Zeyang Yang

Online health information spreads quickly across different publishers and platforms. However, little is known about how information source characteristics, together with cyberchondria, influence individuals’ trust and intentions to share and verify. Using a 2 × 2 × 3 between-subject design, the present study investigated the effects of cyberchondria severity (high/low), publisher type (verified/unverified), and platform type (health app/search engine/social media) on individuals’ trust in health information and their intentions to share and verify that information. A total of 204 participants were included in this experiment. The results show that verified publishers and health apps can increase trust and the intention to share. Individuals with high cyberchondria show higher trust and intentions to share and verify, only when the publisher was unverified. The findings suggest that external authority (verification of accounts and professional content apps) can moderate the effects of cyberchondria or its related anxiety. The present study contributes to the literature on the determinants of online health behaviours by using an experimental design, with practical implications for platform management and online health information publishers.

People have become accustomed to studying or working with the guidance of artificial intelligence (AI) in recent years. Studies have begun investigating the risk factors of AI dependence, though most have used hypothesis-testing methods. The present study aimed to investigate predictors of AI dependence using machine-learning and network analysis, which are data-driven approaches. The included risk factors were Big Five personality traits, self-efficacy, depression, social anxiety, adverse childhood experiences, and AI use motivation, selected based on theories and empirical studies. Participants consisted of 1258 university students (942 females and 316 males) with a mean age of 22.11 years (SD = 2.69). Four machine-learning algorithms were tested, including Elastic Net, Random Forest, XGBoost, and LightGBM. Machine-learning results indicate that escape and social motivation for AI use, along with social anxiety, were the main predictors of AI dependence. Network analysis results show that escape and social motivation were the most central nodes, with the highest Expected Influence (EI) indices. This study indicates that when addressing mental health problems related to AI dependence, it is more effective to focus on emotional isolation and social interaction challenges rather than simply cutting down on AI use.

Short-form video platforms, including TikTok, have rapidly gained popularity. Their emotionally charged, algorithm-curated content can promote addiction and distort time perception. However, there is a scarcity of research on time perception within the realm of short videos. This study aimed to explore the influence of short video emotions, individual emotions, short video addiction, and flow on time perception using experimental and survey methods. Study 1 recruited 158 participants (Mean age = 21.50 years, SD = 2.16), and employed a 2 (time estimation paradigm: prospective, retrospective) × 3 (video emotion: positive, neutral, negative) between-subjects experimental design to investigate the impact of short video emotions, time estimation paradigms, short video addiction, and flow on time perception. Findings revealed that short video emotions influenced time perception, with positive videos leading to an underestimation of duration and a faster perception of time passage. Additionally, both short video emotions and addiction predicted the perception of time passage through flow. Study 2 conducted a survey of 525 participants (Mean age = 21.53 years, SD = 2.14) and utilised structural equation modelling to explore the associations between individual emotions (depression, anxiety, boredom), short video addiction, flow, and time perception. Depression and short video addiction predicted time perception through flow, further clarifying the relationships among these variables. The findings extend the I-PACE model and flow theory to popular short-video media, highlighting the role of flow in linking emotions, addiction, and time perception during short-video use.

Studies have used network analysis to show the progression of problematic internet use (PIU). The present study examined the longitudinal comorbidity network models of specific types of PIU and psychological distress over time. A total of 685 undergraduate students (mean age = 19.12 years) completed three waves of surveys assessing two specific types of PIU (problematic social media use [PSMU] and problematic online video watching [POVW]) and psychological distress (depression, anxiety, and stress). Results showed that the ‘down-heartedness’ symptom of depression and the ‘agitation’ symptom of stress were consistently the most central symptoms in the PIU-psychological distress comorbidity network. ‘Mood modification’ and ‘conflict/functional impairment’ related to social media were the bridging symptoms that connected the other symptoms in the first two waves of studies. The ‘no initiative’ symptom of depression became the key bridging symptom in the third wave. Similarly, cross-lagged panel network analysis showed that PSMU symptoms (e.g., ‘conflict/functional impairment’) were the most influential and susceptible symptoms in earlier waves, while depression symptoms became key predictors of other symptoms over time in the comorbidity network. The findings provide new perspectives and means for the diagnosis and treatment of PIU and associated mental health illnesses.

Short videos’ increasing popularity and increased user engagement have sparked concerns about time perception. While studies have linked gaming or watching TV series to time loss, research on short videos’ temporal impact is scarce. This study aims to investigate the impact of short video use on time distortion (including perceptions of time for experimental tasks and weekly usage) through an experimental design. Fifty-six college students were randomly assigned to two time duration conditions (long-duration for 16 min 9 s or short-duration for 5 min 23 s). Participants in both conditions were instructed to watch short videos and read public articles for the same duration and then estimate the time duration of the tasks. Subsequently, participants completed a questionnaire about their estimated and actual weekly short video use and problematic short watching levels. The results showed that the impact of task duration on time perception was significant. Task type had no significant impact on time perception, with no notable difference in time estimation between conditions involving watching short videos and reading. The interaction between time duration and task type was not significant. Additionally, problematic short video watching and the estimated weekly short video use were not significantly related to time distortion. This study contributes to empirical research on time distortion while watching short videos, providing insights for expanding theoretical models of addictive behaviors and interventions for problematic short video use.

Purpose: While previous cross-sectional studies have investigated the relationship between problematic online video watching and mental health issues, longitudinal studies exploring their bidirectional relationship remain relatively scarce. This study aimed to fill this gap by conducting a two-wave longitudinal design. Participants and Methods: Six hundred and ninety-nine first-year undergraduates voluntarily completed two rounds of questionnaire surveys assessing problematic online video watching, boredom proneness and loneliness with an interval of 4 months. To test the research hypotheses, cross-lagged panel models were performed using AMOS 24.0. Results: Our results revealed the that (1) Boredom proneness and problematic online video watching were mutually influenced. (2) Loneliness at T1 positively predicted the level of problematic online video watching at T2 but not vice versa. (3) Boredom proneness mediated the link between loneliness and problematic online video watching. (4) Chinese first-year undergraduates’ levels of problematic online video watching and boredom proneness significantly increased during their educational transition period from high school to university. Conclusion: The present study enhances our understanding of the dynamic relationships between problematic online video watching, boredom proneness and loneliness, highlighting the targeted prevention and interventions for first-year undergraduates.

Danmaku interaction, an innovative method of engaging with online videos, has rapidly gained popularity. However, research on how danmaku (user-generated comments overlaid on videos) affects cognitive processes, such as attention and time perception, remains limited. This study investigated how the quantity of danmaku, as a factor of intrinsic cognitive load, influences the accuracy of video time estimation. Additionally, within the framework of cognitive load theory, this study explored how flow and addiction levels moderate the effect of effective cognitive load, as measured by actual fixations on danmaku, on time estimation. Findings show that an increase in danmaku quantity decreases the accuracy of time estimation, particularly in high-density danmaku videos, likely due to elevated intrinsic cognitive load. Moreover, flow and addiction levels were identified as key moderators, amplifying or dampening the relationship between effective cognitive load and time estimation. Utilizing eye-tracking technology, we uncovered the mechanisms underlying time estimation in danmaku videos, providing empirical evidence of how danmaku affects attention allocation, time perception, and cognitive load. These results have practical implications for video platform design, content creation, and advertising strategies, contributing to the development of a healthier online environment. Future research could expand on these findings by examining a broader range of video types and contexts to further understand the influence of intrinsic cognitive load on time estimation.