Despite the emergence of the dark web more than 20 years ago, little scholarly attention has focused on identifying potential mental health differences between dark web users and surface web users. Yet, given the pseudo-anonymous nature of the dark web and the purported privacy it provides, individuals with mental health vulnerabilities may be inclined to use the dark web. In the present study, we investigate this matter by drawing on survey data collected in 2024 from a national sample of 2,000 U.S. adults. The results of both bivariate and multivariate analyses indicate that dark web users exhibit greater depressive symptoms and have more paranoid thoughts than surface web users. Likewise, dark web users are more likely than surface web users to report suicidal thoughts, nonsuicidal self-injury, and engagement in digital self-harm. Discussion centers on the implications of these findings for practice as well as avenues for future research.

First responders (FRs) are routinely exposed to traumatic events, increasing risk for posttraumatic stress disorder (PTSD). This study compared heart rate variability (HRV) and skin conductance level (SCL) between FRs with and without probable PTSD at baseline and during a virtual reality (VR) task. Eighty-four FRs completed questionnaires and physiological assessments. Participants with probable PTSD showed significantly lower baseline HRV, indicating reduced parasympathetic modulation. No group differences emerged for HRV during VR or for SCL at either point. The results confirm reduced HRV at rest in PTSD, but further work is needed to clarify why this difference was not observed during the task and why SCL showed no group effects. To advance understanding of these results, future studies should include larger samples, longer baselines, recovery phases, and clinical interviews.

Virtual reality (VR) and biofeedback have emerged as promising tools for mindfulness training. However, their combined effectiveness, compared with Traditional Mindfulness formats, remains understudied. This study investigated the short-term effects of a brief mindfulness intervention delivered via VR, with and without biofeedback, on psychological and physiological outcomes. Seventy-two participants (64.7 percent women; aged 18-57 years, = 24.0) were randomly assigned to one of three groups ( = 24 per group): mindfulness with VR and biofeedback, mindfulness with VR only, and traditional audio-guided mindfulness. Self-report measures assessed negative emotional symptoms, state anxiety, affect, and present-moment awareness before and after the intervention. Heart rate was recorded as a psychophysiological index of arousal. Results indicated significant reductions in stress, anxiety, and heart rate and an increase in positive affect across all groups. The VR + biofeedback group showed significantly greater improvements in receptive awareness and attentional focus compared with the other conditions. These findings support the use of VR-based mindfulness and suggest that integrating biofeedback may enhance present-moment engagement.

While individuals exhibit heightened attentional bias toward social media information, it remains debated whether this reflects a content-specific priority or a broader cognitive adaptation. Using visual search tasks, self-report scales, and the Embedded Figures Test, this study investigated attentional capture by digital icons and the role of cognitive style. Results showed that icons, particularly for field-independent individuals, elicited greater capture, suggesting a generalized digital attentional bias, and cognitive style play a key role in the process.

Previous research on digital media and well-being showed that the impact of media technologies on users largely depends on specific usage characteristics, context, and individual user factors. This research report explores the existence of distinct user profiles in virtual reality (VR), using an existing dataset and applying latent class analysis. The results point to the existence of three classes: two of them primarily include VR game users, who differ in the intensity of their psychological engagement (sense of presence) during gameplay, and a third class that includes users of social VR platforms, who report a relatively very high level of engagement with these platforms. The three groups differ in symptoms of addiction to the technology. These results provide valuable proof of concept for the usefulness of identifying usage profiles as a way to reveal how the relationship between VR and well-being might vary across different usage situations.

Although geographically dispersed organizations increasingly rely on virtual platforms to collaborate, virtual communication can undermine key team processes and outcomes. Prior research has largely focused on individual-level explanations, such as cognitive strain or "Zoom fatigue," for these challenges. We extend this literature by proposing that virtual communication also reinforces hierarchical structures by amplifying disparities in member influence during decision-making. In a controlled experiment comparing video conferencing and face-to-face teams, we find that disparities in member influence are significantly greater in virtual teams, which in turn reduces task performance. These findings highlight a critical, group-level mechanism through which virtual communication shapes communication patterns and outcomes, beyond previously identified individual-level factors. By identifying disparity in member influence as a key mediator, this study advances theory on virtual communication, group hierarchy, and decision-making and offers practical implications for reducing hierarchical distortions and fostering more equal-level conversations that enhance team effectiveness.

Research on dating applications relies heavily on explicitly measured evaluations of potential partners; however, the quick swiping mechanism of Tinder and other popular dating applications may rely on more spontaneous evaluations. We asked heterosexual U.S. undergraduates ( = 135) to rate opposite-gender Tinder profiles using both an implicit reaction time task and explicit self-report. Men had more positive implicit evaluations of sexualized profiles compared with women. Men also showed an implicitly measured preference for sexualized profiles over nonsexualized profiles, while women showed the opposite preference. However, the genders did not differ in evaluation of sexualized or nonsexualized profiles in explicitly measured evaluations. This study highlights the value of using implicit measures when conducting research on mobile dating.

Romantic partner phubbing (RPP) is when one ignores their romantic partner in favor of one's smartphone. Although RPP may be related to poorer mental health, less is known whether social comparison moderates this association. The present study investigated how social comparison moderates the associations between RPP and depressive, anxiety, and stress (DAS) symptoms in Hispanic emerging adults. Hispanic emerging adult college students ( = 210) completed a survey assessing demographics, RPP, social comparison, and DAS symptoms. Results indicated a significant interaction between RPP and social comparison such that lower levels of social comparison (i.e., increased upward comparison) strengthened the positive associations between RPP and DAS symptoms. The clinical implications of these findings are discussed; future prospective studies are warranted to assess temporality.

This study was performed to clarify the relationship between smartphone usage and depression risk among older adults. We also examined smartphone use variations, population heterogeneity, and the mechanisms behind smartphone-related depression symptoms. A total of 10,997 older adults (mean age = 84.31 years) from a national cohort (Chinese Longitudinal Healthy Longevity and Happy Family Study) in China were included. Smartphone use was measured by use status (yes/no) and contents. Among eight dichotomous questions about contents of smartphones, chatting, shopping, traveling, and entertainment were classified as routine-oriented activities. In contrast, health management, financial management, and learning were classified as functional/cognitive-oriented activities. Thus, participants were categorized into four groups, including not using a smartphone, using smartphones both for routine-oriented and functional/cognitive-oriented activities, only for routine-oriented activities, and others. Smartphone users had lower 10-item Center for Epidemiological Studies Depression Scale scores ( = -0.62; < 0.001) and a reduced depression risk (odds ratio = 0.77; < 0.001) compared with nonusers. Furthermore, those using smartphones for routine-oriented and functional/cognitive-oriented activities had even lower risks than those using them only for routine-oriented activities. Second, using smartphones was more sensitive to alleviating depression among older adults aged <80 years or living in rural areas. Third, mediation analyses suggested that using smartphones may be associated with depressive symptoms through communicating with children regularly. Our findings suggest that smartphone use may have a positive effect on mental health, particularly among those who engage in diverse routine-oriented and functional/cognitive-oriented activities. Smartphone benefits are more evident in younger seniors and rural residents, indicating a need for targeted interventions for these groups.

Social media use has brought not only greater connectivity but also growing concerns about its psychological and physiological consequences. Guided by the conceptual framework of the allostatic load model, this study investigated the impact of communication overload-the feeling of receiving more social input than one can manage, as commonly experienced in social media use-on physical health, with social media fatigue and depressive symptoms as serial mediators. A national quota sample of 1625 adults in South Korea completed an online survey measuring communication overload associated with social media use, social media fatigue, depressive symptoms, and self-rated physical health, along with relevant demographic and behavioral covariates. Results revealed a significant indirect pathway from communication overload to physical health, sequentially through both social media fatigue and depressive symptoms. Our findings indicated that the strain of social demands associated with social media use may contribute to deterioration in both mental and physical health. By applying the allostatic load framework, this study contributes to the technostress literature by elucidating how communication overload associated with social media use and its underlying mechanisms affect health, laying the groundwork for future research integrating psychological and physiological perspectives.

While the effects of interpersonal social comparison have been widely explored, the potential consequences of human-machine comparison remain understudied. The present survey research investigates how individuals compare themselves to generative artificial intelligence (GAI), by extending the concept of social comparison orientation to human-GAI interactions. Using a large Chinese sample ( = 1302), we examine the relationships between two subtypes of human-GAI comparison orientation (ability-based and opinion-based), self-esteem, perceived identity threat, anthropocentric beliefs, and personal relative deprivation within the framework of the Rejection-Identification Model. The findings reveal that both ability-based and opinion-based comparisons with GAI positively relate to self-esteem, diverging from relevant findings on interpersonal comparisons. In addition, ability-based comparisons with GAI significantly heighten the perceived identity threat, whereas opinion-based comparisons with GAI may alleviate it. Furthermore, anthropocentric beliefs (i.e., human-first attitudes) suppress the negative effect of identity threat on self-esteem, which is also mediated by personal relative deprivation. In addition, a exploratory multigroup analysis was conducted based on usage frequency. The present study highlights the importance of investigating human-AI comparison and expands the explanatory power of classical concepts and frameworks from the self and identity domains, contributing to psychological interventions aimed at mitigating the potential adverse effects driven by human-machine relationships.

Despite widespread discussions about artificial intelligence (AI) and its impact on society, little work has objectively measured how often people use this technology in the wild. The present article collected up to 90 days of web-browsing data from students (Study 1: 499) and those in the general public (Study 2: = 455), quantifying how often people used AI and evaluating the psychological correlates of such use. Upon coding 4.1 million websites in Study 1 and 9.9 million websites in Study 2, the evidence suggested that AI use was relatively infrequent, totaling 1% of student web-browsing and 0.44% of general public web-browsing, on average. The most consistent predictors of AI use across studies were aversive personality traits (e.g., Machiavellianism, narcissism, psychopathy), albeit the traits were differentially associated with AI use across studies. Demographics were largely unrelated to AI use across studies. Finally, we observed that self-reported AI use and actual AI use were only moderately correlated (ρ = 0.329), suggesting limitations in subjective measures of media use. These findings provide some of the first behavioral measurements of AI in naturalistic settings and establish important benchmarks for understanding the individual differences associated with AI adoption.

As the second most frequently visited website globally, YouTube attracts numerous people who spend significant amounts of time on the platform, potentially leading to problematic consequences. To investigate the underlying mechanism of problematic video-based social media use, the present study translated and culturally adapted the (YAS) into two Chinese versions for people in Taiwan and Hong Kong. Psychometric properties and measurement invariance across regions (Taiwan and Hong Kong) and sex (male and female) were further validated. Participants from Taiwan ( = 887) and Hong Kong ( = 1,008) completed an online survey comprising the , (), and () between September 2023 and June 2024. The results supported adequate concurrent validity of YAS with the ACSID-11 and BSMAS. A robust single-factorial structure with strong factor loadings and good internal consistency was observed among the two Chinese versions. Measurement invariance across regional and sex groups was also supported by all fit indices. Accordingly, the current findings suggest that the YAS can be used as a validated instrument to assess individuals' problematic YouTube use in Taiwan and Hong Kong. Future studies may investigate the psychosocial influences of problematic use of video-based social media platforms or use the scale in studies of active and passive use of social media.

Artificial intelligence (AI) is shaping charitable donations more and more. That means traditional theories about human fundraisers might not apply when AI is involved. In charitable storytelling, fundraisers use first-person or third-person narrative perspectives to convey their messages. Different narratives have varying effects on audiences' donation intentions, yet the most suitable narrative perspective for AI fundraisers remains unexplored. Drawing upon narrative transportation theory (NTT), we ran three studies exploring the interaction between narrative perspective and fundraiser identity (AI vs. human) in shaping people's intentions to donate. Study 1 found that people were more willing to donate money for first-person narratives from AI but third-person narratives from humans. Study 2 found evidence that this difference happens because of empathic concern. The different narrative perspectives influence people's empathic responses, which influence their donation intentions. Study 3 found that the identifiability of the person seeking help is a boundary condition. When the help-seeker's identity is ambiguous, the advantages behind first-person and third-person narratives disappear. This research extends the application of NTT to the field of human-computer interaction. The results provide empirical evidence that nonprofits can use to optimize their use of AI in fundraising.