyads and will individually create and customize a hyper-realistic avatar using dedicated software before entering the virtual environment. This step is intended to support embodiment and self-representation, which are expected to contribute to presence and co-presence during dyadic interaction. Participants will wear a Meta Quest Pro headset and will complete all tasks in a shared virtual room while seated physically in different spaces, ensuring comparable interaction constraints across dyads. Three VR tasks have been designed to systematically vary the type and intensity of interpersonal coordination required: (1) a Collaborative Task (CT), (2) a Creative Collaborative Task (CCT), and (3) a Learning Task (LT). All tasks are performed in pairs and require social interaction (e.g., shared attention, turn-taking, and negotiation), but they differ in goals and in the degree to which success depends on joint problem solving and alignment of mental models. In the CT, participants complete a card-matching game with the objective of minimizing errors while coordinating efficiently across rounds. In the CCT, participants discuss and jointly decide how to visually represent predefined concepts using symbol cards, emphasizing collaborative creativity and shared meaning-making. In the LT, dyad members adopt complementary roles: one participant acts as a “teacher” and delivers a short lesson to the “student” on an unfamiliar topic, encouraging interactive explanation, clarification, and knowledge consolidation. Performance will be assessed using quantitative measures such as completion time and accuracy. Self-report measures will include sociodemographics; affective state (Self-Assessment Manikin, SAM); user experience (User Experience Questionnaire, UEQ); cybersickness (Simulator Sickness Questionnaire, SSQ); presence and immersion (Igroup Presence Questionnaire, IPQ); and co-presence/social presence (Networked Minds Social Presence Measure). Results Data collection is ongoing; therefore, we report planned analyses and expected patterns on at least 5 dyads. We will quantify task performance at the dyad level and test associations with presence, immersion, and co-presence while accounting for usability and cybersickness. We expect higher presence/immersion and higher co-presence to be associated with better collaboration (higher accuracy and/or faster completion in CT and CCT) and stronger learning outcomes in LT. Conversely, greater cybersickness is expected to negatively relate to performance and subjective user experience. Analyses will include correlational tests, enabling comparisons of how experiential constructs are differentially linked to outcomes across CT, CCT, and LT. Conclusion By linking task performance and learning outcomes to presence, immersion, and co-presence, this study will provide foundational evidence on which experiential dimensions matter most for effective collaboration in hyper-realistic VR. These results will inform the next phase of research leveraging the same paradigm to investigate how social and neural synchrony contribute to collaborative behavior and learning in immersive virtual environments.

ABSTRACT. With the growing integration of Virtual Reality (VR) in research, therapy, and entertainment, understanding how individual differences shape users’ affective responses has become increasingly important. While there is growing evidence of VR’s effectiveness, considerable variability exists in how individuals experience and respond to virtual environments (VEs). User-related factors, such as personality traits may contribute to such differences.

Personality traits have been examined in relation to presence, the subjective sense of “being there” in a virtual environment, which is considered a key psychological dimension of VR experience. Among personality dimensions, extraversion and neuroticism are particularly relevant because of their well established links to emotional reactivity and regulation. However, the literature has yet to reach consensus on how personality influences presence, with mixed findings.

This study investigates the combined influence of extraversion, neuroticism, and presence on positive and negative affect across contrasting VR contexts. We used a cross-sectional correlational design to investigate: 1) the relationship between neuroticism, extraversion, and presence, and 2) the impact of neuroticism, extraversion, and presence on positive and negative affect induced by a stressful and relaxing virtual environment. A total of 55 healthy participants aged between 18 - 58 years (M = 24.86, SD = 7.68) took part in the study. Participants were asked to watch a stressful VE, depicting a plane crash (T1) and were subsequently introduced to a relaxing VE that depicted a forest of serenity (T2) via the Meta Quest 2. Neuroticism and extraversion were measured using the Big Five Inventory, affect using the Positive and Negative Affect Schedule, and presence using the Presence Questionnaire. At baseline (T0), we assessed neuroticism, extraversion, and affect. After exposure to a stressful virtual environment (T1), and again after experiencing a relaxing virtual environment (T2), we assessed affect and presence.

Results showed a small positive correlation between extraversion and presence (T1) (r = .20, p<.001) and presence (T2) (r = .18, p<.001). Neuroticism was not significantly associated with presence. Multiple regression analyses showed that neuroticism was a significant positive predictor of negative affect at T0 (β = .388, p < .001) and T1 (β = .299, p = 0.009) and a significant negative predictor of positive affect at T1 (β = -.335, p = .002). Extraversion surfaced as a significant positive predictor of positive affect at T0 (β = .419, p < .001), T1 (β = .510, p < .001) and T2 (β = .427, p < .001). Presence was a significant positive predictor of negative affect at T1 (β = .313, p = .026) and a significant positive predictor of positive affect at T2 (β = .480, p < .001) only.

Overall, our results indicate that extraversion, neuroticism and presence in VR each make distinct, time specific contributions to affective responses in VR, underscoring the importance of user-related factors when evaluating applied VR experiences. More specifically, results showed small positive associations between extraversion and presence in both the stressful and relaxing VEs, suggesting that more extraverted users tend to feel slightly more “there” in VR. Neuroticism consistently predicted more negative affect at baseline and during the stressful VR exposure, and less positive affect during the stressful exposure, whereas extraversion robustly predicted higher positive affect across all time points. Presence, in turn, predicted more negative affect during the stressful VR exposure and more positive affect in the relaxing environment.

Together, these findings show that individual differences are meaningful predictors of how users feel in virtual environments, each exerting independent effects on affect. This has direct implications for the design and evaluation of applied VR systems where tailoring VR content to user characteristics may enhance user experience and intervention effectiveness.

ABSTRACT. Recent advances in communication and cognitive science highlight embodiment as a key mechanism linking physical experience and mediated understanding. Through the lens of embodied cognition, bodily sensations and actions are fundamental for cognitive processing and meaning-making. Embodiment refers to the degree to which users feel physically involved and capable of acting within a digital context, a phenomenon encompassing – but not reducible to – presence, the subjective sense of “being there”, and agency, the perception of control over one’s actions. These processes reinforce one another and promote cognitive engagement. However, little is known about how embodied experiences shape comprehension and sense-making in health communication, where information is often complex, abstract, and difficult to relate to personal experience, particularly for individuals with limited prior knowledge. This pilot study therefore examined whether different levels of embodiment in a digital health task influence knowledge and psychological outcomes, and whether these effects are moderated by participants’ prior health literacy/expertise.

The pilot study employed a one-way between-subjects design with two conditions. Forty-five Italian-speaking adults were randomly assigned to a low- or high-embodiment condition presenting information about the management of chronic cardiovascular disease. Across conditions, informational content, task duration (10 minutes), and interaction frequency were held constant; conditions differed only in the degree of sensorimotor involvement and bodily framing of the information. In the low-embodiment condition, participants completed a desktop-based task in a passive format, reading written slides and advancing with mouse clicks. In the high-embodiment condition, the same content was delivered with a first-person perspective in which the cursor was represented by arms, and participants were invited to complete activities simulating bodily actions to evoke motor involvement and personal ownership of the task. Health literacy/expertise was operationalized as a composite variable (familiarity with the disease and/or healthcare profession). After the task, participants completed self-report measures including the Perceived Qualities Questionnaire (PQQ), the Cardiovascular Risk Knowledge Scale (CARRFKL), the Inclusion of Other in Self scale (IOS) for psychological distance, and the Psychological Empowerment Scale (PES). The study protocol received ethical approval from the Università della Svizzera Italiana on December 17, 2025, and informed consent was obtained from all participants.

Linear models including Condition, Literacy, and their interaction revealed no overall main effect of embodiment. However, significant Condition × Literacy interactions emerged for perceived quality (PQQ) and declarative knowledge (CARRFKL), both with η²p = .168 (p ≈ .006–.007 after FDR correction). Among low-literacy participants, the high-embodiment condition produced significantly higher scores on PQQ (p = .029) and CARRFKL (p = .028) compared to the low-embodiment condition. No significant differences appeared among high-literacy participants. No significant effects or interactions were observed for psychological distance (IOS) or empowerment (PES). Sensitivity analyses (non-parametric tests and alternative literacy operationalizations) confirmed the robustness of the moderation pattern. These preliminary findings provide effect-size estimates to inform sample-size planning for a larger pre-registered study.

This pilot study provides initial evidence that sensorimotor embodiment can enhance perceived quality and knowledge acquisition among individuals with low health literacy, consistent with embodied cognition theory positing that sensorimotor grounding is particularly beneficial when existing schemas are limited. No effects emerged for psychological distance or empowerment in this single-session design. By framing embodiment as a mechanism of cognitive grounding rather than mere interactivity, the findings offer an exploratory account of how digital experiences may support understanding of complex health information.

ABSTRACT. Introduction. Video gaming is a common and diverse leisure activity among adolescents and young adults. Although prior research has identified associations between personality traits and gaming behavior, most studies have relied on cross-sectional designs. Consequently, little is known about how gaming engagement and personality traits co-develop over time, particularly across adolescence and the transition to adulthood, a developmental period characterized by both personality maturation and heightened responsiveness to environmental contexts. Guided by a sociogenomic perspective, the present longitudinal study compares personality development between gamers and non-gamers and examines how gaming exposure and personality traits influence each other over time, conceptualizing personality as a dynamic construct that continuously interacts with the environment. We hypothesize that engagement in video gaming is associated with distinct developmental trajectories of personality across adolescence, reflecting systematic differences relative to non-gamers. Furthermore, we expect bidirectional, within-person associations between gaming behavior and personality traits over time, consistent with dynamic person–environment transactions involving both selection and socialization processes.

Methods. Data will be drawn from a large-scale european longitudinal database investigating factors shaping development from adolescence into adulthood. The data will stem from 1,508 participants (52.25% females) across multiple countries using repeated assessments, of whom 43.30% reported playing video games at some time point. Data collection was approved by the local ethics committees at each research site, and written informed consent was obtained from all participants and from a parent or guardian when participants were minors. The present study focuses on the assessments conducted at ages 14, 16, 19, and 22. Personality traits were self-reported using the NEO Five-Factor Inventory and the Substance Use Risk Profile Scale, either in person or online. Gaming behavior was assessed through a comprehensive questionnaire that captured weekly frequency and daily intensity of play across waves. The data analysis plan will be divided into three stages. Frst, a data quality control will be performed: the dataset will be systematically cleaned, transformed, and evaluated for missing data, outliers, and reliability to ensure it is suitable for analyzing relationships between personality and video gaming over time. Second, Multigroup Latent Growth Curve Models will be used to describe personality development across the study period separately for gamers and non-gamers. Covariates such as gender, age, and centre will be added to the model. Third, Random Intercept Cross-Lagged Panel Models will be estimated to examine reciprocal associations between personality traits and gaming exposure across waves, effectively separating within-person from between-person effects. Alternative strategies may be considered if the models exhibit signs of non-identification, failure to converge, or instability.

Expected Results. Differences in both intercepts and slopes of personality development trajectories are expected between gamers and non-gamers, indicating distinct developmental pathways even when controlling for confounding variables. These trajectories might align with prior cross-sectional findings, particularly higher levels of Openness, Impulsivity and Hopelessness and lower levels of Conscientiousness and Extraversion among gamers. Additionally, reciprocal associations between personality traits and gaming exposure are expected to reflect both selection and socialization processes: changes in gaming behavior will predict corresponding changes in trait levels, and conversely, variations in personality will be associated with subsequent changes in gaming involvement.

Conclusion. By applying longitudinal modeling within a large developmental cohort, this study extends existing research beyond static and pathology-focused perspectives on gaming. The findings are expected to clarify whether associations between personality and gaming exposure reflect dispositional selection, contextual socialization, or dynamic reciprocal processes unfolding during a sensitive developmental window.

ABSTRACT. Background Societal aging is creating unprecedented demands on long-term care (LTC) systems globally, with residents living with mild cognitive impairment (MCI) or mild dementia particularly vulnerable to social isolation. While AI-driven conversational agents show promise for addressing emotional well-being, and immersive virtual reality (IVR) may enhance engagement through heightened social presence, no study has directly compared IVR versus desktop delivery of the same AI conversational agent with people with MCI and mild dementia living in LTC and transitional care.

Objective Preregistered via Open Science Framework, this repeated measures study addresses this gap by evaluating the impact of modality (desktop versus IVR) on social presence, engagement, usability, and attitudes towards AI. The research focuses on AI companion interactions among individuals with MCI and mild dementia and follows up on a protocol presented last year. This research aims to addresses ethical, compassionate, human-centered technology design by clarifying whether immersive delivery meaningfully improves social interaction quality for cognitively impaired older adults, informing evidence-based digital health guidelines and advancing interventions to mitigate social isolation in dementia care.

Methods This four-block randomized study recruits 20 participants from three LTC and transitional care facilities in Ontario, Canada. Inclusion criteria are: diagnosis of mild cognitive impairment or mild dementia, English fluency, adequate sight and hearing, and informed consent from the participant or a substitute decision maker. Sociodemographic data collected include age, gender, loneliness, cognitive status, and functional independence. Each participant engages in two unscripted conversational sessions with "Kiera", an AI companion powered by GPT-4o integrated with speech-to-text and text-to-speech capabilities. The system processes participant speech through automated speech recognition, generates contextual responses via the large language model with tailored prompting for reminiscence-based interaction, and delivers responses through synthesized voice synchronized with a virtual human. In the IVR condition, participants interact within a simulated park environment using a Meta Quest 2 headset with full visual immersion, head-tracking, and spatial audio. In the desktop condition, participants engage via desktop monitor displaying the same environment without head-mounted display. After each session, participants complete questionnaires adapted from the following measures: The Networked Minds Social Presence Measure (NMSPM), User Engagement Scale-Short Form (UES-SF), AI Attitudes Scale-4 (AIAS-4), and Usability Metric for User Experience-Lite (UMUX-Lite). This is followed by a semi-structured interview. Sessions occur within the same week on separate days, with modality order counterbalanced across participants. No time limit is imposed; interaction duration is recorded as a behavioral indicator of engagement. This study was approved by the Bruyère Health Research Ethics Board, with safeguards in place to ensure participants are aware that they will interact with a non-human agent.

Data analysis Analysis will employ Mixed ANOVA with modality as within-subjects factor and presentation order as between-subjects factor to assess main effects and order effects (α = 0.05). Semi-structured interviews explore experiential aspects of presence, engagement, and comfort. Qualitative interview data will undergo thematic analysis using a deductive, a-priori codebook mapping directly onto the quantitative measures used to triangulate convergent or divergent patterns, with inductive coding added to capture any emerging themes not covered by the initial framework.

Expected Outcomes Expected outcomes include higher social presence, engagement, and more positive attitudes toward AI in the VR condition due to immersive technology's encompassing, multisensory stimulation. Longer interaction duration in VR would provide behavioral corroboration of heightened engagement. Qualitative themes are expected to align with quantitative findings while revealing comfort preferences and barriers not captured by standardized measures. Should VR demonstrate superior outcomes, findings would justify immersive technology's resource requirements for LTC implementation. Alternatively, equivalent desktop outcomes would indicate scalable AI companion benefits without immersive barriers, facilitating broader adoption in resource-constrained settings.

ABSTRACT. Introduction The LGBTQ+ community continues to face widespread discrimination and harassment, driven in part by implicit bias, which refers to unconscious automatic associations that shape everyday social interactions. Traditional bias-reduction methods tend to produce only short-term outcomes. Immersive virtual reality (VR) has emerged as a promising alternative, enabling participants to inhabit different bodies and perspectives to foster long-term prosocial behavior. This study explores the Golden Rule Embodiment Paradigm (GREP), in which participants virtually embody another person's perspective to reduce implicit bias and promote prosocial behavior. In GREP, all participants witness a verbal attack by a (virtual) perpetrator on a (virtual) victim. During a replay, participants re-experience the scene from either a first-person perspective (1PP, experimental) condition, in which they first embody an avatar next to the perpetrator before transitioning to embody the victim, or a third-person perspective (3PP, control) condition, in which they simply observe from a distant perspective. While GREP has shown promising outcomes in VR interventions for workplace harassment, domestic assault, and anti-racist training, this study extends its application to LGBTQ+ contexts. The primary objective is to examine whether 1PP embodiment produces greater prosocial behavior toward LGBTQ+ individuals than 3PP, measured by interpersonal proximity, verbal intervention, and visual attention in a second test scenario. The secondary objective is to assess whether 1PP produces greater reductions in pro-heterosexual implicit associations than 3PP, as measured by the Sexuality IAT. We hypothesize that (H1) participants in the 1PP condition will display greater prosocial behavior and (H2) greater reductions in pro-heterosexual implicit associations compared to the 3PP condition.

Methods Fifty participants (aged 18+) will be recruited from the greater Barcelona area. Exclusion criteria include VR-contraindicated medical conditions (e.g., cybersickness, epilepsy), significant cognitive impairments, and inability to communicate in English, Spanish, or Italian. The study was approved by the Comissió Bioètica UB and follows standard ethical protocols. The protocol consists of two sessions conducted one week apart. Participants will be randomly assigned to either the experimental (1PP) or control (3PP) condition. All sessions will begin with an embodiment induction phase to establish avatar ownership. Before session 1, participants will complete a demographic questionnaire and a Sexuality IAT. After session 2, participants will complete a counterbalanced Sexuality IAT, in which the order of attribute category pairings is reversed relative to session 1 to control for order effects. After each session, participants will complete a standard presence questionnaire to verify a sufficient sense of 'being there,' serving as a validity check. Condition-blind human coders will analyze recordings from session 2 to assess verbal intervention and interpersonal proximity toward the victim or perpetrator, as well as visual attention patterns directed toward key targets in the scene. In Session 1, all participants embody an avatar next to an aggressive protester harassing LGBTQ+ Pride marchers on a city street. During the replay, participants in the 1PP condition embody the victim Pride marcher, while participants in the 3PP condition observe from atop a nearby building. In Session 2, participants complete a virtual grocery task interrupted when a customer harasses a transgender woman.

Expected Outcomes Based on previous research, we predict that the experimental group will display greater prosocial behavior and significant reductions in pro-heterosexual implicit associations on the Sexuality IAT compared to the control group. We anticipate that presence scores will be sufficiently high in both conditions, confirming the prerequisite for the study is satisfied. Together, these findings would support VR as a scalable, long-term tool for reducing implicit bias and promoting prosocial behavior in everyday social harm prevention.

ABSTRACT. Background Virtual Reality Hypnosis (VRH) delivers hypnotic induction and suggestions through virtual reality (VR). Although the integration of VR and hypnosis has attracted growing interest among healthcare professionals, scientific understanding of its underlying mechanisms remains limited. Examining individual differences in neurophysiological and psychological factors that may influence its efficacy, and comparing these with hypnosis alone, may help clarify the processes involved. This study aims to determine whether VRH is non-inferior to clinical hypnosis in terms of phenomenological and neurophysiological outcomes, with non-inferiority margins derived from the existing literature. Additionally, EEG activity will be examined to assess whether VRH and hypnosis elicit comparable patterns of functional connectivity and large-scale network organization. Method Sixty-six healthy adult volunteers will be included. Participants had no sensory, neurological, or psychiatric impairments. Hypnotizability and immersion were assessed prior to the experiment. On the experimental day, baseline anxiety and pain were assessed using visual analogue scales (VAS), followed by a 5-minute resting-state EEG recording. Participants underwent two 15-minute conditions (VRH and hypnosis) in a randomized within-subject design, separated by a 30-minute washout period with continuous EEG recording. During the washout, absorption, dissociation and anxiety were assessed via questionnaires. After each condition, subjective experience was assessed using VAS (absorption, dissociation, automaticity, pain, awakeness, anxiety, presence, and time perception) and standardized questionnaires (satisfaction, presence, and cybersickness). Behavioral analysis: Within-subject comparisons will evaluate differences in subjective experience (e.g., absorption, dissociation, automaticity, presence, and time perception) between the VRH and hypnosis conditions. Neurophysiological analysis: Brain activity and functional connectivity will be analyzed across baseline, hypnosis, and VRH conditions. Connectivity patterns will be examined at regional and lobar levels, and network organization will be characterized using graph-theoretical metrics (e.g., measures of segregation and integration). Within-subject comparisons will assess differences between conditions, and associations between EEG measures. Preliminary findings and expected outcomes: Preliminary results indicate comparable effects of VRH and hypnosis on dissociation and absorption. Both conditions are expected to induce similar changes in subjective experience and brain activity, particularly in terms of functional connectivity and network organization. Conclusion: This study investigates the phenomenological and neurophysiological effects of VRH compared to clinical hypnosis, contributing to the understanding of their underlying neural mechanisms. It also aims to determine whether VRH produces effects that are non-inferior to those of clinical hypnosis. These findings may inform the development of VRH as a clinically relevant alternative and support more targeted and effective interventions.

ABSTRACT. Online gambling has evolved into a complex public health issue that extends beyond financial loss, encompassing physical and mental health consequences. Young adults are particularly exposed to online gambling opportunities; however, the relationship between gambling-related symptoms and self-reported health remains insufficiently understood, particularly the mechanisms through which this relationship operates. This study examined the association between gambling disorder symptoms and self-reported physical and mental health, and investigated whether social and familial consequences of gambling mediated this relationship. Data were drawn from a nationally representative telephone survey of Portuguese young adults aged 18–34 (N = 1,993), conducted using Computer-Assisted Telephone Interviewing (CATI). Households were selected through random generation of fixed and mobile telephone numbers, with quota sampling by gender, age group, and NUTS II region. Gambling-related symptoms were assessed using an adapted version of the DSM-IV pathological gambling criteria (Canale et al., 2016), measuring the frequency of seven gambling-related problems over the previous 12 months. This instrument was adopted in the original data collection protocol of the BlindGame national survey (Farias & Antunes, 2024), which preceded the present analyses. Although the DSM-5 (APA, 2013) reclassified gambling disorder, the core symptom criteria remain largely consistent across editions, and the use of this measure ensures comparability with previous population-based research. Importantly, symptoms were operationalized dimensionally as frequency scores rather than as categorical diagnoses, reducing the impact of threshold-related differences between editions. Self-reported health was measured using the SF-12 (Ware et al., 1996), yielding Physical (PCS) and Mental (MCS) Component Summaries. Although gambling symptoms include psychological dimensions, the SF-12 captures broader health-related functioning rather than disorder-specific symptomatology, reducing conceptual overlap between the predictor and outcome. Social and familial consequences were measured using an adapted scale from Hubert (2015). Descriptive statistics for key variables were as follows: gambling symptoms (M = 8.21, SD = 2.19), PCS (M = 18.65, SD = 1.71), MCS (M = 24.44, SD = 3.26), and social/familial consequences (M = 0.47, SD = 1.54). Linear regression analyses indicated that higher gambling symptom severity significantly predicted poorer physical health (PCS: F(1, 658) = 25.768, p < .001) and poorer mental health (MCS: F(1, 685) = 11.324, p < .001). Mediation analysis revealed that social and familial consequences partially mediated the relationship between gambling symptoms and mental health (indirect effect: β = −0.045, SE = 0.014, p = .002, 95% CI [−0.073, −0.017]). No significant mediation was found for physical health. Given the exploratory nature of the study, age-adjusted analyses were not conducted, allowing examination of the primary associations across the full age range. These findings associate greater gambling severity with reduced physical and mental well-being among young adults, and suggest that relational consequences constitute a pathway through which gambling symptoms affect mental health. However, the cross-sectional design precludes causal inference and does not permit establishing the temporal ordering assumed by mediation models. Effective prevention and intervention strategies should address not only individual gambling behaviours but also the social and relational consequences that may amplify health-related harms.

ABSTRACT. Background: Pain is a complex, multifaceted phenomenon shaped by both sensory input and internal body representations. Interoception – the sensing and integration of physiological signals – plays a key role in these representations. Interoceptive Modeling has been proposed as a strategy to modify internal representations and thereby influence pain, but empirical evidence is limited. Immersive virtual reality (VR) provides a platform for embodiment and controlled manipulation of body representations, yet prior applications have largely focused on distraction-based analgesia and have not systematically tested controlled pain paradigms. Objective: This cross-over, single-blind, multicenter randomized controlled trial is the first to empirically test VR-based Interoceptive Modeling in experimentally induced acute thermal pain. We aimed to examine its effects on subjective pain experiences (intensity, unpleasantness, coping) and autonomic nervous system (ANS) reactivity. Methods: Sixty healthy female participants (aged 20–33 years; M = 25.13, SD = 3.04) were recruited in Vienna, Austria and Milan, Italy. Participants were randomized to a cross-over design with a one-week washout period, completing two conditions in counterbalanced order: an experimental condition involving modeling of the painful area and a control condition without modeling. An all-female sample was selected to account for established sex differences in pain perception and coping. Experimental pain was induced by applying a commercial capsaicin cream (0.075%, 750 µg/g), a well-validated pain induction method, to the volar forearm of the non-dominant hand. Participants wore a Meta Quest 3 headset and interacted with a custom-developed PAINEX-VR application, embodying virtual arms matched to their arm size and skin tone. Embodiment was induced through a brief motor task. Using a separate observer interface, experimenters personalized the visual features of the painful area—its location, size, color saturation, and pulsation—based on each participant’s reported pain. In the experimental group, these features were gradually reduced during Interoceptive Modeling, progressively fading the painful area until only 20% remained in the final minute. This aimed to target internal pain representations by gradually altering the visual salience of the pain, rather than providing distraction. In the control group, the painful area remained visually unchanged. In both conditions, participants were instructed to observe the virtual arm without additional guidance, allowing the effects of modeling versus no modeling to emerge from the visual manipulations themselves. Primary outcomes were self-reported pain ratings (VAS), with autonomic nervous system (ANS) parameters as secondary outcomes. Data were analyzed using linear mixed-effects models. Results: A preliminary LMM confirmed the efficacy of capsaicin pain induction. Pain intensity increased significantly from baseline to post-capsaicin. Neither condition nor country, nor the time × condition interaction, were significant, indicating comparable pain induction across groups and sites. Yet, there were marked differences between the two conditions with regards to primary and secondary outcomes: While the control group showed a continued increase in self-reported pain intensity and pain unpleasantness (both VAS) over time, interoceptive modeling in the experimental condition halted the escalation of pain symptoms, leading to a significant stabilization of pain intensity and unpleasantness during the modeling phase. Despite comparable self-reported pain coping across the two conditions, the experimental group with Interoceptive Modeling exhibited superior parasympathetic recovery following the intervention, indicated by significantly higher RMSSD and HF power immediately post-intervention. Conclusions: This study provides preliminary evidence that VR-based Interoceptive Modeling may offer a personalized approach to modulating acute pain by targeting internal pain representations. These findings suggest potential for non-pharmacological pain management, but generalizability is limited by the all-female sample, and further investigation is needed in other pain types, mixed-gender populations, and chronic pain contexts.

ABSTRACT. Background

Sleep plays a central role in emotional regulation and adaptation to daily experiences, yet the influence of emotional experiences on subsequent sleep physiology remains poorly understood. Previous studies have largely relied on subjective sleep measures, clinical populations, or laboratory-based protocols that may compromise ecological validity. In parallel, immersive virtual reality (VR) is increasingly used for emotional induction, entertainment, and therapeutic purposes, but its potential impact on sleep has received limited attention. Understanding how emotionally engaging VR experiences influence sleep, and whether these effects depend on emotional valence, is therefore essential for both basic sleep research and the application of VR technologies.

Objectives

This study protocol aimed to investigate the impact of immersive VR experiences on sleep in healthy adults using a home-based design. Specifically, we examined whether emotional VR exposure alters sleep architecture relative to neutral VR, and whether these effects vary by emotional valence. Based on prior literature, we expected emotional stimulation to modulate REM sleep and hypothesised that deep sleep might be differentially affected by emotional valence.

Methods

The study adopts a mixed experimental design with a between-subject factor (Group: positive vs negative) and a within-subject factor (Condition: neutral vs emotional). Ten healthy adults aged 20–35 years who reported good sleep quality and adequate emotional regulation were recruited and assigned to either the positive- or negative-emotion VR group. Each participant completed three monitored nights at home: one habituation night without VR exposure and two experimental nights following VR exposure, with a 1-week interval between them.

During the habituation night, nocturnal polysomnography (PSG) was recorded without VR exposure, providing an individualised reference of habitual sleep architecture. In the experimental sessions, participants viewed immersive 360° VR video clips at home using a head-mounted display. VR stimuli were selected from a validated open-access database and grouped by emotional valence. All participants were exposed to neutral and emotionally salient VR content, according to their group assignment.

Emotional state was assessed before and after VR exposure using the Self-Assessment Manikin (SAM). User experience, usability, and potential simulator-related discomfort were also evaluated (UEQ, SSQ). On each recording night, participants self-applied a portable PSG system and slept in their natural home environment. Objective sleep outcomes included standard macrostructural (e.g., sleep onset latency, total sleep time, sleep efficiency, wake after sleep onset) and architectural parameters (stage percentages, latencies and durations). Upon awakening, participants completed a sleep diary.

Mixed-design ANOVAs were used to assess the effects of Condition and Group on sleep parameters.

Results

No significant results were observed for macrostructural sleep parameters. A significant main effect of Condition was found for REM sleep percentage (p= .013), with lower REM% following emotional compared to neutral VR exposure. A similar non-significant trend was observed for REM duration (p= .053).

Significant Condition*Group interactions emerged for N3 percentage (p= .021) and REM latency (p= .045), indicating that emotional valence differentially modulated sleep architecture. Specifically, opposite patterns were observed between positive and negative groups. Similar non-significant trends were observed for additional N3-related parameters.

Conclusion

These preliminary findings suggest that emotionally engaging VR experiences before sleep selectively influence sleep architecture. While REM sleep appears to be generally reduced after emotional stimulation, deep sleep shows valence-dependent modulation, with opposing effects for positive and negative experiences. Notably, VR exposure did not appear to disrupt overall sleep continuity. These results enhance understanding of emotion-sleep interactions in ecologically valid contexts and may inform the safe and effective use of VR technologies.

ABSTRACT. Alcohol craving - an intense desire or sense of compulsion to consume alcohol - represents a subjective motivational drive toward alcohol consumption. Craving is recognized as one of the factors in the development of heavy drinking patterns, alcohol use disorder, and the increased likelihood of relapse. Exposure to alcohol-related cues, including visual, olfactory, or contextual stimuli, is known to intensify craving and facilitate drinking behavior, especially among individuals who engage in heavy social drinking. Within cue-reactivity framework, craving is understood as a conditioned response which is triggered by alcohol-related stimuli in the person’s environment. This perspective highlights the importance of ecologically valid cue presentation. Traditional cue-exposure paradigms have typically relied on non-realistic laboratory settings, potentially limiting ecological validity. Novel approaches include the use of virtual reality (VR) technology in designing interventions based on cue-exposure, enabling a high level of experimental control, ensuring intervention standardization and reproducibility, while simultaneously enhancing ecological validity and allowing for flexible personalization of the user experience. On the other hand, many existing studies utilize computer-generated VR environments that may diminish realism and complexity of real-world social situations, therefore limiting immersion. Use of 360-degree videos in virtual reality may represent methodological advancement regarding these limitations, as it allows participants to experience complex photo-realistic immersive environments, while maintaining high experimental control. As part of the research project Nightlife: A Study in Real and Virtual Context - REAL NIGHTS, we developed immersive 360-degree videos in virtual reality, depicting three nightlife social settings - a nightclub, a bar, and a house party. Participants will be able to select their preferred alcoholic beverage (wine, beer, hard liquor) and musical genre (rock, EDM, pop, trap/folk), enhancing personalization of the experience. These VR environments are intended for use in an experimental study examining cue-induced alcohol craving across a protocol including three separate sessions. Prior to the main experimental study, a feasibility study will be conducted to evaluate the developed REAL NIGHTS VR application in terms of user experience and overall quality of VR environments. This feasibility study aims to assess sense of presence and occurrence of general body and eye-related symptoms related to exposure to REAL NIGHTS VR environments. Based on the existing literature on the use of VR in cue-reactivity studies, we hypothesize that participants will report high levels of sense of presence after exposure to REAL NIGTS VR environments. We also expect them to report negligible levels of unpleasant symptoms (e.g., nausea, eyestrain) after exposure. Participants will be recruited using convenience sampling and exposed to one of three REAL NIGHTS VR environments, with approximately equal number of participants assigned by chance to each environment. Before exposure, data will be collected on participants’ sociodemographic characteristics, and their previous experience with VR. Participants will receive standardized instructions on using the VR software. During exposure, participants will repeatedly report their alcohol craving levels using a 0-10 virtual numeric rating scale (VNRS). After exposure, which will last approximately 10 minutes, participants will evaluate their overall experience with REAL NIGHTS VR. This study obtained ethical approval by the Ivo Pilar Institute of Social Sciences Ethics Committee. This abstract describes work in progress. Data collection and analyses are ongoing at the time of submission. Findings from the feasibility study will be presented, providing preliminary evidence regarding the realism and immersion level of the REAL NIGHTS VR environments, as well as their potential to elicit cue-induced alcohol craving. These results will inform the design of subsequent experimental study and contribute to advancing VR-based methodologies in addiction research by validating 360° video as a useful and ecologically valid tool for cue-exposure paradigms.

ABSTRACT. Background: We previously conducted a study of 3MDR delivered in the Computer Assisted Rehabilitation Environment (CAREN). This study featured several novel elements in addition to the novelty of 3MDR itself. First, the study population was 50% female, whereas prior studies had almost exclusively males. Second, all participants had a history of mild traumatic brain injury in addition to PTSD. Finally, study participants were randomized to therapy with or without an eye movement (EM) task. We achieved clinically and statistically significant improvement in participants on the PTSD Checklist for DSM-5 (PCL-5), with scores declining from 52.0 (46.3, 57.7) at baseline to 33.6 (24.3, 42.9) post-intervention, greater with the EM task. Gains were sustained at 3 and 6 months of follow up. Women achieve better results than men. However, given the expense of the CAREN, we sought to demonstrate efficacy with less expensive delivery, to make 3MDR more accessible. Second, given the use of music and pictures, we sought to determine whether providing music and art therapists could further enhance the response to 3MDR.

Methods: CARE4PTSD Computer monitor versus Augmented Reality: Expanding 3MDR Therapy for PTSD: A Randomized Controlled Trial is a clinical trial seeking 60 participants with PTSD. All participants receive 3MDR, but are randomized 1:1 to delivery via either a 34-inch curved screen computer gaming monitor (Dell, $300), or AR head-mounted display (Microsoft Hololens2, $3500). The therapy encompasses 10-14 sessions: three preparatory, 6-10 intervention, and one reconsolidation. Therapist and participant select at least 2 songs and 14 pictures to use in the intervention. Sessions begin with a song to recall the time of the trauma. They then approach a traumatic image until it looms before them, which they talk through with the therapist. This is repeated for up to 7 images, each followed by an EM task. A second song returns them to the present to session’s end. Assessments with the PCL-5 after sessions 6 and 8 guide determinations about whether to conduct additional sessions. The primary outcome is change in CAPS-5 and PCL-5 scores from pre- to post-intervention, with follow-up at 3 and 6 months.

MATE-3MDR Music and Art Therapy to Expand 3MDR Treatment for PTSD is a clinical trial in which participants are randomized to either “traditional” 3MDR as described above, or to have that supplemented by meeting with both art and music therapists during the preparatory phase to help them develop a work of visual art that for use in the intervention sessions, and to choose music for the sessions. In addition, participants in the MATE arm later meet with the art and music therapists to focus on developing another work of art, and choosing another song, that focus on post-traumatic growth. The music is to be used at the end of ensuing intervention sessions, and the art in the final reconsolidation session.

Results: For the first six participants to complete CARE4PTSD, regardless of delivery method, all achieved clinically significant improvement, with mean scores on the PCL-5 improving from 49.7 to 25.3. The first four participants to complete MATE-3MDR have all also achieved clinically significant improvement, with mean scores on the PCL-5 improving from 50.3 to 16.8.

Conclusions: 3MDR consistently achieves clinically and statistically significant improvement in PTSD symptom severity. Both delivery mechanisms, as well as the addition of art and music therapists, have been well-received by participants. By the time of the meeting, sufficient numbers may be achieved in each study to be able to interpret whether one delivery mechanism appears superior to the other, as well as to comment upon the potential added value of art and music therapists.

ABSTRACT. Motion-Assisted, Multi-Modal Memory Desensitization and Reconsolidation (3MDR) has emerged as a promising extension of exposure-based psychotherapy for treatment-resistant posttraumatic stress disorder (PTSD) and related trauma-spectrum conditions. By integrating treadmill-assisted movement with personalized trauma-related visual stimuli and narrative processing, 3MDR aims to enhance emotional engagement, reduce avoidance, and facilitate memory reconsolidation. This presentation reviews key advances in the development, empirical evaluation, and clinical implementation of 3MDR. The intervention originated from adaptations of the Computer Assisted Rehabilitation Environment (CAREN), initially designed for physical rehabilitation and later repurposed for immersive psychological treatment. A growing body of randomized controlled trials conducted in the Netherlands, the United Kingdom, the United States, and Canada demonstrates clinically meaningful reductions in PTSD symptom severity alongside high treatment adherence, particularly among populations that have not responded to conventional therapies. Technological and procedural refinements—including improved software usability, high-definition immersive visuals, and real-time stimulus customization—have further strengthened the feasibility and therapeutic impact of 3MDR. Ongoing extensions include pilot applications in adolescents, individuals with moral injury, and patients with complex PTSD. Implementation efforts are supported by a structured training and supervision framework coordinated through the International Consortium of 3MDR, emphasizing fidelity and therapist competency. Emerging innovations, such as augmented reality delivery via HoloLens, are being explored to increase scalability and accessibility. Overall, 3MDR is presented as a patient-centered, scalable intervention that addresses key limitations of traditional exposure therapies and offers a novel pathway for advancing trauma treatment in clinical and translational contexts. This presentation will also describe experience with adapting 3MDR for use in the Ukraine, and the training of scores of behavioral health providers there to address widespread PTSD caused by the Russian invasion and ongoing destruction during the continuing war. Challenges posed include the continuing traumatization as well as widespread power outages that have required technological innovation to be able to maintain therapy.

ABSTRACT. Background: 3MDR is an innovative intervention that has originally been developed for military members with posttraumatic stress disorder (PTSD) who did not sufficiently benefit from traditional trauma-focused therapies. Delivered in an immersive VR environment customized with personal images and music, the treatment focuses on overcoming persistent avoidance and optimizing engagement. Multiple studies have documented the effects of 3MDR in veterans with treatment-resistant PTSD. So far, no studies were available that made a direct comparison between 3MDR and active trauma-focused psychotherapies as first treatment for PTSD, and no systematic evaluation of 3MDR was available for patients with traumatic grief. Methods: Data are presented from a randomized controlled trial we conducted comparing 3MDR and regular trauma-focused psychotherapy in 134 military veterans and first responders in the Netherlands. We included adults diagnosed with occupational or combat-related PTSD, who had not received prior treatment for their condition. Half of the patients received 3MDR, which was offered in 10 weekly sessions including one preparatory session, six sessions on the platform, and three integrative sessions. The 3MDR platform consisted of a dual-belt treadmill and a synchronized virtual reality environment, comprising a 180-degree projection on either 3 screens by 3 projectors or 6 large television screens connected to each other and a surround sound system. The other half of the patients received up to 16 sessions of a regular manualized trauma-focused treatment as described in the Dutch treatment guidelines. Assessments were performed at baseline, post-treatment, 3 and 6 months after treatment. We will also present data from an ongoing open trial on an adapted version of 3MDR focused on the treatment of traumatic grief, which encompasses symptoms of prolonged grief disorder, major depression, and PTSD. Assessments took place at baseline, post-treatment and at 4-month follow-up. Results: In the randomized controlled trial, 3MDR proved to be non-inferior to standard trauma therapy on self-reported PTSD as measured by PCL-5 at 6 months post-treatment (mean difference -2.91 [95% CI -7.92, 2.10], p=0.25), even though it was given in fewer treatment sessions. Significant time effects were found for both treatment conditions on self-reported and clinician-rated PTSD, comorbid symptoms, avoidance and functioning. Promising improvements were found in 3MDR for traumatic grief (3MDR-TG), corresponding with reliable changes in key traumatic grief symptoms as measured by the TGI-CA at 4-month follow-up. Conclusions: In the randomized controlled trial 3MDR demonstrated to be an effective alternative first-line treatment for veterans and first responders with PTSD. Even though it leans on infrastructure with a treadmill and other hardware components, it may offer an alternative over conventional trauma-focused psychotherapies for PTSD that yields savings of a quarter of time spent within therapy. For traumatic grief, the promising findings pave the way for further systematic study.

ABSTRACT. Background: Multimodal motion-assisted memory desensitization and reconsolidation (3MDR) therapy is a novel, exposure-based treatment for posttraumatic stress disorder (PTSD) conducted in an immersive virtual reality environment. 3MDR augments the patient’s visual and auditory immersion by incorporating patient-chosen pictures and music, as well as physical movement (walking), to facilitate accession of traumatic memories. The key elements of 3MDR posited to achieve reduction in PTSD symptoms are (a) minimizing patterns of avoidance; (b) increasing efficacy in engaging traumatic memories; (c) experiencing and labeling trauma-related emotions in the here and now; and (d) restoring the memory, sensory, and affective information associated with these memories. Evidence from four randomized controlled trials support 3MDR as an efficacious treatment for PTSD.

Experiential avoidance (EA) describes a person’s unwillingness to remain present with bodily sensations, emotions, thoughts, and memories and their tendency to engage in various forms of avoidance. Associations between EA and PTSD are well established, even when controlling for the avoidance symptoms that characterize PTSD. Decreasing EA through exposure-based treatment of PTSD may lead to reductions in PTSD symptoms.

Materials and Methods: This presentation will report preliminary data on PTSD symptoms, measured by the PTSD Checklist, and EA, measured by the Acceptance and Action Questionnaire-II, from participants completing 3MDR as part of an ongoing feasibility study conducted at the Naval Health Research Center. To date, 15 active-duty service members (M=34.00±8.64 years) have completed the treatment. A linear mixed model was conducted to examine the main effect of time on PTSD symptoms and on EA. A lagged regression model was conducted to examine the time-lagged effect of EA on PTSD symptoms and the time-lagged effect of PTSD symptoms on EA.

Results: PTSD symptoms significantly decreased over the course of treatment (B=-0.78, SE=0.25, t=-3.09, p=.008, d=-1.13) as did EA (B=-0.53, SE=0.17, t=-3.19, p=.007, d=-1.16). Session-to-session reductions in EA significantly predicted successive reductions in PTSD symptoms (B=0.48, SE=0.14, t=3.41, p=.002, d=1.24). In contrast, session-to-session reductions in PTSD symptoms did not significantly predict successive reductions in EA (B=0.07, SE=0.05, t=1.39, p=.16, d=.51).

Conclusions: These preliminary findings suggest that EA may be a mediator of 3MDR treatment effects on PTSD. Further study is warranted with larger samples.

ABSTRACT. Extended reality (XR) is poised to play a critical role in the emergence of precision medicine, informing, enabling, and optimizing activities from discovery to real-world implementation. Despite rapid advances in genomic and other -omic technologies, and the proliferation of targeted therapeutics, major gaps remain in translating individualized data into routine clinical practice. Challenges include the complexity of multi-dimensional data, limited tools for dynamic clinical decision support, suboptimal patient understanding and engagement, and difficulty forecasting real-world implementation outcomes. XR offers embodied, interactive, and context-aware tools that may overcome limitations associated with existing approaches in certain domains.

This presentation introduces a comprehensive framework for integrating XR throughout the precision medicine ecosystem. Conceptually, the framework spans key stages from early discovery, through clinical trial design and optimization, to point-of-care decision support, patient-facing assessment and education, and longer-term prediction. It considers where linking XR platforms with genomic, imaging, sensor, electronic health record, and other data streams may be particularly promising.

The objectives of this presentation are to, first, articulate a structured framework for XR deployment across the precision medicine pipeline, and second, present empirical evidence demonstrating feasibility, validity, and effectiveness of XR-based tools for these purposes.

Empirical data from several ongoing and completed projects will be presented to illustrate the capabilities and limitations of XR at each stage of the framework. Examples will include: validation of XR tools to measure clinically relevant patient behavior and functioning (e.g., assessment of food choice behavior in the context of pharmacological trials); trial data from assessment of XR-based clinical interventions in ADHD, anxiety, and other clinical domains; evaluation trials assessing the effectiveness of immersive patient education experiences in genomics education; and creating interactive simulations to forecast clinical integration outcomes and develop best practices (e.g., in provision of polygenic scores to promote disease prevention).

By anticipating and experimentally probing the convergence of XR with sensors, imaging, AI, and other emerging technologies, we can optimize and create new approaches for applying -omic and individualized data in healthcare to enable precision medicine practice. The presentation will showcase research examples while critically examining both potential and inherent risks.

ABSTRACT. This paper asserts that the reason virtual reality (VR) interventions are piloted in clinical settings but not always then embedded in services is, in part, a failure of researchers to consider the service fit from the perspective of both potential users and clinical staff. This paper also discusses a study in progress that considers dual perspectives and logic model validation as a foundational work towards an RCT for inpatients with Spinal Cord Injury (SCI) in a rehabilitation hospital in Ireland. Inpatients in rehabilitation hospitals, adapting to often traumatic sudden disability that comes with spinal cord injuries, have the potential to experience significant psychological distress. Loss of control and freedom, often associated with inpatient stays, may compound the distress. While much VR-based research has taken place with this cohort (Alashram, 2025; Qian et al., 2020), most are one-off small sample pilots that focus on the technology, motor and cognitive rehabilitation and not the underlying psychological mechanisms of well-being that psychological rehabilitation requires. A recent SCI may involve a sudden biographical shift that requires new identity construction (Bourke et al., 2015); therefore, generalised VR well-being tools cannot be considered without being filtered through the specific needs of this population. The current lack of robust RCTs in SCI-VR research is driven by low sample sizes and poor methodology, leaving clinicians without evidence-based protocols (Lanier et al., 2019). While pilot studies reported improved quality of life for SCI participants, critical clinical questions remain: How should evidence-based VR protocols be introduced? Are they effective? And at what stage of rehabilitation is hedonic or eudaimonic VR most appropriate? This study adopts key aspects of an established health intervention Feasibility Framework (Bowen et al., 2009), focusing on the domains of practicality and limited efficacy, while the Theoretical Framework of Acceptability (TFA) (Sekhon et al., 2017) guides evaluation of the potential burden for both patients and staff. The VR experience is grounded in the Cognitive Vitality Model (CVM) (Howlin & Rooney, 2020). The study design is a mixed-methods, dual stakeholder exploratory feasibility study in preparation for a pilot intervention evaluation study. The study involves a sample of 15-20 inpatients and 5-10 clinical staff from the Spinal Cord Injury (SCI) programme at Ireland’s National Rehabilitation Hospital (NRH). A choice-based intervention where participants select from three distinct, theoretically driven environments: 1) Calm: Focused on emotional regulation. 2) Awe: Focused on "transportation" and perspective-shifting. 3) Narrative Meaning: Eudaimonic experiences promoting identity and biographical reconstruction. Data is collected with all participants assuming a dual role after undergoing the VR experience. Phase 1 (The Participant): Captures immediate phenomenological and emotional data regarding the VR experience. Phase 2 (The Consultant): Participants pivot to an advisory role, using lived (patient) or professional (staff) expertise to evaluate service fit, accessibility, and optimal timing within the rehab cycle. Qualitative thematic analysis of feedback, alongside scales relating to emotional experience and pain levels, informs the VR protocol refinement. Quantitative feasibility metrics, recruitment rates, retention, and scores on the Digital Health Acceptability Questionnaire DHAQ (Haydon et al., 2023) address research questions relating to feasibility and practicality in the clinical setting. By situating the investigation within a live clinical service and adopting a dual-stakeholder approach, this study explores psychological mechanisms while simultaneously addressing practical implementation concerns, thereby contributing knowledge that is relevant to both theory and practice.

ABSTRACT. Brain–computer interface (BCI) technologies and non-invasive neuromodulation have become increasingly important tools in cognitive therapy, physical rehabilitation, and digital mental health interventions. In particular, transcranial direct current stimulation (tDCS) has demonstrated potential benefits in the treatment of depression, emotional dysregulation, and a range of neurological disorders by modulating cortical excitability and functional brain networks. Despite promising laboratory and clinical findings, the real-world translation of tDCS-based interventions remains limited by the size, rigidity, and poor digital integration of existing stimulation devices. In this work, we present a miniaturized, battery-powered wireless direct current stimulation platform designed to function as a brain–computer interface node for cognitive and physical therapy applications targeting depression, emotional disorders, and neurological conditions. The proposed system integrates a precision constant-current stimulation module, real-time electrical sensing, and low-power wireless communication into a compact wearable device suitable for daily-life and home-based use. At the hardware level, the system employs a low-offset operational amplifier–based current source with a high compliance voltage, enabling stable and accurate stimulation across varying electrode–skin impedance conditions commonly encountered in long-term and non-clinical settings. An embedded monitoring module continuously measures stimulation voltage and current, providing safety-aware feedback and supporting adaptive regulation during stimulation. This design addresses a key limitation of conventional open-loop tDCS devices, which often lack real-time awareness of stimulation conditions and user variability.System validation results demonstrate that the device successfully achieves a miniaturized form factor of 5 × 5 × 2 cm while maintaining a stable constant current output from 0.5 to 2 mA with a high step precision of 0.1 mA. Performance testing under simulated high-impedance loads confirms that the integrated 24V Boost converter and ADS1115-based feedback loop maintain an output error within ±10, even in non-sensor modes. The hardware exhibits an exceptionally rapid response time of 10 ms for current stabilization and a wireless control latency of ≤100 ms via WiFi, facilitating reliable real-time remote parameter adjustment. Battery life tests show that the system can operate continuously for over 4 hours at maximum stimulation intensity on a single charge. Furthermore, biological verification using a mouse model of depression indicates a significant mitigation of depressive-like behaviors, as assessed by standardized behavioral metrics, with no observable skin damage or physiological toxicity reported . Beyond standalone neuromodulation, the proposed device is explicitly designed for integration into closed-loop brain–computer interface and digital therapy frameworks. Wireless connectivity allows seamless interaction with external software platforms, enabling remote supervision by clinicians, therapists, or researchers, as well as cloud-based data storage and analysis. Such connectivity supports emerging therapeutic paradigms in cyberpsychology and cybertherapy, including remote treatment of depressive symptoms, adaptive emotion regulation protocols, and long-term neurobehavioral monitoring in naturalistic environments. The system architecture supports personalization and adaptability, which are critical for mental health and neurorehabilitation applications. Stimulation parameters can be dynamically adjusted based on behavioral performance, user-reported emotional states, or physiological signals acquired from complementary sensing modalities. This capability enables the development of closed-loop, data-driven intervention strategies that align neuromodulation with cognitive and physical therapy goals. Key innovations include system-level miniaturization of a fully functional constant-current stimulation device, real-time monitoring and safety-aware closed-loop operation, and a clinically oriented design philosophy emphasizing usability and personalization for depression and neurological disorders. The proposed system provides a technological foundation for future research in brain–computer interfaces and supports scalable, personalized interventions for neurobehavioral rehabilitation.

ABSTRACT. Introduction: Augmented reality (AR) and virtual reality headsets have expanded the capabilities of clinical care across a variety of fields including traumatic brain injury, vestibular disorders, Parkinson’s disease, and posttraumatic stress disorder. However, technical limits of these systems, specifically a restricted field of view (FoV) and synthesized 3D audio, may interact with preexisting sensory impairments. To safely implement AR in warfighter rehabilitation, it is essential to understand the working field within a headset and its implications for user performance. This study serves as a critical first step in a multi-process research initiative. We sought to develop a comprehensive "performance map" within a 180° working space to define the normative biomechanical and efficiency baselines in healthy service members. Establishing this baseline is a clinical prerequisite, as it allows for the future identification of "performance gaps" in patients with TBI or vestibular disorders (study currently underway) who may lack the compensatory strategies observed in healthy cohorts. Materials and Methods: Healthy active duty participants (n=18) completed visual search tasks and audio localization tasks where targets were presented in a 180 degree working space. Tasks were completed on a Magic Leap One AR (horizontal FoV, ±20 degrees; vertical FoV, ±15 degrees) headset and a Computer Assisted Rehabilitation Environment (CAREN) system, used to mimic real-space performance. To understand the effects of audio within virtual spaces, a subset of participants completed tasks with real speakers versus 3D audio. All tasks were performed while standing and walking to monitor biomechanical stability (head rotation, step length, step time, and base of support). Results: In visual search tasks, targets outside the +35 degrees and -8 degrees azimuth and -11 degrees elevation resulted in significantly longer completion times (p<0.001) in AR compared with the CAREN. To compensate for the limited FoV, participants exhibited significantly greater head rotation for vertical targets (p<0.001). Reliable audio cues tended to mitigate the effects of limited FoV in AR. Despite the increased visual search burden, gait parameters (step length, step time, and base of support) did not differ significantly between AR and CAREN, suggesting that healthy participants are able to prioritize locomotor stability when sensory input is constrained. Conclusions: This performance map, scalable to other AR devices, demonstrates that hardware constraints can be leveraged as programmable clinical parameters. Clinicians can use these boundaries to gradually challenge visual and auditory systems to encourage therapeutic head movement or sensory weighting.

ABSTRACT. Mild cognitive impairment (MCI) is a clinical state between normal aging and dementia and a key window for early detection. In MCI, while episodic memory is often most affected, executive dysfunction is also associated with an increased risk of progression to severe cognitive impairment and could become the predominant pattern, with broader cognitive inefficiency and early behavioural changes. Cognitive decline is also strictly linked to age-related changes in physical health, psychological wellbeing, and social functioning. Traditional neuropsychological tests like the Trail Making Test (TMT) and Modified Five Point Test (MFFT) often lack sensitivity and ecological validity, missing everyday executive challenges. Digital health technologies offer a promising way to address these limitations by collecting quantifiable measures called digital biomarkers. In particular, immersive virtual reality (VR) can recreate realistic environments, modulate distraction, and capture performance metrics and kinematic signals. Several studies have confirmed the validity and feasibility of digital biomarkers in MCI, highlighting the predictive value of motor variables, such as gait speed, stride length, and stride variability, in detecting cognitive decline related to executive dysfunction. Digital biomarkers offer more sensitive, less invasive, and more accessible measures than standard clinical biomarkers, while generating datasets to feed Artificial Intelligence (AI) and Machine Learning (ML) models able to detect early, subtle changes. However, how these digital biomarkers integrate with psychosocial determinants of cognitive decline remains underexplored. This paper presents the protocol of a cross-sectional study which identifies digital biomarkers of executive functioning through immersive virtual reality, integrating psychosocial and health-related risk factors to improve early MCI detection. Some preliminary findings are also provided. The overall study pursues three aims: (i) to determine whether VR-derived executive biomarkers discriminate older adults with MCI from healthy and frail older adults; (ii) to examine convergent validity between VR metrics and traditional executive-function measures; and (iii) to test whether integrating VR biomarkers with psychosocial and health-related factors increases predictive accuracy using ML classification models. The study adopts a quantitative, cross-sectional design and recruits participants aged ≥65 years, distributed across three groups: healthy older adults, frail older adults, and individuals with MCI. Participants complete a neuropsychological battery, followed by self-report questionnaires capturing health-related and psychosocial variables. Then, two VR tools are used to obtain a richer phenotype of executive functioning: (i) the Nesplora IceCream Executive Functions tool offers a structured scenario assessing key executive components (e.g., planning, working memory, cognitive flexibility) via objective indices, while capturing digital biomarkers from head and hand kinematics; (ii) the Picture Interpretation Test 360° (PIT 360°) assesses executive functioning by making participants scan a 360° scene using natural head movements with contextual cues to form an interpretation, yielding metrics such as time to correct interpretation, information sampling strategy, and head-movement biomarkers. Data analysis includes group-level comparisons across healthy, frail, and MCI participants and convergent-validity analyses to test the associations between VR metrics and standard measures of executive function. Incremental, feature-wise comparisons are used to test whether integrating VR-derived biomarkers with psychosocial and health-related factors improves the predictive accuracy of ML-based classification models, compared with models based on digital biomarkers alone. Within this framework, psychosocial and health-related factors may help explain additional variability in inefficient performance and altered motor patterns captured in VR, thereby improving risk stratification through a more contextualized and holistic profile of cognitive vulnerability. So far, preliminary convergent-validity analyses on a subsample of 31 participants have shown encouraging associations between VR-based indices and paper-and-pencil executive measures, particularly for working memory and planning. These findings support the ecological value of VR assessment and its potential for informing predictive models of cognitive vulnerability, although confirmation in larger samples is needed.

ABSTRACT. Background and Objectives: Cyberscam survivors often experience significant financial losses and devastating psychosocial consequences. However, victim aftercare is often overlooked, with limited evidence-based therapeutic interventions available. A novel cyberscam psychosocial intervention (“Smooth Sailing After Scams”; SSAS) was co-designed and piloted. The SSAS program consisted of ten weekly group or individual 1-hour therapy sessions was designed to address 1) cybersafety, and cyberscam-related impacts on 2) finances, 3) emotions, 4) relationships and lifestyle. Allied health clinicians delivered the program either in-person or via teleconferencing to scam survivors and their family/close others. Peer facilitator “Scambassadors” contributed to up to four sessions and helped engage participants in the program. Due to earlier findings of increased scam susceptibility in individuals with acquired brain injury (ABI), the program was initially piloted with this cohort. The SSAS pilot demonstrated encouraging quantitative findings regarding its potential to address scam-related impacts in a single-case-experimental design evaluation. Given individual responses to interventions vary, it is important to understand participants’ perspectives of the intervention. This study aimed to explore participants’ qualitative experiences of the SSAS program. Method: Institutional ethics approval was obtained with all participants providing informed consent. Twenty-two adults (68% female, Mage=54.7, SD=17.5) participated including clients (n=8) and Scambassadors with ABI (n=3), close others (n=8) and clinicians (n=3). Semi-structured 1:1 interviews were conducted via videoconferencing, recorded and transcribed verbatim. Memory prompts were provided to maximise participation of participants with ABI and social-communication difficulties. Braun and Clarke’s six-phase reflexive thematic analysis was used to iteratively identify and refine latent themes that reflected a range of participant experiences. Results: Clients had experienced buying and selling, investment and cryptocurrency, charity, toll, and dating and romance scams. Scam-related financial losses ranged from $0 (near-misses) to over AUD$300,000. Participants entered the intervention with varying levels of scam engagement-disengagement, insight, distress and adjustment. Overall, participants regarded the SSAS program as a valuable experience. Seven themes were generated reflecting factors supporting intervention process enablers, experience enhancers and outcomes. Intervention Enablers encompassed: 1) “Sense of community”, demonstrating group intervention factors that supported or hindered participant engagement, 2) “Therapeutic alliance”, highlighting the important role of a trusting clinician-participant relationship to foster a supportive environment, and 3) “Well-run and well-managed”, indicating the organised delivery of the intervention within the service. Therapeutic Enhancers comprised themes that were both intervention facilitators and outcomes: 4) “Increasing insight through the scam adjustment journey”, demonstrating the use of a conceptual formulation infographic in understanding post-scam experiences and recovery, and 5) “It’s brought us closer”, showing the social benefits of close other involvement as a mechanism improving social connection and collaborative scam adjustment. These resulted in Positive Outcomes: 6) “I learnt how to stay safe online” (i.e., increased cyberscam knowledge) and 7) “I don’t feel shame” (psychological adjustment). Practical recommendations for program refinement were also provided. Conclusions: In-depth qualitative interviews with scam-survivors with ABI, close others, Scambassadors and clinician participants involved in a novel cyberscam psychosocial adjustment intervention indicate it resulted in positive outcomes in terms of cybersafety, emotional wellbeing and social support. Despite varying levels of initial scam experiences, losses and adjustment, the intervention was widely reported to provide meaningful benefits. Outcomes were driven by a group format, ABI-specific cohorts, individualisation, conceptualised post-scam adjustment as a journey and strong therapeutic alliance. These reflections provide corroborative evidence for the feasibility and acceptability of the SSAS Program, with additional feedback and recommendations providing valuable direction for program refinement. Given the subjective nature of qualitative research and the small previous pilot, this program now requires more rigorous evaluation through a randomised controlled trial. Importantly, this study advances the evidence-base towards a scalable empirically-validated scam aftercare support program.

ABSTRACT. Whilst video games are increasing in popularity, with 1.93 billion players worldwide (Clement, 2025), there are still concerns surrounding how video games influence human behaviour, with parents (Ofcom, 2017, 2024), the justice system (Brown v. EMA, 2011) and researchers (Olejarnik & Romano, 2023; Lacko et al., 2024) being weary of video game impacts. As the wellbeing literature is fragmented and delivers mixed results, there is a need for a comprehensive investigation of how video games influence different facets of player wellbeing. This work-in-progress presentation showcases our findings to date on whether video games influence multidimensional wellbeing – validation of a multidimensional wellbeing framework, cross-sectional analysis of gaming and wellbeing data, and panel data analysis of gaming, behavioural problems and wellbeing data.

As video game psychology overfocusses on psychological wellbeing, we began by recontextualising wellbeing as multidimensional, as is done in other areas of cyberpsychology. We did so by drawing on literature on how video games influence wellbeing, formulating the Player Multidimensional Wellbeing (PMDWell) framework (Olejarnik & Romano, 2025), which posited that player wellbeing is comprised of four dimensions: social functioning, mental health, physical health and life circumstances. We validated the PMDWell framework by conducting confirmatory factor analyses using data from 443 players worldwide, offering a 54-item questionnaire to probe multidimensional wellbeing.

Using a subset of the same wave of data (N = 289), we conducted a cross-sectional study to investigate how gameplay hours influence the interaction between wellbeing dimensions and gaming disorder (GD) symptoms. Using regression analyses we found that the higher the gametime, the worse the mental health. Simultaneously, the higher the GD symptoms, the worse the physical health and higher the online social functioning. Follow-up regression analyses, splitting the sample into Youth (≤21 years old) and Adults, found that GD symptoms were positively associated with online social functioning in Adults, but not Youth. A path analysis, linking hours of gameplay by device (Console/PC and Mobile) and game mode supported (Singleplayer, Multiplayer, Co-operative) found that the more time spent playing Console/PC games, the worse the life circumstances, leading to worse mental health, worse offline social functioning and worse physical health, greater gaming disorder symptoms and better online social functioning. These findings suggest that increased engagement with dedicated gaming devices is linked to worse multidimensional wellbeing outcomes, which could lead to increased addiction symptoms, where for Adults more online social capital could be a maintenance factor of GD.

We then conducted a UK-wide panel data analysis using the Understanding Society (University of Essex, 2024) Youth and Adult datasets. We found that cross-sectionally in Youth, increased video game engagement on a school day is predictive of greater behavioural problems, with strong between-person effects, even when accounting for covariates. However, we found that this pattern is not replicable prospectively for lagged analyses of Youth gaming patterns onto behavioural problems 2/4 years later, or longitudinally for lagged and cumulative exposure analyses of Youth gaming patterns onto Adult wellbeing 3, 6 and 10 years later. These findings suggest that at population level within the UK, the negative effects of video games are time-locked – they are short-term and do not influence behavioural difficulties or overall adult wellbeing in the long-term.

Work to date suggests that video games could be a potentially debilitating force in the short-term, but not the long-term, with increasing engagement interacting with external and internal wellbeing to increase GD symptoms. Continuation of the work will aim to investigate the impacts of wellbeing in the short-term (up to 6 months) and to interview gamers to tackle the perception of gamer wellbeing and how it is affected by video games.

ABSTRACT. Video games have become an increasingly central force in contemporary global society, both in economic and socio-cultural terms. In addition to entertainment and leisure, their daily use extends to use as tools for learning, promoting general well-being, and even for clinical interventions. At the same time, studies also highlight problematic aspects with concerns related to excessive internet use, the potential for addiction, or encouraging antisocial behaviours. Accordingly, it is crucial to clarify when gaming functions as an adaptive strategy and when it may shift into a problematic pattern, especially considering the psychological and contextual complexity surrounding gameplay. The present study aims to examine potential relationships between gaming motivations and an individual’s ability to maintain a healthy balance between digital life, physical life, and well-being. Furthermore, a specific focus is on whether specific gaming motivations can predict levels of psychological well-being and digital life balance (DLB). Accordingly, four specific research questions were defined. The first two investigated the potential association between gaming motivations and both psychological well-being (RQ1) and DLB (RQ2). The third and fourth investigated gaming motivations as predictors of both psychological well-being (RQ3) and DLB (RQ4). Cross-sectional data were gathered via an online survey, with participants being recruited through social media platforms (e.g., Facebook, Discord), online forums, and mailing lists. The survey items included measures of sociodemographic data, the Psychological Motivations for Playing Video Games (PMPVGs) scale, the DLB scale, and the Flourishing Scale (FS). Inclusion criteria, besides being of legal age, included being fluent in Italian and having played video games in the last six months. The final sample consisted of 285 participants: 146 males (51.23%), 120 females (42.11%), and 19 participants (6.67%) who did not disclose their gender; the mean age was 25.3 years (SD = 9.54). After data collection, correlation and multiple regression analyses were performed. Results of the correlation analysis showed significant negative correlations between well-being (flourishing) and the gaming motivations of boredom, diversion, habit, and escapism. A larger variety of associations were observed, with significant negative correlations between DLB and the motivations of boredom, competition, mood/stress, diversion, escapism, habit, and self-esteem. Regression analysis revealed that both escapism and habit were strong negative predictors of well-being (flourishing). The overall model with gaming motivations explained R2 = .204 of the variance (adjusted R2 = .169). In respect to DLB, both diversion and mood/stress emerged as significant negative predictors. Here, the model explained R2 = .258 of the variance (adjusted R2 = .225). Results of the correlation analysis suggest that individuals who play games out of boredom, routine, or to avoid real-life concerns tend to report lower levels of psychological well-being. This is supported by the subsequent regression analysis, which confirms that frequently relying on games as a form of avoidance or ingrained routine is associated with undermined psychological well-being. Regarding DLB, mood/stress and diversion were particularly relevant, suggesting that stress-regulation and disengagement-oriented gaming may undermine the balance between digital engagement and daily usage. Thus, DLB reflects not only digital use but also how well it integrates with daily life. This study contributes to the ongoing discussion on adaptive versus maladaptive digital gaming by showing that the consequences of play depend largely on the underlying psychological motives. Considering gaming motivations with respect to DLB, this work suggests a nuanced framework for identifying risk-related patterns of game use beyond traditional addiction-focused perspectives. This has practical implications for preventive interventions and psychoeducational strategies aimed at fostering healthier digital lifestyles, particularly among individuals who use gaming as an avoidance-oriented coping mechanism. Finally, there is a need to explore these factors from other geographic and/or social contexts.

ABSTRACT. The continuing growth in online and mobile gaming is evidence of its appeal to players. However, the presence and severity of toxic behaviour (e.g., harrassment, hateful content, and abuse) in the online gaming ecosystem is increasing. Although this is true of many digital environments, online multiplayer gaming is particularly challenging in this regard, with many problematic behaviours normalised as new players enter established communities characterised by competitiveness and both formal and informal hierarchies.

The potential for toxicity is enhanced through online disinhibition and anonymised communication (1), meaning that players are psychologically disconnected from both the targets and consequences of toxic behaviour. Toxicity is problematic for the games industry as it increases player churn and reduces monetisation, while also seriously impacting players’ mental health and well-being.

While most online games incorporate some form of moderation, existing methods have been criticised for their predominantly reactive and punitive approach. Indeed, many place the burden on players to initiate moderation procedures. Furthermore, these processes have been criticised as slow, inconsistent, and for not addressing the underlying causes of toxic behaviour.

This research seeks to understand how best to utilise proactive strategies for combatting toxicity in online gaming environments, thereby benefitting companies and improving player well-being. Accordingly, the research is structured around the following research question:

“What proactive strategies are used to mitigate toxic behaviours in online gaming environments, and what are the benefits and limitations of these strategies?”

Recent research shows academic literature addressing proactive approaches is extremely limited (1), meaning systematic or scoping reviews are not applicable due to the dearth of publications. Therefore, this research used narrative review, incorporating both scientific and grey literature. The method is particularly suited to evaluating this diverse body of material as it is both a flexible and practical approach to providing an overview of the current state-of-play; furthermore, a key strength is a focus on interpretation, rather than evaluation, to identify gaps and outline a future research agenda.

No time restraints were applied to the search which identified 15 relevant materials in academic databases using the following query: ((online OR multiplayer) AND (gam*) AND (toxic*) AND (proactiv* OR preventativ*) AND (moderat*) AND (tools OR strateg* OR tech*)).

Grey literature (26 items) was identified using Google searches and cross-referencing; materials were deemed relevant if they addressed either theorised or actual strategies.

Four contemporary strategies for proactive moderation were identified: Real-Time Moderation (RTM); Shaping Community Norms (SCN); Change Individual Behaviours (CIB); and Predictive Modelling and Intervention (PMI). In RTM, Artificial Intelligence (AI) is used to improve current moderation approaches by limiting exposure to toxicity. SCN attempts to reduce toxicity through promoting prosocial behaviours among online communities. CIB combines aspects of both RTM and SCN to deliver behavioural interventions to individuals who perpetrate toxic behaviour. While PMI builds models trained on player data to predict potential toxicity before it occurs. Additionally, four themes were identified which can be used not only to contextualise how these strategies have developed and how they operate, but also to identify potential new strategies.

This work makes several practical and theoretical contributions. First, it produces knowledge about the current state-of-play by identifying contemporary strategies for proactive moderation of toxicity and their contextual relevance. Second, it highlights core themes which provide a frame for interpreting proactive moderation strategies. Finally, it establishes a research agenda that can guide future work.

1. Wijkstra, M., Rogers, K., Mandryk, R. L., Veltkamp, R. C., & Frommel, J. (2024). How to tame a toxic player? A systematic literature review on intervention systems for toxic behaviors in online video games. Proceedings of the ACM on human-computer interaction, 8 (CHI PLAY), 1-32.

ABSTRACT. Digital mental health technologies and serious games for youth are often developed in response to crisis, pathology, or clearly visible high-risk behaviour. While these approaches are important, they often overlook a large group of non-high-risk young adults who appear to function well in schools and communities yet experience early psychosocial strain. When such pressures go unrecognised, they may intensify over time. Because these young adults do not fit crisis-based or deficit-oriented definitions of risk, they frequently remain invisible within existing screening tools, intervention programmes, and game-based health systems. Within the Society 5.0 paradigm, this highlights the need to rethink how serious games are designed, shifting towards ethical, human-centred approaches that support well-being through engagement, dignity, and agency rather than surveillance, labelling, or clinical framing.

This qualitative study examines how digital serious games, understood broadly as experiences that support reflection and meaning-making, can operate as preventive psychosocial spaces for non-high-risk urban poor youth in Malaysia. It focuses on how play enables emotional expression, reflection, and engagement. Drawing on youths' lived experiences, the study explores how play is understood and valued as a means of coping with stress, expressing emotions, and making sense of identity within socially and economically constrained environments. Rather than positioning games as tools for assessment or treatment, this research treats play as a psychosocial practice through which early needs and vulnerabilities can surface naturally and safely.

The study adopts a qualitative, interpretivist research design grounded in the Ecological Transactional Framework and Society 5.0 human-centred design principles. Twenty participants were recruited through purposive sampling via educational institutions and youth support organisations serving urban poor communities: fifteen young adults aged 18 to 24 (mean age 20.4 years; 8 female, 7 male), identified as non-high-risk based on continued educational participation, the absence of overt behavioural concerns, and functional daily engagement, and five educators and youth practitioners including school counsellors and youth programme facilitators. Data were gathered through individual, in-depth, semi-structured interviews. Interviews centred on participants' experiences with play and games, including both digital and non-digital forms, with digital serious games constituting the primary focus. Youths reflected on emotional responses to games, social interaction through play, and differences between gaming spaces and school, home, or formal support settings. Perspectives from educators and youth practitioners were included to strengthen ecological interpretation.

Interview data were analysed using narrative thematic analysis, allowing accounts to be understood as coherent stories shaped by context rather than isolated behavioural indicators. This enabled identification of recurring patterns related to emotional safety, agency, belonging, and expression as they emerged in narratives about play. Three themes emerged. The first examines how young adults experience play and serious games as emotionally safe spaces distinct from evaluative, disciplinary, or performance-driven settings. The second explores how early psychosocial needs and vulnerabilities become visible through play-based engagement without risk labels or deficit-based language. The third identifies ethical design principles to guide serious games for health, centring voluntary and self-paced participation, narrative ambiguity that supports player-led meaning-making, and the deliberate avoidance of mechanics that generate a sense of monitoring or assessment.

Findings indicate that ethically designed serious games can function as low-threshold preventive environments supporting emotional expression, reflection, and resilience among non-high-risk young adults. In practice, this points toward design approaches such as optional narrative pathways in place of required emotional check-ins, ambient storytelling that allows meaning to emerge indirectly, and progression systems that do not surface personal data to the player. This study offers a qualitative framework for understanding serious play as a preventive, human-centred practice aligned with the Society 5.0 vision and grounded in inclusivity, ethical engagement, and psychological safety in youth-focused game design.

ABSTRACT. Violent video games (VVGs), an increasingly popular category of game, have been the subject of extensive debate surrounding their potential to negatively effect the players. In contrast, the potential for VVGs to positively impact the player remains underexplored and little regard is given to the attitudes and perceptions of the players of such games. The current research utilised a mixed-methods approach to explore player experiences and perceptions of playing violent video games through the lens of self-determination theory’s basic psychological needs and through the players’ subjective accounts.

An online survey comprised of two questionnaires on basic psychological needs (in general life and through playing VVGs) and four open-ended questions was completed by 101 participants aged between 18 and 52 (M = 27.25, SD = 7.00). The sample consisted of 76 men (75.2%), 22 women (21.8%), two identified as non-binary (2%) and one participant chose not to disclose gender (“prefer not to say” 1%).

The quantitative results demonstrated that participants exhibited low needs frustration overall and that playing VVGs was associated with psychological needs satisfaction. Additionally, needs satisfaction through playing VVGs was found to be predictive of violent video game playing frequency (measured by the average amount of hours spent playing VVGs per week). A qualitative manifest content analysis revealed three categories regarding the players subjective experiences and attitudes towards the impacts of playing violent video games: (1) Beyond the Violence discusses the popularity and factors that influence individuals to play violent video games, (2) Perceived Benefits of Playing Violent Video Games illustrates the range of benefits that players believe violent video games can offer, and (3) Violent Video Games are NOT the Issue presents the players’ attitudes towards the debate surrounding VVGs. Overall, the findings presented a novel perspective on VVGs, demonstrating that players may view these games as a positive, needs-fulfilling activity, that can potentially provide a range of social, emotional, and cognitive benefits.

ABSTRACT. This study developed and validated the effectiveness of a prevention education program designed to address new forms of digital violence among upper-grade elementary school students. First, considering the evolving nature of digital violence, we discussed the need for such a program, analyzed the key components and limitations of existing domestic programs, and established the direction for the development and structure of this program. Based on this, we finalized the program’s objectives, overall session structure, specific activity materials, teacher’s manual, and case-based lesson plans. As a result, we identified three core elements—‘Understanding Digital Violence’, ‘Attitudes Toward Digital Violence’, and ‘Coping Skills for Digital Violence’—and developed a two-session (80-minute) in-person classroom program based on these elements. Session 1 was designed to enhance participants’ motivation to join the program and improve their digital literacy. It included activities to understand the definition of digital violence and real-life cases such as illegal online gambling, as well as to analyze the perpetrators’ motives, victims’ emotions, and bystanders’ perspectives within various scenarios. Session 2 focused on understanding five representative methods for preventing and resolving digital violence, followed by activities where participants identified their own vaccine against digital violence and made a personal commitment. To verify the effectiveness of this program, it was implemented with 243 upper-grade elementary school students. Subsequently, to address sample size imbalance, 59 participants were randomly selected from the 243 using the random number generator function in SPSS 22. This resulted in a final sample of 59 students in the experimental group and 59 in the control group, which were analyzed using ANCOVA and MANCOVA. The research instrument was developed based on existing prior research, the program’s content and objectives, and the Seoul Internet Addiction Prevention Counseling Center’s scale for verifying the effectiveness of prevention services. To ensure content validity, the instrument was reviewed by two experts in counseling psychology. The questionnaire consisted of a total of 10 items (9 items for the pre-test, excluding the satisfaction item) and was organized into four subdomains: understanding, attitude, coping ability, and satisfaction regarding digital violence among upper-grade elementary school students. The overall reliability of the scale was confirmed with a Cronbach’s α value of .91, and the reliability of each subscale ranged from .70 to .82. The analysis results showed statistically significant improvements in all domains, with particularly notable effects observed in the improvement of coping skills. Additionally, satisfaction with the program averaged 4.69 points, suggesting that, overall, the program was effective in motivating the target group while also delivering educational benefits. Furthermore, feedback from the instructors who delivered the program was collected and analyzed. It was noted that activities designed to spark student interest—such as group discussions, presentations, and creative hands-on activities—led to high levels of engagement and participation. It was also highlighted that the educational effectiveness was further enhanced by allowing students to directly practice self-expression skills and problem-solving strategies, going beyond mere conceptual learning. This study is significant in that it addresses the limitations of existing cyberbullying-focused prevention education, which has failed to adequately address new forms of violence arising from changes in the digital environment. It presents an activity-centered structure suitable for the developmental level of upper elementary students and develops a more effective short-term prevention education program, thereby enabling high on-site accessibility and the potential to establish a multi-layered support system. Furthermore, the study is significant in that it reinforces the acquisition and internalization of response skills through scenario-based practical activities tailored to specific situations, and in that it designs and delivers activities that take into account the developmental stage and media usage patterns of upper-grade elementary students.

ABSTRACT. Self-representation in individuals with substance use disorders can be explored through digital tools that enable the construction of identity-based digital representations, namely avatars. Recent research suggests that such technologies are of interest to mental health professionals, who may use avatars for assessment, intervention, and training purposes in cybertherapy contexts. Indeed, according to avatar self-discrepancy theory, avatars customized by users to represent ideal/actual self can give relevant information on users’ self-perception, which is demonstrated to act as a predictor of psychopathology such as addiction. Building on this background, 20 patients from a residential recovery community were recruited for the study. All participants were undergoing treatment for diagnosed addiction disorders, including alcohol (30%), drugs (60%), and gambling (10%). Using a commercially available platform, participants created three avatars corresponding to three experimental conditions. Following a paradigm previously applied in pathological populations and consistent with avatar-self discrepancy studies, the avatars represented the Actual Self, the Ideal Self, and the Addiction Self (i.e., a self-representation associated with substance use and addiction experience). For each avatar, we recorded creation time (in seconds) and the number of appearance modifications during customization, as a marker of second-thought and uncertainty. Participants also completed Likert-scale attitude measures for each avatar (self-representativeness, attractiveness, and perceived difficulty of creation) and reported emotions experienced during the creation process (valence, arousal, and dominance) using the Self-Assessment Manikin. Demographic and socio-clinical variables were collected (e.g., age, sex, occupation, marital status, type of substance, and time since last use). Results show that avatar creation time was significantly shorter for Actual Self avatars, while Addiction Self avatars were rated as significantly less pleasant. No significant overall effect emerged for the number of modifications; however, a linear trend suggested an increase from Actual Self to Ideal Self to Addiction Self avatars. Participants with more recent substance use reported significantly more intense emotions during Addiction Self avatar creation, as well as more difficult-to-manage emotions across all avatars compared with participants with longer abstinence periods. Further analyses revealed significant differences in emotional dominance related to avatars as a function of occupation, type of addiction, parenthood, and recency of last use, indicating that self-representations expressed through avatars may be sensitive not only to addiction, but also to broader aspects of social role and recovery trajectory. Overall, the present study supports avatar creation as a promising tool for capturing self-representation in clinical populations. Although findings should be interpreted cautiously due to the small sample size and exploratory nature of the study, they provide preliminary evidence for the utility of avatars in cybertherapy of addiction, as they allow professionals to identify differences in self-representations both at the behavioral and the emotional level. Therefore, avatars could be possibly integrated with assessment tools for addiction management and recovery stages.

ABSTRACT. As interactive technologies increasingly aim to induce specific psychological states, from mindfulness applications to virtual reality therapeutic interventions, a critical methodological challenge emerges: how do we validate that a digital experience produces its intended psychological state? This question is relevant for complex emotional states like awe, a self-transcendent emotion associated with psychological well-being but difficult to reliably induce through technology. Standard validation approaches rely heavily on quantitative self-report scales, yet these measures face limitations in cross-cultural contexts where emotional concepts are language-specific and may not fully capture people’s lived experience.

Contemporary emotion science frameworks offer a methodological solution: emotional experiences cluster around recognisable categories while showing substantial individual variation, exhibiting both structure and idiosyncratic expression (Cowen & Keltner, 2021). This means that if a stimulus successfully elicits awe, its phenomenological features (vastness, accommodation, small self) should appear in spontaneous descriptions even without prompting the label "awe." This study applies this phenomenological approach by asking an open-ended question and coding responses for awe's phenomenological dimensions, examining whether participants spontaneously described their experience in awe-consistent terms (e.g., feeling small, connected, part of something larger).

We designed "Passenger," a breath-responsive installation that operationalises awe's phenomenological components (vastness, self-diminishment, connectedness, cognitive accommodation) into concrete interactive design features. We focus here on validation methodology and qualitative findings rather than technical implementation.

Fifty-five participants (M_age = 34.97, SD = 11.53; 28 female, 24 male, 3 non-binary) experienced the installation individually and responded to a single non-directive question in French: "In a few words, how did this experience make you feel?" This open-ended qualitative approach elicits participants’ experience without priming specific emotional labels or biasing responses toward researcher's expectations.

We employed mixed deductive-inductive thematic analysis, combining theory-driven codes from the Awe Experience Scale (self-diminishment, vastness, cognitive accommodation, connectedness, physical sensations) with inductively identified themes that emerged from the data itself (ambivalent feelings, calmness/relaxation, aesthetic appreciation). Each response was binary coded by two independent coders, achieving substantial inter-rater agreement (Cohen's κ = 0.70, range: 0.50–0.84).

Results revealed that 72.7% of participants spontaneously described experiences matching at least one awe dimension. Themes arose in these proportions: calmness/relaxation (52.7%), ambivalent feelings (43.6%), cognitive accommodation (38.2%), vastness (36.4%), and connectedness (27.3%). Notably, ambivalent feelings, the simultaneous experience of emotions of opposite valence, emerged as the second most frequent dimension, aligning with findings that ambivalence is central to awe. However, only 32.7% reported multiple dimensions simultaneously, and just 9.1% showed the prototypical vastness-accommodation combination, suggesting the installation elicited different awe facets in different individuals.

The wide range in dimensional presence (0-4 dimensions) indicates strong individual differences in receptivity to technology-mediated emotional experiences, likely influenced by trait characteristics, prior meditation experience, or media habituation. One participant explicitly noted: "I'm so used to different images (even exceptional violence) online that I have trouble feeling anything", highlighting how digital media exposure may shape emotional responsiveness to interactive experiences.

The high calmness rate (52.7%) alongside awe dimensions suggests what we term a "contemplative pathway" to self-transcendent emotions in interactive contexts: physiological regulation and outward attentional orientation create conditions under which awe-related experiences can emerge for receptive individuals. However, calmness alone does not constitute awe; it may be necessary but insufficient, as 20% showed only calmness without other awe dimensions.

This work demonstrates that phenomenological validation through qualitative analysis provides a rigorous, culturally sensitive alternative to scale-only approaches for validating technology-mediated psychological states. The method is transferable to other interactive contexts where researchers aim to induce specific emotional or cognitive states and need to validate experiential outcomes without relying solely on quantitative self-report measures.

ABSTRACT. Awe is considered a self-transcendent emotion, meaning, an emotion that could promote behavioral change and transformation. Awe is associated with reduced self-absorption, a heightened sense of connection, and a greater tendency toward prosocial behavior (Stellar et al., 2017). Psychological research has shown that awe can be elicited by a variety of stimuli, both natural and artificial, characterized by breadth and symbolic meaning, producing transformative effects on cognitive, emotional, and moral levels (Keltner & Haidt, 2003; Piff et al., 2015). In recent years, virtual reality (VR) has emerged as a particularly effective tool for experimentally inducing awe, thanks to its ability to create immersive experiences and support the emergence of sense of presence and profound involvement, especially when the virtual simulation is designed taking into account users’ intentions (Triberti, Sapone, & Riva, 2025). Secondarily, the experience of awe has been associated with improvement in morality decision-making and prosocial behavior, although the specific mechanisms involved in such a relationship are not completely clear (Scuotto et al., 2024a). This pilot study investigates whether a single immersive VR experience specifically designed to elicit awe, through a narrative-driven journey through space can effectively induce this emotion and influence implicit prosocial attitudes. An experimental protocol was developed that included two conditions: an awe-inducing virtual reality condition and a control condition based on a virtual reality experience tested for relaxation. Twenty-one participants (N = 21) were randomly assigned to the two conditions. Participants completed self-report questionnaires and implicit measures to assess emotional experience, sense of presence, and implicit prosocial attitudes: specifically, they were administered an Implicit Association Task validated for prosociality (Marvel & Resh, 2018) that assessed their tendency to associate self/other representation to either prosocial/egoistic behaviors. Regarding awe elicitation, results indicate that the awe-inducing virtual reality elicited significantly higher perceptions of vastness than the control condition, a major component of awe appraisal (Yaden et al., 2019). Furthermore, the experimental group reported higher levels of curiosity, interest, and surprise, consistent with the emotional profile typically associated with awe (Chirico & Gaggioli, 2021). Importantly, the manipulation check revealed no significant differences between conditions in terms of sense of presence, suggesting that the observed effects were not driven by the immersion factors. Regarding implicit attitudes towards prosociality, the experimental group showed a moderate increase post-intervention (Cohen’s d = .52), although the effect was not statistically significant. Overall, these preliminary results suggest that immersive virtual reality represents a promising tool for eliciting transformative emotions and investigating intuitive moral processes and implicit prosocial tendencies. However, given the small sample size, these findings should be considered exploratory. Further studies with different methodologies and larger samples are needed to assess whether extended reality for complex emotion elicitation could be used within clinical and educational settings in order to improve individuals’ ability to deal with moral issues (Scuotto et al., 2024b).

ABSTRACT. Psychosocial stress is a major determinant of eating behavior, yet its effects on food craving are often assessed using methods with limited ecological validity. Virtual Reality (VR) allows the simulation of realistic decision-making environments in which stress-related behaviors can be examined more accurately. This study explored food-craving patterns following VR-induced psychosocial stress in adolescents, using an immersive VR supermarket paradigm. The objective of the study was to understand food choice in adolescents after social stress. A total of 38 adolescents were recruited from a high school in Iași, Romania, with written informed consent obtained from participants and their parents. Eating behavior and perceived stress responses were assessed using an immersive virtual reality protocol. Participants were exposed to a standardized 10-minute virtual reality protocol based on the Trier Social Stress Test, delivered via an Oculus Quest 2 headset. Stress induction took place in a virtual classroom environment, developed by Cliniques et Développement IN VIRTUO. The environment included a virtual teacher and seven virtual classmates. Participants were instructed to deliver a self-descriptive speech, followed by a mental arithmetic task involving serial subtraction under time pressure. Throughout the tasks, the virtual teacher and classmates displayed controlled, standardized social cues (e.g., neutral expressions, boredom, brief interest, disapproval), used to induce acute social stress. Immediately after stress exposure, participants were immersed in a virtual supermarket, created by the ‘Gheorghe Asachi’ Technical University of Iași, Faculty of Automatic Control and Computer Engineering. Participants were exposed to the environment for 5 minutes to assess stress-related food choice and craving. They explored the supermarket layout and selected food items from predefined categories (high-fat/salty, sweet, and healthy foods), while rating craving and anxiety using visual analog scales. Eating behavior traits were assessed using the Three-Factor Eating Questionnaire for Children (TFEQ-R21C), while perceived stress and emotional responses were evaluated using validated self-report scales. Anthropometric measurements, including body mass index, were obtained to characterize nutritional status. A 2 × 3 mixed-model Repeated Measures ANOVA revealed a strong main effect of food type on post-stress cravings (F(1.93, 435.41) = 168.98, p < 0.001, η²p = 0.43). Sweet foods elicited the highest craving ratings, followed by fatty foods, while healthy foods were least desired. A significant interaction between food type and Uncontrolled Eating was observed (F(1.93, 435.41) = 16.49, p < 0.001, η²p = 0.07), indicating that adolescents with higher UE scores experienced greater cravings for sweet and fatty foods following VR-induced psychosocial stress. These results suggest that VR-induced psychosocial stress selectively alters the profile of food cravings rather than overall appetite. The VR supermarket represents an ecologically valid, technology-mediated environment for assessing stress-related eating tendencies and may support the development of personalized, digitally delivered nutritional interventions.

ABSTRACT. Background. Virtual reality (VR) has become an increasingly prominent tool for the induction of acute stress in laboratory contexts, allowing for strong experimental control while preserving ecological validity. Within cyberpsychology, social-evaluative stress which is characterized by social scrutiny, rejection, or interpersonal threat, represents a particularly relevant domain, as it closely mirrors digitally mediated social interactions. While a growing body of research has employed VR-based paradigms to elicit social-evaluative stress, there is substantial heterogeneity in paradigm design, outcome measurement, and the consideration of individual vulnerability factors. A systematic mapping of this literature is therefore needed to clarify how social-evaluative stress is operationalized in VR and which stress response domains are most consistently engaged. Objective. This scoping review aims to map virtual reality–based social-evaluative stress paradigms, focusing on core paradigm characteristics, stress outcome domains, and individual difference moderators. Methods. A scoping review was conducted in accordance with PRISMA-ScR guidelines. Searches were performed across major scientific databases, including Web of Science, ScienceDirect, SpringerLink, and Google Scholar. Eligible studies employed VR-based paradigms explicitly designed to induce stress and reported at least one stress-related outcome. Following systematic screening and full-text evaluation, 21 studies were deemed eligible for inclusion. Data charting and comparative analysis were conducted to examine paradigm characteristics, stress induction mechanisms, outcome domains, and moderators specifically within social-evaluative VR paradigms. Results. Studies of social-evaluative VR paradigms predominantly employed designs based on avatar-mediated social evaluation, interpersonal threat, and social exclusion. The most frequently used paradigms were virtual adaptations of the Trier Social Stress Test (TSST-VR), involving public speaking and mental arithmetic in front of virtual evaluators, as well as social exclusion tasks such as Cyberball. Across studies, stress induction consistently relied on two core psychosocial components: uncontrollability and social-evaluative threat. Subjective stress and anxiety measures showed consistent and robust increases following VR stress induction. Autonomic markers, including heart rate, electrodermal activity, and heart rate variability, also demonstrated reliable stress reactivity. In contrast, hypothalamic–pituitary–adrenal (HPA) axis activation, indexed by salivary cortisol, appeared more heterogeneous and sensitive to specific design features, such as the realism and interactivity of virtual evaluators. Several studies reported a dissociation between strong subjective and autonomic responses and more variable endocrine responses, indicating differential sensitivity of outcome domains to paradigm characteristics. Multiple studies examined individual difference moderators. Sex differences in autonomic reactivity were reported, and early adverse experiences, including childhood trauma and emotional neglect, emerged as relevant moderators of stress responses and coping behaviors. Additional moderating factors included social support and neurobiological modulation (e.g., intranasal oxytocin), highlighting the sensitivity of these paradigms to vulnerability-related processes. Conclusions. These findings indicate that VR-based social-evaluative stress paradigms provide a flexible and ecologically meaningful framework for investigating psychosocial stress. Beyond confirming robust subjective and autonomic responses, the review highlights systematic differences in how outcome domains respond to specific design features, offering practical guidance for selecting appropriate measures in future studies. In addition, the identification of key paradigm components such as social-evaluative threat and uncontrollability supports more targeted and theory-driven design of VR stress paradigms. By integrating evidence across studies, this review contributes to refining methodological choices and advancing the precision of VR-based stress research within cyberpsychology.

ABSTRACT. Background: Sports injuries are critical events in athletes' careers, significantly impacting not only physical functioning, but also psychological well-being. During rehabilitation, athletes may experience high levels of stress, self-criticism, difficulty regulating emotions, and reduced self-esteem. These factors can hinder their return to competitive sports. In this context, psychological resources such as self-compassion and resilience can serve as protective factors. Meanwhile, a growing body of evidence underscores the significance of the mind-body connection and autonomic regulation as key indicators in stress adaptation processes. Virtual reality (VR) is an innovative tool that integrates psychological and physiological dimensions by offering emotionally engaging, immersive experiences to support rehabilitation. Objective: This pilot study explored the effects of a brief virtual reality-based psychological intervention on psychological variables (e.g., self-compassion, self-esteem, and emotional coping) and cardiac regulation indices in injured athletes. Specifically, the study used subjective self-report measures and physiological indicators of heart rate variability (HRV) to evaluate the intervention's impact over time. Methods: The overall sample consisted of 20 competitive athletes (aged 18-29; M = 24.5, SD = 4.2) in post-injury recovery. All participants followed an identical protocol consisting of four sessions involving exposure to immersive virtual reality scenarios related to the sporting context (one session per week for a total duration of one month) and a follow-up two weeks later. These scenarios were integrated with compassionate self-talk exercises and emotional regulation techniques. The study adopted a within-subjects design with repeated measures. Psychological variables were assessed using validated questionnaires before and after the intervention. Cardiac activity was simultaneously recorded during the sessions using wearable devices and analyzed using heart rate variability (HRV) indices in the time domain, including mean heart rate (HR_mean) and root mean square of successive R-R intervals (RMSSD). Results: Overall, the results showed positive trends for psychological variables, including an increase in self-compassionate behaviors, as well as an improvement in self-esteem and emotional regulation. From a psychophysiological point of view, heart rate variability (HRV) analyses revealed systematic changes in cardiac indices between the baseline and intervention phases. Specifically, RMSSD decreased during the intervention compared to the baseline, suggesting autonomic involvement associated with emotional activation induced by the immersive experience. Meanwhile, the average heart rate varied consistently with the different experimental conditions. Conslusion: Overall, the results suggest that a VR-based intervention could be a promising tool for supporting the psychological recovery of injured athletes by promoting positive subjective changes and measurable physiological responses. However, the exploratory nature of the study and the limited sample size require caution when interpreting the results. The integration of psychological and cardiac measures provides an innovative, multidimensional perspective. These data lay the foundation for future controlled studies on larger samples aimed at clarifying the roles of virtual reality and self-compassion in post-injury recovery processes.

ABSTRACT. Current presence research in virtual reality relies predominantly on declarative assessment and group-level analyses (Grassini & Laumann, 2020), adopting technocentric approaches that obscure individual differences (Triberti et al., 2025) and ignore how users project subjective meaning onto virtual environments. Our findings reveal critical limitations driven by three core problems: methodological reductionism (purely declarative measures miss embodied experience), decontextualization (treating presence as isolated variable), and group-level erasure (averaging obscures individual projective processes).

We therefore propose rethinking presence evaluation through an integrated framework. Drawing on Haddouk's (2017, 2018) symbolic third-space concept and Lewin's (1936) field theory, presence emerges from dynamic person-environment interaction as subjective appropriation, participants invest virtual environments with personal meanings through projection, reinforcing VR "reality" by attributing unconscious contents to immersive scenes. This requires methodological transformation toward MCIL approaches (Multimodal, Contextual, Individual, Longitudinal).

Thirty French young adults (15 women, 15 men, aged 19-30) completed psychometric assessments at baseline (BFI-10, STAI, HADS), followed by three randomized VR sessions (soothing, neutral, anxiety-inducing) at biweekly intervals. Presence (ITC-SOPI), personality, anxiety, and emotional valence/intensity were assessed during and after each immersion; HRV (SDNN, LF-HF) was continuously recorded via EmbracePlus (Empatica). Closure interviews captured post-immersion memory and projective content. A two-level analysis combined group-level correlations and intra-individual trajectories.

Multimodal assessment is necessary because individual presence manifest through psychophysiological patterns invisible to declarative measures alone. Case analyses revealed opposing embodied profiles masked by group averages. Participant D3 (low trait anxiety, low neuroticism) exhibited concordance: highest presence in soothing session (4.0) with decreased heart rate variability when anxious (SDNN: 52ms). Conversely, D4 (high trait anxiety, high neuroticism) showed discordance: highest overall presence in soothing (3.63) yet paradoxically greater engagement (4.53) and spatial presence (3.36) during anxiety-inducing immersion with increased HRV (SDNN: 121ms) despite elevated anxiety. Among concordant participants, neuroticism correlated negatively with HRV (r = -.80, p < .05) while extraversion showed positive associations (r = .96, p < .01), revealing distinct psychophysiological styles that purely declarative assessment cannot capture may have an impact on presence.

Contextual assessment is essential because presence emerges from situated person-environment interaction. While group analyses confirmed contextual effects (presence higher in soothing versus anxiety-inducing conditions, p = .02; emotional intensity predicted engagement R² = .61, spatial presence R² = .71), individual examination revealed radically different patterns across contexts within the same person. D4's paradoxical profile, highest global presence in soothing yet maximal engagement in anxiety-inducing environment, demonstrates that presence must be studied as anchored in specific environmental interactions.

Individual assessment must capture projective processes through which each person co-constructs experience, as group means erase these dynamics entirely. Closure interviews exposed striking projections: D4 recalled vivid human screams absent from VR scenes, A6 perceived silhouettes despite no avatars being present. These memory distortions reveal presence not as passive response but as active projective attribution, internal psychic contents invested into immersive environments. Moreover, individuals differ fundamentally in which contexts facilitate their presence: D3 experienced maximal presence in soothing environments with physiological alignment, while D4 paradoxically engaged most deeply in anxiety-inducing contexts. Such individual variations in contextual sensitivity disappear in group averages yet constitute fundamental mechanisms of presence emergence.

Longitudinal assessment enables distinguishing stable traits from contextual states. Repeated observations across four sessions revealed D3 consistently exhibited concordance while D4 consistently showed discordance, establishing these as stable psychophysiological styles rather than situational fluctuations.

These findings demonstrate that presence assessment requires moving beyond group-level psychometric approaches toward MCIL frameworks capturing presence as an embodied, projective phenomenon co-constructed through person-technology-context interaction. Future research should extend this framework to augmented reality, where virtual elements embedded in personally meaningful physical spaces may intensify projective processes through richer contextual anchoring.

ABSTRACT. Numerous studies in cyberpsychology focus on the sense of presence, a particular phenomenon observed during use of virtual reality. However, recent technological developments raise new questions about the sense of presence, such as the use of augmented reality. In the scientific literature, the definition of the sense of presence is still not universally agreed upon (Schuemie et al., 2001). As a result, the tools used to assess this feeling vary, particularly with regard to the dimensions considered and observed. The assessment of presence in VR has long relied on standardized questionnaires such as the Immersive Tendencies Questionnaire (Witmer and Singer, 1998) or the ITC Sense of Presence Inventory (Lessiter et al., 2001), as well as related scales. However, their extensive use has also revealed significant methodological shortcomings that deserve to be re-examined (Grassini & Laumann, 2020). Recent work (Triberti and al., 2025) challenge the traditional approach of presence in VR centered on technological factors (such as graphic quality or latency), arguing that it is primarily a psychological phenomenon. Even though presence is a psychological phenomenon, it seems to lie at the interface of the relationship between humans and technology. This is why assessing presence in virtual reality and augmented reality can hardly be done with the same questionnaire, as the user experience is fundamentally different, particularly in terms of immersion. It should be noted that few questionnaires have already been developed within the framework of game-based research, to assess immersion in augmented reality (Georgiou and Kyza, 2016). However, no presence scale is currently validated in psychology for the use of augmented reality. Our study aims to test the qualities of the AR Presence Questionnaire, a scale that we designed to assess the sense of presence in augmented reality. Several dimensions of the sense of presence in AR are assessed in the questionnaire. like: immersion / realism, engagement, social presence/co-presence, embodied presence, narrative. Our sample consists of 100 participants, aged 18 to 30. We included psychologically and physically healthy subjects in the study and we tried to include 50% men and 50% women. After inclusion in our protocol, participants were randomly assigned to two groups. The first group (n=50) filled a personality questionnaire (BFI), then participated in an augmented reality immersion, during which they interacted with an avatar visible in the room where the experiment took place. They spoke with this avatar for a few minutes about their personal well-being. After the immersion, they filled a cybersickness questionnaire (CSQ-VR), the AR Presence Questionnaire, and an acceptability questionnaire (SUS). The same tasks were repeated two weeks after. In the second group (n=50), participants filled a personality questionnaire (BFI), were taken to an environment of their choice (several options were available) where they spoke with the same avatar for a few minutes about their personal well-being. After the immersion, they filled a cybersickness questionnaire (CSQ-VR), the AR Presence Questionnaire, and an acceptability questionnaire (SUS). The same tasks were repeated two weeks after. Our data collection is still in progress, but the expected results aim to validate the statistical reliability and the temporal stability of the AR Presence Questionnaire. To this end, internal consistency tests and test-retest correlation will be applied to the results, depending on the different dimensions assessed. These exploratory results will also allow us to observe any possible correlations between the personality scores, acceptability and cybersickness with the AR presence scores. These results will allow us to consider future research, in order to use the AR presence questionnaire in other protocols and to integrate complementary psychometric assessments and physiological measures such as eye-tracking and heart rate variability (Morel et al., 2025).

ABSTRACT. From a pragmatic perspective, human–AI dialogue can be understood as a cooperative interaction governed by conversational norms such as relevance and clarity (Grice, 1975; Clark, 1996). Prior studies highlight that conversational style—including tone, structural organization, and repair strategies—plays a key role in shaping perceptions of agency, trust, and competence in AI systems (Luger & Sellen, 2016; Guzman & Lewis, 2020). At the same time, conversational fluency may create an illusion of understanding, potentially obscuring reasoning errors or epistemic limitations (Bender et al., 2021). By integrating pragmatic theory with fine-grained conversational analysis and multimodal interaction analysis, the approach we are introducing demonstrates that conversational design choices shape perceived agency, trust, and competence, which are fundamental components for the sense of 'being there' or 'being with' the other, independently of objective reasoning performance. This perspective moves beyond traditional conceptualizations of presence and social presence, which have primarily focused on user experiences in virtual environments. The notion of 'enunciative co-presence' is here proposed as a more nuanced framework for understanding the reciprocal awareness and mutual acknowledgment manifested through the exchange of communicative acts in human-AI dialogue. This conceptual shift signals the need to go beyond established presence paradigms and recognize human-AI interaction as a novel challenge requiring innovative methodological approaches. This interdisciplinary perspective (drawing on psychology, pragmatics, and computer science) suggests that human-AI interaction represents a new frontier for presence research, calling for a reconceptualization of the construct to account for the embodied, contextual, and dynamic nature of presence in evolving technological landscapes. Such an approach may contribute to a more comprehensive understanding of the sense of presence in technologically-mediated contexts.

ABSTRACT. Background

The accelerated integration of technology into everyday life shapes not only how people access information but also how they emotionally interpret and evaluate societal innovations and environmental solutions. Bioplastics have been framed as a sustainable alternative to conventional plastics, but public perceptions remain uneven and emotionally complex. Given the central role of digital communication in forming and expressing societal views, examining emotional responses to bioplastics in online spaces contributes to understanding human behaviour in the context of socio-technical transformations.

Objective

This study explores public sentiment toward bioplastics by analyzing a 10-year period of online discussions on Reddit (2014–2024). Through this approach, it aims to identify dominant emotions and examine temporal variations associated with major environmental and policy-related milestones.

Methods

The data collection used Python-based web scraping via the Reddit API (PRAW). Data was retrieved using 32 predefined keywords (obtained through a benchmark list validated by team members), such as biomaterial and polylactic acid. To ensure a robust approach, a rigorous preprocessing and cleaning procedure was implemented, extracting a total of 11,738 comments and then removing 6,697 duplicates, resulting in a final dataset of 5,041 comments. Textual processing was used through the Natural Language Toolkit (NLTK) by removing non-alphabetical characters and stopwords. Sentiment analysis was conducted using NRCLex, mapping affective vocabulary to Plutchik’s eight basic emotions. Emotion scores were normalized by word count to ensure comparability across comments with different lengths.

Results

Public emotional responses toward bioplastics were predominantly positive, with trust, anticipation, and joy consistently exhibiting the highest scores. Although negative emotions (fear, sadness, and anger) were less prevalent, they showed a significant increase in 2018. An additional temporal analysis revealed that peaks in comment volume and sentiment level frequently co-occurred with major milestones, such as the Paris Agreement, the EU Plastic Strategy, and the EU Single-Use Plastics Directive.

Discussion

These findings suggest a theoretical correlation between significant environmental milestones and public sentiment changes, highlighting how global policy announcements and media coverage could potentially influence public engagement. However, heightened public awareness (e.g., post volumes) does not necessarily translate into an increase in sustainable behavior. These results further highlight persistent misconceptions and the need for clearer and more trustworthy information to mitigate public skepticism and uncertainty. Furthermore, several methodological limitations must be acknowledged. The exclusive use of Reddit as a data source limits the representativeness. Additionally, the lack of socio-demographic control (e.g., age, gender, country) and the potential emotional amplification within digital debate forums may influence the generalizability of the results.

Conclusions

This study underscores the importance of real-time social media monitoring for identifying consumer concerns and informing industry communication strategies. To facilitate the transition to an environmentally friendly economy, stakeholders should adopt transparent, emotionally aligned communication strategies to build trust and prevent material misconceptions among end-users. Finally, these results support that close collaboration between companies and consumers is vital for designing sustainable solutions that are directly aligned with public needs.

ABSTRACT. In cyberpsychology, nonverbal signals such as heart rate variability (HRV) and eye movement patterns provide measurable indicators of stress during (Lee & Jung, 2020). These physiological indicators can be used to assess virtual reality (VR) immersion, particularly in VR interventions targeting stress (Meshkat et al., 2024). It also seems promising to use them in psychotherapeutic interactions, for example, to monitor in‑session change, client–therapist alliance, promote physiological synchrony, and to inform VR‑based or technology‑mediated treatments (Tschacher et al., 2025). Thus, the use of biometric sensors could allow us to understand psychological phenomena such as stress, emotional valence, attentional engagement, and interpersonal synchrony in new ways during therapeutic interactions, whether remote or in person (Morel et al., 2025). Physiological synchrony in VR has recently been shown to be comparable to face-to-face interactions (Streuber et al., 2026). Underlying physiological and psychological processes still require investigation. The aim of this study is to evaluate whether physiological signals, specifically HRV and eye-movement data, can be used to identify patterns associated with interpersonal engagement, disengagement, reactivity, and potential risk of alliance rupture during remote psychotherapeutic interaction. Thereby offering interpretable measures that may support the assessment of intersubjectivity. This study is exploratory and has limitations related to the small scale of the data collection. Also, we will rely on exploratory interviews rather than real clinical sessions. This is a proof-of-concept study; the findings should be interpreted as more indicative than definitive and provide a foundation for larger-scale, real-world investigations. The expected results from this exploratory study should provide further insight into psychological variables such as stress, engagement, and alliance rupture. The expected complementarity between physiological and self-reported data collected during interviews may improve understanding of the complexity of interpersonal dynamics in therapeutic contexts.

ABSTRACT. Immersive technologies enable the controlled induction of psychosocial stress while maintaining ecological validity, offering valuable tools for cyberpsychology research focused on human–technology interactions. Given that stress reactivity has been associated with maladaptive eating behaviors, including uncontrolled eating, this study also explored whether physiological stress responses relate to individual eating behavior traits. This study investigated cardiac autonomic responses to VR-induced psychosocial stress using a Virtual Reality adaptation of the Trier Social Stress Test (VR-TSST), from a physiopathological perspective. Thirty-eight adolescents (mean age 15.8 ± 0.6 years) were exposed to the VR-TSST while electrocardiographic (ECG) parameters were recorded before and immediately after stress exposure using a wearable device. Heart rate (HR), corrected QT interval (QTc, Fredericia formula), and PQ interval were analyzed as markers of autonomic and cardiac conduction responses. Subjective stress perception was assessed, and baseline eating-behavior traits were measured prior to VR exposure using the Three-Factor Eating Questionnaire (TFEQ-R21C). Exposure to VR-induced psychosocial stress elicited significant autonomic activation, reflected by an increase in HR (mean ΔHR ≈ +8 bpm, p < 0.01) and QTc prolongation (mean ΔQTc ≈ +12 ms, p < 0.01), while PQ intervals remained unchanged. Partial correlation analyses controlling for sex and body mass index showed that stress-related changes in HR and QTc were positively associated with Uncontrolled Eating scores (QTc: r ≈ 0.42, p < 0.01; HR: r ≈ 0.38, p < 0.05). Regression analyses confirmed that ECG responses to stress significantly predicted uncontrolled eating tendencies. These findings indicate that VR-induced psychosocial stress acts as a potent stimulus capable of eliciting measurable physiological responses in adolescents. From a cyberpsychology perspective, the integration of immersive VR environments with wearable biosignal monitoring offers a robust framework for identifying objective markers of stress susceptibility and for advancing the understanding of mind–body interactions in technology-mediated contexts. These findings highlight the potential of VR-based paradigms as digital tools for identifying individuals at risk for stress-related maladaptive behaviors and for developing personalized interventions in technology-mediated environments.

ABSTRACT. Specific phobias (SP) are a prevalent disorder affecting between 3% and 15% of people worldwide each year. The gold standard treatment is in vivo exposure therapy (IVET), which has demonstrated effectiveness in reducing emotional response to the feared stimulus. However, IVET has certain limitations due to the negative perceptions of both patients and therapists, which reduces the search for and correct application of the treatment, in addition to being affected by other factors (safety, environmental control, and stimulus variability). Augmented Reality Exposure Therapy (ARET) emerges as a potential alternative with advantages in usability, adaptability, and logistics that could overcome these limitations. Previous studies have shown the effectiveness of treating arachnophobia using a mobile-based AR intervention. Moreover, evidence shows that Attentional Biases (AB) are a key factor involved in the development and maintenance of SP. In this line, several studies have demonstrated that Attention Bias Modification (ABM) focused interventions with participants who fear arachnids, could help to maximize the effectiveness of exposure therapy. This work presents the study protocol of a research project aimed at exploring the feasibility, usability, and acceptance of an AR-based mobile app for cockroach exposure (PhobiApp). This mobile app will include two main functionalities: One, for conducting the exposure treatment, in which the app will integrate a 3D model of the insect in the user's physical environment, along with other activities (serious games, gamified levels, reinforcement messages, self-instructions, videos, etc.); another focused on foster ABM through training activities based on the dot-probe task mechanics. Additionally, as a secondary aim of the project, the potential efficacy of both functionalities will be explored. The study will be registered on clinicaltrials.gov and will be conducted following the CONSORT (Consolidated Standards of Reporting Trials) guidelines for pilot and feasibility studies. This project was competitively awarded funding by the Generalitat Valenciana under the CIAICO call for Consolidated Research Groups (CIAICO/2024/66) and the application for approval by the Universitat Jaume I ethics committee is in progress. To achieve the stated objectives, two studies will be conducted. First, the Study 1 will focus on Usability and will consist in a pilot study for the usability and ergonomics of the app. It will require 5-10 participants who will evaluate the PhobiApp user experience. The app will be assessed for its positive, simplicity, and user-friendly nature, and the System Usability Scale (SUS) will also be administered. Second, the Study 2 will be a two-arm randomized study with a sample of 40 participants. Participants who meet the inclusion criteria (being over 18 years of age, meeting DSM-5 criteria, having had the phobia for more than one year, and signing informed consent) will be randomly assigned to one of the following two conditions: 1) PhobiApp with exposure treatment functionality; and 2) PhobiApp with exposure functionality plus ABM training functionality. The evaluation protocol will include diagnostic and other relevant clinical measures related to small animal phobias (beliefs, fears, behavioral avoidance), as well as attentional biases (assessed by an eye-tracking specifically design task), adherence to the App, and measures of expectations, satisfaction, and opinion. All these measures will be assessed at pre-treatment, post-treatment, and at follow-up (6 months). We expect that this project will provide feasibility, usability and acceptance data of the FobiApp, as well as preliminary efficacy data related to its potential effectiveness. These results would help to keep improving and maximizing exposure treatments.

ABSTRACT. Positive technology (PT) is a growing interdisciplinary field concerned with the use of technology to promote human wellbeing. Drawing on positive psychology, which emphasises holistic wellbeing, PT adopts an integrated model in which hedonic, eudaimonic, and social needs are viewed as central to wellbeing, and positions technology as a uniquely capable medium for supporting their development. This commitment is reflected in the conceptual distinction between hedonic, eudaimonic, and social technologies, designed to support affective, meaningful, and social engagement respectively (Riva et al., 2012). This position paper traces the evolution of positive psychology and positive technology frameworks, examining how their shared theoretical commitments have been reflected in research outputs to date. Within positive psychology, it is increasingly recognised that eudaimonic concerns, particularly meaning and purpose are underrepresented and methodologically underserved, with outcomes typically framed in hedonic terms. As a result, a gap has been identified between positive psychology’s theoretical commitment to an integrated model of wellbeing and its research and applied outputs, which remain largely centred on subjective wellbeing and other hedonic outcomes. Wong (2011, 2019) has argued that traditional positive psychology approaches, which emphasise positivity as the primary route to wellbeing, are poorly suited to the development of holistic wellbeing. In particular, Wong suggests that eudaimonic development cannot be achieved through an exclusive focus on positive experience, as the pursuit of personal meaning, values, goals, and purpose often requires engagement with existential concerns, trauma, and other challenging aspects of the human condition that involve discomfort or suffering. This paper argues that a similar gap between theoretical ambition and empirical output can be observed within positive technology research. Although PT as a framework is well positioned to address challenging and meaningful subject matter, particularly through immersive technologies such as virtual reality, research outputs have yet to fully reflect this potential. Virtual reality, in particular, offers the capacity to immerse users in challenging, thought-provoking, and meaningful experiences that may be better suited to supporting eudaimonic engagement. This paper examines the extent to which PT research to has realised the vision of hedonic and eudaimonic technologies, identifies key gaps in current approaches, and outlines the theoretical foundation for three experimental studies. These studies investigated differences in psychological responses to hedonic and eudaimonic virtual reality stimuli, with a particular focus on understanding the processes and mechanisms through which engagement with positive technology designs influences psychological outcomes. Critically examining the relative importance and more importantly, effectiveness of hedonic and eudaimonic technology designs on a range of wellbeing outcomes has the potential to provide more tools and insight to designers and practitioners looking to use the PT framework to support user wellbeing.

ABSTRACT. Research on digital distraction and smartphone use has predominantly relied on self-report questionnaires to assess individuals’ perceived attentional difficulties, problematic use, and emotional dependence on mobile devices. Such instruments provide valuable insight into subjective experiences associated with technology use; however, their relationship with objectively measured cognitive performance remains insufficiently understood. This dissociation is particularly relevant within the framework of attentional fragmentation, where frequent digital multitasking may alter subjective perceptions of focus without necessarily manifesting as deficits in core attentional networks. In parallel, cognitive science offers well-established experimental paradigms for assessing attention through behavioral indices such as reaction times and error rates. Despite the widespread use of both self-report and experimental approaches, relatively little work has examined the extent to which these measurement modalities converge. The present study addresses this methodological gap by investigating whether commonly used self-report measures of digital distraction meaningfully reflect objective indices of attentional performance. Data were collected from a laboratory study originally designed to examine smartphone related phenomena. The present work constitutes a secondary, methodologically focused analysis addressing a distinct research question concerning measurement convergence. The sample comprised 144 young adults aged 18 - 25 years who participated in a controlled experimental setting. Participants completed a battery of validated self-report questionnaires assessing perceived smartphone-related distraction, nomophobia, and problematic smartphone use. Objective attentional performance was measured using the Attention Network Test (ANT), which provides indices of reaction time and accuracy across multiple attentional components. The analysis focused on the relationships between self-assessment measurements and their correspondence with attention behavior indicators. Initially, the interdependencies between the self-assessment tools were examined in order to evaluate the internal consistency of the subjective concepts related to digital distraction. Next, the correlations between self-assessment indicators and objective measures of attention (ANT reaction times and error rates) were evaluated. It is important to note that the analysis is not intended to test experimental results or differences between groups, but to examine the pattern and strength of correlations between subjective and objective measurement methods. Results reveal a consistent and interpretable pattern. Self-report measures of digital distraction, nomophobia, and problematic smartphone use showed strong intercorrelations, suggesting a consistent subjective experience of technology-related attentional difficulties. In contrast, associations between these self-report indices and objective attentional performance were weak and inconsistent. Correlations with ANT reaction times were minimal, across all three attentional networks (alerting, orienting, and executive control), suggesting little correspondence between perceived distraction and processing speed. Correlations with error rates were slightly stronger but remained small and heterogeneous across measures. Overall, subjective reports of digital distraction did not reliably correspond to objective measures of attentional efficiency as recorded by a standard experimental task. These findings have important methodological implications for research on attention in digital environments. They suggest that self-report instruments may primarily capture experiential, affective, or metacognitive aspects of perceived distraction rather than core attentional performance. At the same time, brief laboratory-based attention tasks like the Ant may be relatively insensitive to the multidimensional and socially-cued interruptions characteristic of everyday smartphone use . The observed dissociation underscores the need for caution when interpreting self-reported digital distraction as a proxy for objective attentional decline in controlled settings. By foregrounding the convergence - and lack thereof - between subjective and objective measures, this study contributes to ongoing discussions about construct validity and measurement in cognitive science. The results highlight the importance of multi-method approaches and suggest that a more comprehensive understanding of attention in digital contexts may require the integration of self-report, behavioral, and ecologically valid assessment methods.

ABSTRACT. The rapidly increasing integration of digital technologies into everyday life has reshaped the conditions under which attention develops, is regulated, and is disrupted. University students, in particular, engage in learning and daily activities in environments characterized by constant connectivity, multiple digital platforms, and increased cognitive demands. Although previous research has extensively examined individual aspects of technology use, such as smartphone use, social media participation, or multitasking, such concepts are often studied in isolation, limiting our understanding of how attention regulation works in the broader digital environment of everyday life. This study is the first phase of an ongoing doctoral-level longitudinal research project that aims to examine how students regulate attention, cognitive functioning, and behavior in high-tech environments. Rather than focusing on a single device or application, the research takes a whole-picture approach to everyday technology use, including multiple forms of digital engagement and their interaction within a continuous multitasking environment. The participants are first-year university students who were recruited from face-to-face courses. Data collection was carried out using printed questionnaires, which were completed in multiple sessions within the classroom, with the aim of reducing fatigue. Participation was anonymous, and responses were linked between sessions using a self-generated code according to instructions, which participants were asked to remember for the following years of participation in the study. In total, fourteen validated self-report questionnaires were administered, covering a wide range of concepts related to attention, self-regulation, and daily technology use. The tools assessed, among other things, multitasking behaviors, daily attention disruptions, distraction due to digital devices, academic motivation, metacognitive beliefs, cognitive flexibility, self-control, problematic forms of technology use, nomophobia, fear of losing online content, goals for technology use during lectures, and digital skills. All scales were culturally adapted according to international guidelines, and their internal consistency (Cronbach’s α) was examined at the subscale and scale levels before calculating the scores. This first phase is designed as a mapping and profiling stage, providing a comprehensive baseline of individual differences prior to further experimental and longitudinal assessments of the project. To identify distinct patterns of attention regulation, a person-centered analytical strategy was employed using K-means Cluster Analysis on the standardized scores of the measured constructs. Preliminary findings from the first phase (N=233) identified five distinct student profiles, ranging from "Digitally Resilient" (high self-control, low distraction) to "High Digital Vulnerability" (high nomophobia, frequent multitasking, and low attentional control). By placing attention regulation within the broader context of everyday technology use, this study seeks to advance the current understanding of how individuals adapt to constant digital demands. The profiling approach of the first phase lays the empirical foundation for the subsequent phases of the doctoral program, which will incorporate experimental tasks and longitudinal follow-ups to examine the stability, cognitive associations and functional consequences of the identified profiles over time.

ABSTRACT. The body, both one’s own and that of others, plays a crucial role in shaping how individuals relate to one another and to the environment, in both physical and digital contexts. Based on Embodied Social Presence Theory, this randomized between-subjects experimental study aims to investigate how different body representations and embodied cues of virtual agents can impact perceived social presence, as well as self- and other co-presence. Eighty participants are randomly assigned to complete a computer-mediated dialogical task with a virtual agent in one of four conditions: interacting with a disembodied voice in an empty virtual room; a voice associated with an empty virtual chair; a virtual spherical agent displaying eyes and a mouth; or a virtual 3D female agent. The virtual environment, voice, and task remain constant across conditions. The conversational task involves a reciprocal exchange in which the virtual agent asks participants about their personal interests and preferences, and both parties share their viewpoints. The study integrates self-report and psychophysiological measures, by also collecting eye-tracking data during the conversations, to identify gaze patterns associated with social presence. Participants first provide their demographic information and are then randomly assigned to one of the dialogical conditions. Immediately after the interaction, they complete a survey to assess their experience. This includes the Social Presence Questionnaire (Herrera et al., 2020), the Copresence Scale (Nowak & Biocca, 2003), the Godspeed Questionnaire Series (Bartneck, 2023), the Inclusion of the Others in the Self Scale (Aron et al., 1992), and the Uncanny Valley Effect Scale (Ho & MacDorman, 2017). Data are collected at a single time point. Eye-tracking patterns are recorded using the Web-Agile Facial Emotion Recognition and Eye-Tracking System (WAFER-ET; Gursesli et al., 2023). At the time of submitting the abstract, data collection is nearly completed and statistical analyses are forthcoming. Based on previous literature, the main hypothesis predicts that participants will report higher social presence and co-presence in experimental conditions characterized by a greater degree of embodied cues from the agents (i.e., the human-like agent), due to the potential activation of cognitive models underlying human social interactions. This is provided that the uncanny valley effect does not emerge. However, if such an effect occurs, leading participants to perceive the most anthropomorphic agent as uncomfortable, a complementary hypothesis anticipates higher ratings in less embodied conditions, such as interaction with the spherical agent, where reduced anthropomorphic cues may elicit fewer social interaction expectations. The perceived intelligence, likeability and emotionality of the agents are also measured to test their role as mediators in these hypotheses. Additionally, the study expects to detect specific eye movement patterns that could predict social presence scores. Overall, the study aims to explore the modeling of social presence through the integration of self-report and psychophysiological data. Moreover, the research seeks to deepen the understanding of how bodily cues may shape social presence in human-agent interaction, while offering design-relevant insights into the use of embodied features in virtual characters. The findings could be particularly relevant in light of current developments in embodied artificial intelligence.

ABSTRACT. As part of a broader research initiative investigating how Recommender Systems encode human psychological constructs, this work in progress introduces an expert-validated dataset. By pairing individual personality traits with specific persuasion principles, this research enables the development and evaluation of more transparent, psychology-aware Recommender Systems.

Currently, research in the field of Recommender Systems is primarily focused on the optimization of model performance and predictive accuracy, as the underlying reasons for model behavior and the specific impact on the end user are rarely prioritized. This dataset is being developed to address this gap by providing a controlled environment where the influence of human psychology can be directly measured. We are thus shifting from performance-based evaluation to providing a framework for assessing the ethical footprint of recommender systems.

The primary objective of this study is to develop and validate a dataset that maps Five Factor Model personality traits to Cialdini’s principles of persuasion, providing a benchmark to audit the behavioral intent and ethical impact of AI-driven recommendations.

The experimental framework is structured into three sequential phases:

- Phase I: Stimuli Generation and Psychological Grounding. Advertisements for mundane, high-utility household goods were synthesized using a Generative AI pipeline. For each product, six variations were engineered to systematically isolate one Principle of Persuasion (e.g., Social Proof, Scarcity, Authority), ensuring that user responses are a direct reaction to the persuasive frame, rather than the product's utility or external brand influence. This corpus was then subjected to a double-blind annotation process by expert psychologists to verify the active persuasive "hook" establishing the ground-truth persuasion-embeddings.

- Phase II: Data Collection. A large-scale user study (N=300) is conducted where participants are exposed to a randomized selection of the stimuli. Additionally, personality factors were measured using the Ten-Item Personality Inventory (TIPI), which created a dense interaction matrix where user traits (Five Factor Model) are mapped directly to behavioral responses to the persuasive content.

- Phase III: Statistical Analysis and Behavioral Mapping. The dataset is validated by mapping the recorded interactions against established psychological theory. Successfully mirroring these relationships will serve as the primary validation of the dataset's quality, transforming the raw interaction data into a verified benchmark to audit how recommender systems process human traits.

The engineering and expert validation of the advertisement stimuli have been successfully completed, providing a scientifically grounded foundation for the interaction dataset. Recruitment for the behavioral data collection study, involving N=300 participants, has been initiated. Preliminary results from the expert review process indicate that high-fidelity persuasive content can be effectively synthesized while maintaining strict alignment with psychological frameworks.

Ultimately, this dataset will serve as a tool for the community to audit the "why" behind algorithmic recommendations. By providing a clear map of how traits like the Big Five interact with persuasive hooks, this work creates a path for designing systems that are psychologically transparent and ethically sound.

AI Disclosure: Generative AI was utilized for the synthesis of advertisement stimuli and to assist in the linguistic refinement of this abstract. All psychological annotations and experimental designs were developed by the authors.

ABSTRACT. Background. Digital reading has become a default mode for academic and leisure texts, yet empirical work often reports a “print advantage” in comprehension and memory. A persistent challenge is that “print vs. digital” comparisons frequently bundle together multiple factors: screen luminance, scrolling vs. paging, distraction potential, and—crucially—differences in embodied interaction. From an embodied-cognition perspective, physical handling of a text (turning pages, holding the object, micro-adjusting posture and grip) may support attention regulation and the construction of a coherent situation model by providing stable spatial-temporal cues and sensorimotor feedback. In tablet reading, however, these cues can vary substantially depending on whether the device is held and actively manipulated or positioned on a stand with limited tactile engagement. In parallel, individual differences may shape digital reading outcomes: habitual social media use has been discussed as a potential risk factor for sustained attention during longer-form texts, and readers’ medium preferences may influence engagement and motivation—both of which are known to support deeper comprehension and richer mental representation. The present poster integrates coordinated analyses from a shared experimental dataset to disentangle (i) medium effects from (ii) embodied interaction constraints, while also examining the roles of (iii) social media addiction symptoms and (iv) preference–medium match in narrative engagement and spatial situation modeling. Methods. University students read the same 19-page literary excerpt (H. Hesse, Steppenwolf) under between-subjects random assignment to: print book (n=51), iPad handheld (n=51), or iPad on a stand with minimal touch interaction (n=51). After reading, participants completed a 14-item comprehension test (true/false; 0–14) without access to the text. A subsample comparing print vs iPad handheld (N=102) additionally completed the Bergen Social Media Addiction Scale (BSMAS) and measures of spatial representation (total and subscales) and narrative engagement, including a preference–condition match indicator. Results. Comprehension differed by reading condition (Welch’s ANOVA: F(2, 98.7)=6.41, p=.002). Performance was lowest in the iPad-on-stand condition (M=7.06, SD=1.91), compared with both print (M=8.27, SD=2.57; post-hoc p=.02) and iPad handheld (M=8.35, SD=2.11; post-hoc p=.004), while print vs iPad handheld did not differ (p=.98). In the print vs iPad handheld subsample, comprehension again did not differ (t(100)=0.169, p=.567, d=0.03), and BSMAS scores were not associated with comprehension (r(100)=0.053, p=.703). Spatial representation and narrative engagement showed no significant main effects of medium (spatial: t(100)=−1.03, p=.305; engagement: t(100)=−1.61, p=.111). However, in print, spatial representation correlated with engagement (total: r=.31, p=.027; narrative subscale: r=.34, p=.013), whereas this coupling was not observed in the digital condition. Preference mattered: participants who read in their preferred medium reported higher engagement (t(94)=2.33, p=.022, d=0.47). Conclusions. Overall, the pattern of results indicates that comprehension differences often attributed to “digital reading” may be better understood as consequences of how digital reading is physically enacted. When tablet reading preserved active, handheld interaction, performance was comparable to print, whereas restricting tactile manipulation (tablet positioned on a stand, minimal touch) was associated with poorer comprehension. This supports the view that sensorimotor involvement and stable spatial cues can scaffold narrative processing in longer texts. Social media addiction symptoms were not reliably linked to comprehension in this sample. Preference–medium match predicted higher narrative engagement, suggesting motivational alignment can enhance immersion even without comprehension gains. Practically, findings support e-reading setups and interfaces that preserve tactile control and spatial landmarks.

ABSTRACT. Introduction: Digital mental health interventions have expanded rapidly and offer considerable potential to promote well-being and resilience in vulnerable populations, such as migrant communities. However, a persistent challenge in digital interventions is sustaining user engagement over time, particularly among groups facing social, economic, and structural barriers to accessing care (Dobalian & Rivers, 2008; Jorm, 2012)​. Low adherence and early dropout remain common, limiting the effectiveness of otherwise evidence-based interventions. Addressing engagement is therefore a central design and implementation challenge rather than a secondary outcome. Recent progress in artificial intelligence offers new opportunities to support engagement in digital mental health interventions through adaptive, personalised delivery ​(Aschentrup et al., 2024). Machine learning techniques can be used to tailor content, format or the timing to users’ changing needs and contextual states, potentially reducing burden and increasing perceived relevance. However, technological solutions alone are insufficient if interventions are not grounded in the lived experiences, cultural contexts or the preferences of the populations they aim to serve (Bernal et al., 2009). In response to these challenges, the European project Reconnected integrates participatory co-creation with artificial intelligence-driven personalisation to develop and deliver digital microinterventions for Latin American migrants living in Spain. By combining stakeholder involvement in content development with adaptive algorithms that support engagement, the project aims to enhance both the accessibility and sustained use of digital mental health resources within a scalable platform. Objective: The objectives of this work are: (1) to describe the co-creation process used to develop and adapt microintervention content for Latin American migrants living in Spain; and (2) to present artificial intelligence aimed at fostering engagement and personalising the delivery of digital microinterventions. Method: Four co-creation workshops of approximately 90 minutes were conducted with end users from the target population and professionals working with migrant communities. During the workshops, participants were presented with microintervention content in different formats, including text, images, and videos. Structured and open-ended questions were used to elicit feedback on clarity, relevance, cultural appropriateness, and perceived usefulness of the materials, as well as suggestions for improvement. Qualitative material collected during the sessions is currently undergoing analysis to inform adaptations to language use and cultural relevance of the microinterventions. In addition, a machine learning based system is being developed to personalise the delivery of microinterventions. The system is designed to adapt delivery format (text, audio, or video), recommend grounding-focused microinterventions under predicted stress conditions, and deliver automated motivational messages. Motivational messages may be informative, affirming, or motivating, with message type determined algorithmically. A large language model is used to adapt the tone and content of messages based on target group characteristics (i.e., age or gender). Results: Qualitative material related to cultural adaptation of the microinterventions was collected across four co-creation workshops and is currently undergoing qualitative analysis. The digital support system including adaptive delivery functionalities has been implemented within the Moodbuster 2.0 platform. The system includes personalised selection of delivery format, stress-informed prioritisation of grounding-focused microinterventions as well as motivational messages. Moodbuster 2.0 is available as both a web-based and a smartphone application (Kleiboer et al., 2016; van de Ven et al., 2017; Warmerdam et al., 2012). Conclusions: Integrating stakeholder co-creation with artificial intelligence driven personalisation represents a promising strategy for addressing engagement challenges in digital mental health interventions for migrant populations. The combination of culturally adapted content and adaptive delivery mechanisms within a scalable digital platform has the potential to support the promotion of well-being and resilience, while accounting for individual needs and structural barriers to accessing mental health resources. This approach may also facilitate implementation across diverse settings and populations.

ABSTRACT. In the pursuit of improved quality of life and socioeconomic opportunities, migration has increased in recent years across the European Union (Esteves & Rauhut, 2023; European Commission, 2025). Given the complex psychological challenges associated with migration, it is essential to develop adequate tools to support the acculturation process (Cervantes et al., 2010). There is currently limited research on extended reality (XR) applications concerning the experience of migration that are not focused on specific refugee or national groups. This becomes problematic when one considers the limited generalisability of such interventions to cross-national contexts, targeting the wider migration experience. This study aims to develop an XR tool for clinicians and therapists to integrate in their practice, to support immigrants’ adaptation to their new circumstances. XR was chosen as the main technology due to its immersive nature, allowing for the elicitation of emotions that migrants would experience in their host countries, such as loneliness or homesickness (Rowan et al., 2020). Focus groups will be conducted with clinicians to understand the psychological challenges that this population encounters. 15-20 participants, selected via convenience sampling, will engage in semi-structured interviews. The meetings will be recorded and transcribed, with the resulting data going through thematic analysis (Braun & Clarke, 2006). The objective of these conversations is to inform the development of the XR tool, which will then be validated within a clinician sample, followed by a case study implementation to determine its utility and effectiveness from a user perspective. The development and validation of such a tool will contribute to the personalisation of interventions, thus allowing for improved individual support of immigrants and subsequently greater integration in their host environments.

ABSTRACT. As digital health technologies become increasingly integrated into everyday settings, there is growing demand for neurotechnological tools that are not only accurate but also accessible. In mental health applications, this places particular emphasis on minimizing system complexity without compromising clinical relevance. EEG-based biomarkers for major depressive disorder (MDD) hold promise as objective, behaviorally relevant indicators of cognitive and affective dysfunction. However, their translation into clinical and real-world practice remains limited. A key barrier is the continued reliance on high-density EEG systems, which are costly and time-consuming to set up. While ultra-low-channel EEG systems (e.g., 2–3 electrodes) offer a more practical alternative, they are typically built around fixed, convention-driven electrode placements, with limited empirical validation regarding whether these locations are optimal for classification. This raises a fundamental question for applied EEG research: how minimal can an EEG system be while still providing reliable and meaningful information about mental health? Addressing this question requires moving beyond tradition-driven design toward data-driven, evidence-based optimization of electrode count and placement. Here, we show quantitatively that simple ERP indices extracted from the dot-locked P300 (0.3–0.6 s) during an emotional attention dot-probe task are sufficient for competitive MDD/healthy control (HC) classification, and that using all 128 EEG channels is not only unnecessary but can be detrimental. Instead, a small set of 3 electrodes selected via a data-driven optimization strategy can preserve or even improve classification performance relative to high-density approaches. We analyzed high-density 128-channel EEG and behavioral data from the publicly available MODMA dataset (Cai et al., 2022; n = 53; MDD = 24, HC = 29). Within the P300 window, we computed channel-wise P300 peak latency, Hjorth mobility, and positive area (Li et al., 2018) and performed MDD/HC classification using logistic regression with leave-one-out cross-validation (LOOCV). To identify optimal electrode configurations for a given channel count k, we evaluated up to 50,000 random channel subsets drawn from the full 128-channel set or exhaustively assessed all possible combinations when feasible. Channels were ranked based on their frequency of appearance in top-performing subsets. Using all 128 channels resulted in near-chance performance (52.8% accuracy, AUC = 0.527). In contrast, three optimally selected electrodes (E124, E109, E95) achieved 75.5% accuracy (AUC = 0.767). Notably, these three electrodes are all located in the right hemisphere. Performance among the top 1% of channel subsets peaked at k = 3 (70.9 ± 1.8%) and declined steadily with increasing channel count (k = 64: 63.0 ± 1.2%; k = 128: 54.7%), indicating diminishing returns and increased overfitting with larger montages. We further compared this data-driven electrode set with a commonly used frontal 3-channel montage (Fp1/Fpz/Fp2), which is often favored for wearable and clinical applications due to ease of placement. Fourteen ERP metrics per channel (Li et al., 2018) were extracted, and all feature-subset combinations were evaluated. Importantly, even when the same optimal feature combinations identified for the frontal montage were applied to our data-driven channel set, the latter consistently achieved higher classification performance. This indicates that electrode placement itself, rather than feature selection alone, is a critical determinant of performance. While conventional frontal electrodes remain attractive for wearable EEG due to their practicality, our results demonstrate that non-frontal, right-lateralized posterior regions can carry complementary and highly informative signals. In summary, increasing sensor density does not necessarily improve EEG-based MDD biomarkers. Reliable discrimination between individuals with MDD and healthy controls can be achieved with as few as 3 strategically selected electrodes, and data-driven optimization of electrode placement provides a systematic approach toward efficient and scalable low-channel EEG systems.

ABSTRACT. Introduction The rapid ageing of the population is associated with a growing burden of cognitive decline and dementia, highlighting the need for accessible tools for early detection and longitudinal monitoring. Eye-tracking (ET) has emerged as a promising method to capture real-time markers of attention and cognitive processing. However, its clinical adoption remains limited due to the high cost, technical complexity, and limited scalability of traditional ET hardware. Recent advances in webcam-based eye-tracking algorithms offer a low-cost and widely deployable alternative, but their feasibility and clinical relevance in older populations are still underexplored.

Objectives The primary aim of this study is to evaluate the feasibility of integrating an affordable, webcam-based eye-tracking system into digital cognitive assessment in an older population. Specifically, the study seeks to examine the association between eye movement patterns extracted from standard laptop webcams and overall neuropsychological performance. Secondary objectives include comparing ocular metrics between cognitively healthy older adults and individuals with Mild Cognitive Impairment (MCI), in order to assess the discriminatory potential of webcam-based eye-tracking measures. Exploratory objectives focus on investigating relationships between specific eye-tracking parameters and distinct cognitive domains, such as attention, executive function, and memory, as well as examining longitudinal changes in cognitive performance and associated ocular markers over time.

Methods This study employs an observational cross-sectional design with a longitudinal follow-up component. Participants aged 65 years and older will be recruited from a tertiary geriatric center. The study population will include cognitively healthy controls, individuals with subjective cognitive decline, patients diagnosed with MCI, and individuals with mild dementia, allowing for a broad representation of cognitive functioning in later life. All participants will undergo a standardized digital neuropsychological assessment administered on a laptop computer. During the assessment, eye movements and facial features will be registered using a software-based eye-tracking application that relies exclusively on the device’s built-in webcam, without the use of dedicated or high-cost eye-tracking hardware. The system will extract multiple ocular metrics including eye movement amplitude, movement velocity, variability in amplitude and velocity, and the absence of long-fast eye movements. Key methodological considerations include potential variability in webcam quality across devices and susceptibility to noise from head movement and environmental conditions, which will be monitored and accounted for in analyses. Planned statistical analyses include descriptive statistics to characterize the sample, group comparisons to identify differences between cognitive status groups, correlation analyses to explore associations between eye-tracking metrics and cognitive scores, and generalized linear mixed models.

Expected Outcomes This study is expected to demonstrate the feasibility and potential clinical value of low-cost, webcam-based eye tracking as a scalable and accessible tool for cognitive assessment in older adults. By reducing technological, financial, and logistical barriers, this approach may support broader implementation of eye-tracking–enhanced cognitive screening in routine clinical practice and facilitate longitudinal monitoring of cognitive trajectories. Ultimately, such tools could contribute to earlier detection of cognitive decline and more personalized, data-driven approaches to geriatric cognitive care.

ABSTRACT. Harmful behaviors in gaming communities and social media platforms compromise user wellbeing and weaken psychological safety. Most moderation systems focus on responding through punitive actions such as muting or banning. These reactive approaches prioritize enforcement over behavior change and do little to address the emotional and regulatory processes that sustain harmful interaction patterns. These limitations have prompted calls for psychologically informed tools that move beyond punishment toward mechanisms supporting proactive and adaptive behavior change. This poster presents validation of an AI-powered system designed to evaluate emotional tone in online text. The system scores thirteen attributes grounded in psychological and communication theory, including empathy, encouragement, sarcasm, hostility, and social judgment. These attributes are organized into three interpretable categories—Friendly, Tense, and Toxic—which together represent community-level emotional tone. The framework draws on established research in affective computing while emphasizing constructs with direct relevance to online community dynamics. We evaluated alignment between automated assessments and trained human judgment using publicly available YouTube comments. Two independent reviewers selected two videos as stimulus sources based on contrasting interaction climates. In both cases, the content creator established the dominant tone. One video represented a predominantly negative context in which the creator engaged in sarcastic, dismissive, and personally demeaning communication that shaped subsequent audience responses. The second video represented a predominantly positive context in which the creator modeled humor, warmth, and playful engagement that fostered supportive exchanges among commenters. Both videos focused on casual lifestyle content to reduce topic-related variation in emotional tone. The sampling frame consisted of top-level comments visible at the time of extraction. The analysis included only primary comments authored in direct response to the video and excluded reply threads to maintain independence across observations. After data cleaning, the negative-context video contributed 110 comments and the positive-context video contributed 101 comments, yielding a total corpus of 211 unique interactions. Four raters with graduate training in psychology independently coded each comment after completing a structured calibration session using shared operational definitions for all thirteen attributes. Inter-rater agreement reached strong levels, with a mean intraclass correlation coefficient of .81 and a range from .74 to .91. Aggregated human ratings served as the benchmark for evaluating automated performance. The AI system analyzed identical comments using the same attribute framework, and raters remained blind to automated outputs throughout. Results demonstrated strong correspondence between human and automated assessments. In the negative interaction context, correlation between AI scores and aggregated human ratings was r = .86, p < .001. In the positive interaction context, correlation reached r = .97, p < .001. Joint analysis across all 211 comments yielded r = .87, p < .001, indicating large effects and stable alignment regardless of interaction climate. Attribute-level analyses revealed higher agreement for constructs with clear lexical signals, including hostility, sarcasm, and encouragement, and greater variability for constructs requiring contextual inference, such as empathy and social judgment. Comment-level analyses showed consistent sensitivity with no evidence of systematic bias. Several limitations merit consideration. The dataset consisted of comments drawn from two videos, which limits the representativeness of interaction patterns across broader online contexts and content genres. The sample size, while sufficient for validation and correlation analyses, remains modest and restricts generalization to larger and more heterogeneous populations. In addition, the use of binary or threshold-based attribute coding may oversimplify complex emotional expressions. Therefore, replication using larger samples, diverse platforms, and alternative coding schemes is necessary. The findings show that a psychologically grounded AI system closely aligns with human judgment in assessing emotional tone across online interactions, while human interpretation remains essential for contextual nuance.

ABSTRACT. Public skepticism towards U.S. initiatives over seas shows the limits of generic outreach strategies. Many platforms do not consider differences in personalities, backgrounds, and financial interests. This project explores how artificial intelligence driven personalization is able to create more effective and trustworthy user experiences. A web platform was developed using Python Flask as the framework with Jinja2 templates and a SQLite for the database. When users sign up they complete a short 15-question personality assessment based on the OCEAN psychological model. This identifies dominant personality traits and investment preferences. Using this information, along with basic demographic data, content is tailored across the platform. AI language models, like Llama 3, generate personalized responses. This is done through prompt engineering rather than model retraining. This allows custom messaging tones and custom content to align with individual user profiles while remaining scalable. The platform also integrates AI generated images using the Replicate API (Flux Schnell model). Users see a sequence of personalized images that slowly shifts from culturally neutral visuals to positive cultural themes. These images are paired with simple explanations of investing concepts. The user interface was redesigned with a clean layout, smooth animations, and a persistent chatbot available on every page to improve usability and trust. Users are also given the opportunity to support causes benefiting foreign communities through verified organizations. In conclusion, this project shows that, with the help of AI, personalization may help increase engagement and trust, by using personality analysis, language generation, and personalized visuals for every page, making the website more usable and trusted.

ABSTRACT. Every voice is a fingerprint. It carries frequency and timbre, the physiological trace of a body that has lived. It is also, increasingly, a dataset: collected, labeled, aggregated, and fed into systems that learn from it, produce new voices, and make decisions about the people whose speech provided the raw material. This study asks what happens when a voice becomes data: what is gained, what is lost, who benefits, and who pays.

The study draws on 14 semi-structured expert interviews conducted between February and March 2026, which were analyzed using Reflexive Thematic Analysis (Braun & Clarke, 2006, 2021). Participants were purposively selected to bring together three communities of knowledge that rarely engage directly: policy and legal researchers, empirical and field researchers (including NLP scientists and investigative journalists), and digital artists working with voice and sound. They span nine countries and seven mother tongues; professional contexts include EU regulatory bodies, open-source NLP development in East African language communities, humanitarian data practice, and sound art. A supplementary quantitative layer — VADER sentiment scoring and K-Means linguistic clustering applied to interview transcripts — was used diagnostically to surface discourse texture and test whether professional clusters correspond to linguistic patterns. Quantitative discourse clusters do not reproduce professional segmentation, confirming that what people say and how they say it are independent analytical dimensions.

Eleven themes emerged across four groupings. The first traces a fifty-year genealogy of voice datafication and examines how the somatic cascade — the physiological sequence through which voice acts on the body before the mind interprets it — renders rational-choice consent frameworks structurally inadequate for technologies that act before deliberation begins. The second document surveillance mechanisms and structural asymmetries along lines of language, gender, and postcolonial geography, including accent modification software that erases workers' linguistic histories in real time and measurable erosion of politeness toward female-coded voice assistants after sustained use. The third map shows governance failures and identifies a functioning community data license, built collaboratively with East African language communities, as an empirically grounded alternative to state-centered regulation. The fourth addresses institutional silences, the absence of longitudinal evidence on psychological effects of chronic voice surveillance, the ethical and relational dimensions of posthumous voice synthesis, and the emergence of non-human voice rights as a theoretical frontier.

Five cross-cutting findings are reported: voice datafication is a fifty-year infrastructure now operating at normalised scale, not a recent disruption; its consequences are structurally asymmetric along lines of language, gender, and postcolonial geography; existing governance frameworks are inadequate and internally contested, with the disagreement between reformist and structural-critical positions not resolvable through evidence alone; the longitudinal evidence base for psychological effects does not exist, leaving a scientific vacuum that policy discourse fills with undemonstrated claims; and the most disruptive implication of voice AI may not be the surveillance it enables but the forms of expression it makes possible for entities — posthumous, non-human, and otherwise silenced — that have never had one.

Three practical implications follow: synthetic voice must be disclosed in any context that presents itself as authentic; sub-acoustic neuromuscular signals must be regulated as biometric data; and governance frameworks must be built from the perspective of the most structurally vulnerable data subject.

ABSTRACT. In the current era of the Human 5.0 paradigm, which is characterised by an unprecedented density of information flows and deep integration of humans into algorithmic environments, the cognitive system of the subject faces the achievement of a "metabolic ceiling". Contemporary cyberpsychology traditionally focuses on semantic data processing, often ignoring the energy cost of pre-semantic perception of visual symbols. This work presents the Cognitive Ease Light (CEL) index, an innovative metric designed to quantitatively assess the ‘metabolic costs’ (physiologic process of adenosine triphosphate consumption) of brain neural networks during the decoding of glyphs and forms before the analysis of their semantic content begins. The CEL index is developed based on the idea of 'linguistic-somatic resonance'. This is a new combination of Rudolf Arnheim's Gestalt aesthetics with the cross-modal correspondences of the 'Buba-Kiki' (maluma-takete) effect. Mathematically, the index describes the relationship between visual density, angularity, and gravitational weight of a form on the one hand, and the speed of its cognitive assimilation on the other. CEL is a reading metric that goes beyond standard metrics in that it takes into account the recipient's somatic response. Instead of considering the font as a mere carrier of information, it is regarded as a 'force field' that modulates the level of primary trust or anxiety. As outlined in the operationalised model, there are three key functional ranges of CEL index values that determine the psychophysiological mode of user interaction with the font interface: Cognitive ease mode: CEL>0.85 The use of rounded 'Buba' shapes (sonorous visual patterns) has been shown to minimise cognitive resistance. In this mode, information is perceived as safe, which puts the subject into a 'flow state' and blocks the risk of 'amygdala hijack' when encountering potentially triggering content. Cognitive mediation point: CEL<0.60 The resulting typographic friction forces the brain to switch to active analytical analysis. The incorporation of 'Kiki' stimuli (angular, dynamic shapes) has been shown to enhance alertness and focus. However, a lack of congruence (e.g., an aggressive font with empathetic text) can create an effect of 'visual sarcasm' that undermines trust in the source. Somatic hyper-mass mode: CEL 0.45-0.55 The deliberate creation of significant cognitive friction is achieved through the concept of the 'Sovereign Glyph'. The visual symbol is given virtual weight and density, activating the lateral-occipital tactile-visual area (LOtv). This approach transforms the process of reading into an act of 'bodily contemplation', acting as an antidote to 'digital anorexia' — the devaluation of meaning in sterile digital interfaces. The practical application of the CEL index facilitates the design of interfaces that preserve an individual's biological sovereignty, protecting their attention from the algorithmic hyper-speed of AI systems. The metric allows for precise customisation of the user's 'depth of immersion', tailored to the specific clinical or communicative task at hand. The validity of the index has been confirmed in interdisciplinary agent-oriented studies, where optimisation of CEL parameters reduced the level of 'structural friction' in conflict sessions by 22-25%, stabilising interaction even in conditions of deep axiological division. The proposal of the CEL index as a basic standard for designing secure and trustworthy communication protocols of the future is a key recommendation. The index would return the 'corporeality' of perception to the field of cyberpsychology.

ABSTRACT. Museums and memorials play an important role in helping people understand and connect with the past. Today, many of these spaces are beginning to incorporate digital tools to make visits more engaging. One such tool is augmented reality (AR), which overlays digital elements onto real environments and can transform how people experience historical places. This project sits at the intersection of cyberpsychology and heritage studies. As immersive technologies become increasingly present in everyday life, understanding how they shape cognition, emotion, and social attitudes is a growing priority for the field. Yet despite the rapid adoption of extended reality tools in cultural institutions, empirical research on their psychological effects in memorial and historical contexts remains limited. This study directly addresses that gap. The project explores how AR can be used in a memorial context, particularly when presenting difficult historical events such as deportation during World War II. The study focuses on four key dimensions of the visitor experience: historical empathy (both emotional engagement and cognitive understanding of historical context), overall emotional impact, and generalization, meaning whether the effects of the visit extend beyond the immediate experience to other contexts. Using a between-groups experimental design, young adult participants will be assigned to either a traditional non-AR condition at the museum or one of two AR conditions that vary in their level of narrative presence. Participants will complete psychological measures immediately after the experience and again one week later, allowing us to assess both the immediate impact and the durability of the effects. We expect that AR, particularly the condition designed to foster deeper narrative engagement, will lead to stronger emotional responses, greater historical empathy, and better contextual understanding compared to a traditional visit. We also anticipate that these effects will persist over time and generalize to other historical and contemporary contexts involving persecution and displacement. As an ongoing project, this research aims to better understand how immersive technologies can support meaningful engagement with history, rather than simply making visits more interactive.

ABSTRACT. By leveraging technologies like Python, and the use of the Flask framework, we aim to create an enhanced Military Information Support Operation (MISO) campaign. The purpose of the campaign is to provide adversaries positive information about the United States and to convince them to donate funds to support United States operations. Using Open AI's ChatGPT, and other large language models (LLMs), we created a clean User Interface for an adversarial citizen interested in investing, transitioning from establishing an account to a tailored visual presentation based on personal characteristics. The first component gathers personal information about the user. Following this, the user undergoes a 15-question personality test centered on the OCEAN personality characteristics model. On the basis of the OCEAN model results, they are sorted into one of three different financial profiles: over-controlled, resilient, and under-controlled (Exley et al., 2022; Campbell et al., 2023). The second component is centered on providing positive stimuli associated with the United States. The interface reveals a series of five-images paired with statements in progression, which are generated and tailored to each user. These images range from featuring slightly attractive to extremely attractive individuals who are increasingly aligned with pro-American imagery and symbolism, starting from a neutral background to a distinctly pro-American one. Each image contains two statements on basic investing and how U.S. policies improve lives in adversarial countries. The final component asks the user to donate to a cause supporting adversarial citizens, with links given to a charity and a legitimate investment company, regardless of their decision to donate to a cause.

ABSTRACT. The proliferation of health misinformation, especially on social media, has led scholars to focus on ways to mitigate adverse effects (Krishna & Thompson, 2021; Nan et al., 2022). Yet relatively little research has examined misinformation related to mental health (cf. Bizzotto et al., 2023; Starvaggi et al., 2024), which is a critically important public health issue. A growing body of research has examined factors that contribute to combatting misinformation, including responses by social media users. This study addresses what motivates people to challenge, correct or otherwise respond to mental health misinformation (MHM) on social media.

The influence of presumed media influence (IPMI) model describes an indirect media effect in which people respond to media messages in part based on their perceptions of how the messages impact others (Gunther & Storey, 2003). Based on the IPMI model, perception of harm to others from MHM may promote efforts to respond or support for content regulation/restrictions. Some people, especially those dealing with mental health issues, may regard the messages as threatening due to anticipated effects on others and protect themselves by not responding publicly. However, MHM could motivate people to publicly express themselves, to alert others to a public threat and reduce potential harms. This study examined how perceptions of MHM on social media predicted willingness to challenge/correct misinformation and support for content moderation by social media platforms. The role of online health literacy was also explored.

Social media users (N=174) recruited on MTurk completed a Qualtrics survey. After receiving a definition of mental health misinformation (MHM), respondents reported (on multi-item scales, all reliable) the presumed harm to others and perceived personal threat of such messages on social media. They also rated their online health literacy. Three measures assessed their responses to MHM on social media: likelihood of criticizing/correcting the misinformation publicly and privately, and support for content regulation by social media platforms. Personal mental health experience was also reported (60% had a mental health condition).

Participants with personal mental health experience reported greater presumed harm of MHM on others and greater perceived personal threat. Hierarchical regression analyses (controlling for background variables) examined predictors of responses to MHM on social media. Personal mental health experience and presumed harm of MHM were associated with greater willingness to provide public (but not private) critique, and greater support for content moderation. Perceived personal threat positively predicted all three response measures. Personal mental health experience did not moderate the relationship between perceived personal threat and intention to critique MHM. Online health literacy predicted higher scores on all three response measures.

Predictions grounded in the IPMI model were confirmed. Presumed harm to others was associated with greater intention to publicly–but not privately–critique misinformation. Publicly challenging misinformation can potentially reach a wider audience, thus reducing others’ susceptibility to influence, whereas private comments reach only those who posted false/misleading messages. Consistent with prior research, presumed harm to others was also associated with greater support for moderation by social media platforms. Concern about social harms appears to underlie preference for some level of regulation. The fact that people with mental health issues were more willing to challenge publicly suggests that, rather than withdrawing for self-protection, they were willing to alert others and potentially help reduce harm. Being open about mental health issues and challenging misinformation are ways some people resist stigma. Much additional research on this topic is needed. For example, studies should explore the role of factors such as message source and comments of other users affect how people respond to MHM.

ABSTRACT. Background Personality traits and trait emotional intelligence are widely treated as stable psychological antecedents of mental health outcomes. Computational models in digital mental health reinforce this assumption by encoding, for instance, personality and emotional intelligence as fixed input features while symptom scores in anxiety or depression serve as the predicted output. However, accumulating longitudinal evidence challenges this unidirectional framework. There is a possibility that sustained psychopathology might itself reshape personality over time through a process known as the scar effect, whereby mental health symptoms would produce measurable increases in neuroticism and decreases in conscientiousness within the same individual across time. Whether trait emotional intelligence might undergo comparable changes remains entirely uninvestigated. Anxiety and depressive symptom time-series features computed from repeated measurements represent a theoretically grounded and computationally tractable candidate pathway. Three features characterize each individual's time series: variability (SD: how much symptoms fluctuate overall), instability (MSSD: how erratically symptoms shift from one assessment to the next), and inertia (AR-1: the tendency for symptom state to resemble the previous one, reflecting temporal persistence in symptom levels across assessments).

This study will investigate whether anxiety and depressive symptom variability, instability and inertia, derived from fortnightly longitudinal time series of anxiety and depressive symptom assessments, are associated with within-person changes in personality traits and trait emotional intelligence.

Methodology Data will be drawn from the RABSYPO study, a prospective longitudinal cohort of 1,049 adult participants from all autonomous communities of Spain, with sex, age, and urbanicity distributions matched to national population characteristics [DOI: 10.1016/j.sjpmh.2024.08.003] who answered fortnightly online questionnaires about anxiety (GAD-7) and depressive symptoms (PHQ-9) every two weeks over one year. 942 participants completed the follow-up, yielding 26 symptom assessments per participant. Participants also completed two measurements ofParticipants also completed two measurements of personality traits (IPIP-50) and trait emotional intelligence (TEIQue-SF) collected 26 weeks apart (one started at week 6 and one at week 34).

Three time-seriestime-series features — SD, MSSD, and exploratorily AR-1 — will be extracted from each participant's GAD-7 and PHQ-9 time series across the 17 fortnightly assessments preceding the second IPIP-50 and TEIQue-SF measurements. In Latent Change Score Models (LCSM), these features will serve as predictor variables, whereas within-person changes in personality traits and trait emotional intelligence between baseline and follow-up will serve as outcome variables. Sociodemographic variables and mean symptom levels will be included as covariates. Multiple imputation will be used to handle missing data, and FDR will be used to correct for multiple testing.

Results We expect that participants with greater GAD-7 and PHQ-9 SD and particularly MSSD and AR-1 will be associated with end-of-study increases in neuroticism and decreases in conscientiousness and trait emotional intelligence. Effect sizes are expected to be small-to-moderate, consistent with prior longitudinal research on symptom-driven trait change. Model fit will be evaluated using Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardised Root Mean Square Residual (SRMR), with unstandardised and standardised coefficients reported alongside 95% bootstrap confidence intervals.

Conclusion Symptom time-series features may serve as candidate markers of trait-level personality and emotional intelligence change over time. If MSSD and AR-1 extracted directly from longitudinal GAD-7 and PHQ-9 time series capture this process, they could represent not only theoretically meaningful indicators but potentially actionable signals for digital mental health systems.

ABSTRACT. Advances in generative artificial intelligence (AI) have transformed the landscape of phishing attacks, enabling the creation of highly polished and linguistically flawless phishing emails that obscure many of the cues traditionally used to signal deception. At the same time, everyday email use often occurs under varying degrees of cognitive load (i.e., the mental effort required to maintain and process information in working memory), which can constrain attentional resources and impair analytical processing. The present study examines how cognitive load influences individuals’ ability to detect phishing emails, and whether this effect differs for AI-generated versus human-generated messages. Using a fully within-subjects 3 × 2 × 2 design, participants completed a dot-matrix working-memory task to induce low, moderate, or high cognitive load while evaluating the suspiciousness of 48 emails varying by email status (phishing vs. safe) and source (AI-generated vs. human-generated). Detection accuracy, confidence ratings, and performance on the cognitive-load manipulation were recorded. Data collection is now complete, and analyses using preregistered generalised linear mixed-effects models are currently being conducted to account for the repeated-measures structure of the data and trial-level variability. Based on preregistered hypotheses, we expect (i) reduced detection accuracy under higher cognitive load, (ii) lower accuracy for AI-generated compared with human-generated emails, and (iii) the greatest impairment when participants judge AI-generated phishing emails under high load. Full results, including tests of predicted two- and three-way interactions, will be presented. This work aims to clarify how cognitive load and increasingly sophisticated phishing content jointly shape susceptibility to digital deception.

ABSTRACT. The gut-brain axis plays a central role in emotional regulation and decision-making. Most serotonin is produced in the gastrointestinal tract, modulating visceral sensitivity and bidirectional communication with the central nervous system via the vagus nerve. Yet the specific contribution of gastric physiological activity to stress-related decision-making remains unexplored, a gap with clinical relevance given altered autonomic signalling in major depressive disorder (MDD) and bipolar disorder (BD). This cross-sectional psychophysiological study recruits 45 adults (15 healthy controls; 15 diagnosed with major depressive disorder; 15 diagnosed with bipolar disorder, aged 18 to 45) from Portugal. In a single lab session (45 to 65 min), participants completed the EADS-21 (designed to measure the severity of depression, anxiety, and stress), underwent sensor setup, and provided a 5–10 minute physiological baseline. Stress was induced via the virtual-reality Trier Social Stress Test (VR-TSST), with electrogastrography (EGG), electrocardiography (ECG), and electrodermal activity (EDA) recorded continuously. Immediately after, participants completed the Balloon Analog Risk Task (BART) as a measure of stress-affected decision-making. The change in scores from baseline are computed for EGG, ECG, and EDA; regression and mediation analyses examined whether gastric reactivity predicts stress-affected BART performance, with depression and bipolar disorder status as a moderator. An LLM was used to bootstrap the raw psychophysiological data collected, and the model's imputed outputs will be compared against the empirical results upon study completion to evaluate the accuracy and reliability of LLM-based data bootstrapping.

ABSTRACT. Psychological language has travelled a long way from the therapist's chair to the TikTok floor, becoming the dominant vocabulary through which romantic relationships are interpreted, evaluated, and increasingly acted upon. From "avoidant attachment" to "knowing your worth", the informal use of clinical and therapeutic language (therapy-speak) now structures how millions of users on TikTok understand intimacy. Yet how this language is organised, circulated, and taken up in everyday relational life has received little empirical attention in cyberpsychology. This study addresses that gap.

Using summative content analysis and reflexive thematic analysis of 31 TikTok videos and associated comments, this research offers an empirical mapping of relationship-related therapy-speak as a distinct cyberpsychological phenomenon. The dataset spans content from self-identified qualified therapists and non-qualified creators, published in 2025–2026 and selected against systematic engagement thresholds to ensure sufficient interactional context.

The findings reveal a recurring discursive formula: relational problems are identified, reframed through therapeutic concepts, and resolved through individual self-work. An original six-domain taxonomy was developed — spanning nervous system regulation, attachment-based labelling, self-development, relational evaluation, explicit clinical terminology, and love languages. This formula was reproduced with equal fluency by qualified therapists, unverified creators, and — in an unexpected finding — accounts that appeared to be AI-generated, presenting relational prescriptions with full authority and no disclosure. Therapeutic authority on TikTok, the analysis shows, is performed rather than credentialed.

The discourse is also deeply gendered, framing relational responsibility and self-improvement in essentialist terms, and it is actively co-constructed in comment sections, where therapeutic frameworks are applied to real relationships and partners evaluated along a spectrum from "the bare minimum" to "princess treatment." Audience data reveal that therapy-speak is not merely consumed but enacted: users described deploying the language of needs and autonomy to renegotiate real partnerships in real time. Therapy-speak does not remain on screen; it extends beyond TikTok's platform and into everyday relational life.

Interpreted through Self-Determination Theory, this discourse invokes needs for autonomy, competence, and relatedness, but constructs a motivational orientation closer to introjection than to intrinsic motivation, in which relationships are evaluated against external standards rather than experienced as inherently satisfying. Drawing on Eva Illouz's concept of emotional capital, relationship-related therapy-speak functions as social currency: a vernacular through which individuals signal self-awareness, benchmark relational effort, and position themselves within the moral economy of modern relationships. Now that this currency is also being minted by AI systems optimised for engagement rather than insight, questions of authenticity, authority, and accountability in platformised intimacy become increasingly pressing.