Telematics and Informatics最新文献

Among the numerous ways to communicate anonymously in a virtual environment, many people prefer using an avatar as a tool to represent themselves. Use of an avatar allows people to reveal their psychological ownership (PO) as well as to create a visual presentation of themselves. Thus, avatars are a means to differentiate the self from others and to attract attention. This study determines if a person with an avatar that they designed to resemble themselves experiences PO. This study verified the direct effect through 2 (personalization vs. customization) x 2 (ideal self vs. actual self) analysis of variance. There was a direct effect on the PO of the avatar only when it was customized. In addition, among the four IVs, the PO of the avatar was highest in cases of customization x actual self (Group 3). Moreover, the serial multiple mediator model was verified through Hayes process macro model 6. A positive mediating effect was shown in the serial mediator model where IV affects the PO of the avatar via attachment, but there is no mediating effect via avatar identification. The results of this study suggest the use of PO to motivate users to continue to consume content.

With an experimental design, this study examined the effect of source cues (Human vs. AI) on hostile media bias through heuristic machine evaluation of machine and human social media profiles. This study also explored the effects of affective and cognitive involvement as moderators along with media source-self ideological incongruity (source incongruity). A 2 (human vs. AI) x 3 (CNN vs. USA Today vs. Fox News) experimental study was conducted (n = 434). Participants exhibited less hostile media bias when presented with a news story with AI source cues through heuristic machine evaluation. The mitigating effect was stronger for those viewing news from an incongruent news source. Such moderated mediated effect was further moderated by two types of involvement (i.e., affective and cognitive). Implications for future research surrounding the two types of involvement, source incongruity, machine heuristic evaluations, and hostile media bias are discussed in light of our findings.

Today the sharing economy, involving the online sharing of under-utilized assets with unknown others, has increasingly become a prevalent e-commerce business model. While numerous studies have explored determinants of a guest’s booking intention in the sharing economy, little attention has been given to what drives a host to accept a booking request from a guest. A host endeavors to gather and assess information about a guest to decide whether to accept a request, and ultimately, a host's perceptions of a guest are likely to impact their willingness to share accommodation online. This study investigates the impact of a host's perceptions of guests on their intention to share an unused property with an unknown guest. It aims to present and test a conceptual model of accommodation sharing by examining the effects of a guest profile and guest reviews on a host's trust in a guest and, subsequently, on a host's intention to share accommodation based on signaling theory. This research collected data from 270 Airbnb users, aged 20 to 70, using an online survey on Amazon MTurk. The results indicated that a host's trust in a guest was positively affected by perceived guest favorability from guest reviews and perceived credibility from guest profiles. However, the perceived fit between a host’s guest preferences and the portrayed guest identity did not have a significant effect on trust. Furthermore, it was found that trust in a guest had a positive relationship with sharing intent but a negative relationship with perceived risk. Based on the results, this study offers practical as well as academic implications and suggests directions for future research.

Both in China and across the rest of the world, livestream shopping is developing at a considerable rate. Non-transactional consumer engagement has been demonstrated to be essential to the long-term growth of livestream commerce; however, few studies have addressed this topic. Using the socio-technical approach, the present study proposes a theoretical framework to explore the antecedents of non-transactional consumer engagement in livestream commerce. To obtain comprehensive insights into these antecedents, we employed structural equation modeling (SEM) and fuzzy set qualitative comparative analysis (fsQCA) for data analysis. According to the SEM results, multiple cues, personal focus, and identification with streamers all have a positive and significant influence on emotional support and informational support, which in turn lead to greater consumer engagement. However, immediate feedback is associated with informational support, but not with emotional support. The results of fsQCA confirm the importance of the proposed antecedents and provide configurations that increase consumer engagement. The study findings not only enrich the literature regarding livestream shopping but also provide practical insights for practitioners to make consumer engagement strategies based on their own strengths. In addition, this paper demonstrates the feasibility of hybrid methods in the livestream shopping research.

COVID-19 pandemic restrictive measures curtailed environmental stimuli and social contact while encouraging media consumption. This study investigated isolation, media use, sleep and stress factors related to involuntary phenomena with and without media content. Video game players answered an online survey (N = 397; males 70.8 %; mean age = 28.04). Experiencing intrusions during lockdown (mean = 65.81 min outdoors per day) was common: earworms (92 %), Game Transfer Phenomena (GTP) (e.g., re-experiencing game content such as images, sounds, thoughts and impulses) (91 %), intrusive thoughts/imagery (ITI) (90 %), mobile phone intrusions (MPI) (e.g., vibration, ringing) (67 %) and hallucinations (49 %). All intrusions were correlated; however, different factors were related to the likelihood of experiencing them. Time playing and time using social media were correlated with intrusions with and without media content. No correlations were found with watching something or using a computer. Older age decreased the odds of GTP, ITI and hallucinations, although factors such as stress and meeting people outside the household had higher effects. Stress, rather than isolation, media use or sleep factors, appear to have been a key factor for most of the intrusions during lockdown. Particularly younger players appear to be susceptible to intrusions with and without media content.

Deepfake videos are highly realistic manipulated videos that are often difficult to distinguish from authentic videos. The technology rapidly evolves, making deepfake videos increasingly realistic. Most research on deepfake videos is focused on the algorithmic detection of deepfakes. Less is known about the human recognition of deepfake videos. The aim of the current study was to investigate the predictors of human performance at recognizing deepfake videos. Our findings show that humans perform better at recognizing deepfake videos of familiar persons compared to deepfakes of unfamiliar persons. Next, our findings show a positive relationship between time spent on social media and deepfake detection performance, as well as conspiracy thinking and deepfake detection performance. No relationships were found between age and gender and deepfake detection performance.

Crowdfunding, as enabled by donors worldwide, has played a critical role in financially supporting military forces, NGOs, humanitarian groups, and civilians in the Ukraine-Russia Crisis. This global mobilization underscores the evolving dynamics in the modern geopolitical landscape, emphasizing the transformative power of online collective action in fundraising and organizing. This study draws on signaling theory to pinpoint success indicators in crowdfunding for Ukraine projects on GoFundMe. Recognized as an effective explanatory framework, signaling theory is used as it bridges fundraising tactics with donation behavior in crowdfunding contexts. Particularly, the research examines how emotions and semantics contributed to (or hindered) fundraising success through sentiment analysis, topic modeling, and paragraph vectors. Findings revealed that trust showcased in fundraisers’ texts and videos amplified donations. Conversely, expressions of sadness deterred donations. Furthermore, the strategic deployment of an anti-Putin narrative notably enhanced fundraising results, but this strategy was way more likely to be deployed by fundraisers based in the United States. The distinct linguistic patterns of US campaigns were so pronounced that machine learning classifiers could predict the fundraiser’s origin (US vs. non-US) with 82% accuracy based solely on language. Overall, the study sheds light on the importance of emotion regulation and “cheap talk” as functional signaling strategies that facilitate value alignment. It provides new theoretical insights and methodological feasibility of examining how strategic communication affects financial outcomes in crowdfunding.

Virtual humans’ communication of emotions in social media posts are considered heuristics of human behavior. The current study extends the Cognitive-Functional Model of Emotions to virtual influencers and investigates how social media users engage with virtual human emotions. In an experiment (N = 362), non-emotional and emotional (i.e., happy, sad, and lust) social media posts are examined for their influence on Generation Z and Millennial (18–40 years old) users’ affective responses and subsequent attitudinal and behavioral outcomes within the moderation of eeriness based on the uncanny valley paradigm. Findings indicate affective responses and attitudes are serial mediators between a virtual human’s display of emotion and social media users’ behavioral intentions, and eeriness was a significant moderator. Lust has the strongest effect, followed by happiness, sadness, and no emotion. This research elevates the Cognitive-Functional Model to discrete virtual human emotions with approach tendencies. Implications for understanding the impact of virtual humans’ emotions and sexual appeal on social media users’ engagement are discussed.

The potential of Generative AI, such as ChatGPT, has sparked discussions among researchers and the public. This study empirically explores the capabilities and limitations of ChatGPT, specifically its portrayal of environmental justice issues. Using OpenAI’s ChatGPT API, we asked ChatGPT (GPT-4) to answer questions about environmental justice issues in 3,108 counties in the contiguous United States. Our findings suggest that ChatGPT provides a general overview of environmental justice issues. Consistent with research, ChatGPT appears to acknowledge the disproportionate distribution of environmental pollutants and toxic materials in low-income communities and those inhabited by people of color. However, our results also highlighted ChatGPT’s shortcomings in detailing specific local environmental justice issues, particularly in disadvantaged (e.g., rural and low-income) counties. For instance, ChatGPT could not provide information on local-specific environmental justice issues for 2,593 of 3,108 counties (83%). The results of the binary logistic regression model revealed that counties with lower population densities, higher percentages of white population, and lower incomes are less likely to receive local-specific responses from the ChatGPT. This could indicate a potential regional disparity in the volume and quality of training data, hinting at geographical biases. Our findings offer insights and implications for educators, researchers, and AI developers.