Xiaobei Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiaobei Chen has a total of 33 co-authors. Here are the top co-authors by number of shared publications:
- Kun Xu
College of Journalism and Communications, University of Florida , Gainesville, FL
3 shared publications - Fanjue Liu
USC-SJTU Institute of Cultural and Creative Industry, Shanghai Jiao Tong University
2 shared publications - Aantaki Raisa
Washington University in St. Louis
1 shared publication Alicia Papanek
Department of Family, Youth, and Community Sciences University of Florida Gainesville Florida USA
1 shared publicationBenjamin K. Johnson
University of Florida
1 shared publication- Benjamin Lok
University of Florida, Gainesville, USA
1 shared publication
Latest Publications
- 28 July 2026
- Published by Oxford University Press (OUP) in Annals of the International Communication Association
Charting the evolution of human–machine communication: a systematic review of 25 years of empirical research
This study systematically examines trends and patterns in human–machine communication (HMC) research from 2000 to 2024. Based on a comprehensive screening of studies published in 28 communication and interdisciplinary journals and 2 foundational books, we analyzed 209 peer-reviewed articles based on their metadata, theoretical frameworks, research methods, technologies of interest, and research themes. The findings show steady growth in HMC scholarship following its formal recognition as a subfield with the establishment of the ICA Interest Group in 2019, continued reliance on social scientific research designs, and increasing interdisciplinary contributions. A BERTopic analysis identified 9 coherent research clusters, reflecting a gradual shift from robot-focused investigations toward a broader range of studies on virtual agents, chatbots, and generative AI. Overall, this review maps HMC’s intellectual trajectory and highlights key directions for future theoretical, methodological, and interdisciplinary development. It also discusses the increasingly ambiguous boundaries of HMC given the growing use of large language models and the convergent spaces between HMC and AI-mediated communication.
- 28 July 2026
- Published by Oxford University Press (OUP) in Annals of the International Communication Association
Charting the evolution of human–machine communication: a systematic review of 25 years of empirical research
This study systematically examines trends and patterns in human–machine communication (HMC) research from 2000 to 2024. Based on a comprehensive screening of studies published in 28 communication and interdisciplinary journals and 2 foundational books, we analyzed 209 peer-reviewed articles based on their metadata, theoretical frameworks, research methods, technologies of interest, and research themes. The findings show steady growth in HMC scholarship following its formal recognition as a subfield with the establishment of the ICA Interest Group in 2019, continued reliance on social scientific research designs, and increasing interdisciplinary contributions. A BERTopic analysis identified 9 coherent research clusters, reflecting a gradual shift from robot-focused investigations toward a broader range of studies on virtual agents, chatbots, and generative AI. Overall, this review maps HMC’s intellectual trajectory and highlights key directions for future theoretical, methodological, and interdisciplinary development. It also discusses the increasingly ambiguous boundaries of HMC given the growing use of large language models and the convergent spaces between HMC and AI-mediated communication.
- 25 April 2026
- Published by Taylor & Francis in Health Communication
- p. 1-16
See That Post, Get That Shot: Content and Time-Series Analyses of Health Belief Model Constructs on CDC’s Instagram
COVID-19 has posed unprecedented challenges to our public health system but also offered opportunities for us to examine how the Centers for Disease Control and Prevention (CDC), the flagship health authority in the United States which was considered the gold standard globally before the pandemic, designs persuasive vaccine-related messages over time. Grounded in the health belief model (HBM), this study employed quantitative content analysis of 858 vaccine-related Instagram posts from the CDC's @cdcgov account between 2021 and 2023. Then, a time-series analysis was conducted to consider the effect of CDC messaging trends on real-world vaccine behaviors. We found that the CDC posted more frequently between June 2021 and January 2022, and most posts utilized at least one HBM construct. Among six HBM constructs, benefit, self-efficacy, and cues to action were three dominant constructs that appeared on Instagram posts throughout the pandemic. Time-series analysis and Granger causality tests indicate significant causal links between all HBM constructs and newly administered vaccines, except for severity. Interestingly, acknowledging barriers was the only construct that immediately produced a positive effect on newly administered vaccines. In contrast, benefit, susceptibility, and self-efficacy messages exerted short-term negative impacts, yet showed positive influence in the following days. Theoretical contributions and implications are discussed.
- 25 April 2026
- Published by Taylor & Francis in Health Communication
- p. 1-16
See That Post, Get That Shot: Content and Time-Series Analyses of Health Belief Model Constructs on CDC’s Instagram
COVID-19 has posed unprecedented challenges to our public health system but also offered opportunities for us to examine how the Centers for Disease Control and Prevention (CDC), the flagship health authority in the United States which was considered the gold standard globally before the pandemic, designs persuasive vaccine-related messages over time. Grounded in the health belief model (HBM), this study employed quantitative content analysis of 858 vaccine-related Instagram posts from the CDC's @cdcgov account between 2021 and 2023. Then, a time-series analysis was conducted to consider the effect of CDC messaging trends on real-world vaccine behaviors. We found that the CDC posted more frequently between June 2021 and January 2022, and most posts utilized at least one HBM construct. Among six HBM constructs, benefit, self-efficacy, and cues to action were three dominant constructs that appeared on Instagram posts throughout the pandemic. Time-series analysis and Granger causality tests indicate significant causal links between all HBM constructs and newly administered vaccines, except for severity. Interestingly, acknowledging barriers was the only construct that immediately produced a positive effect on newly administered vaccines. In contrast, benefit, susceptibility, and self-efficacy messages exerted short-term negative impacts, yet showed positive influence in the following days. Theoretical contributions and implications are discussed.
- 5 April 2026
- Published by Wiley in Social Policy & Administration
- Vol. 60 (5) , 974-990
Queering the Social Media Narrative: Influencers as Public Health Authorities
Heteronormative public administration assumes that LGBTQIA+ communities will reflect their heterosexual counterparts. The resulting deficient outreach to vulnerable communities has worsened health disparities, as seen in the 2022 Mpox epidemic, in which gay and bisexual men experienced the highest infection rates. This research applied queer theory to examine social media efforts to eliminate Mpox spread. This study critically assessed 591 social media messages from influencers, non‐profits, and government for queer‐focused and heteronormative discourse. Findings revealed that most governmental and non‐profit accounts failed to augment content for the at‐risk audience. Gay and bisexual men had to navigate heteronormative content to locate resources critical to their well‐being. Conversely, influencers acted as a public health system through repeated messaging and by supplementing governmental guidance with steps tailored for a queer audience. Influencers' value came from leveraging their lived experiences to challenge discriminatory societal norms, highlighting the crucial role of culturally relevant communication.
- 5 April 2026
- Published by Wiley in Social Policy & Administration
- Vol. 60 (5) , 974-990
Queering the Social Media Narrative: Influencers as Public Health Authorities
Heteronormative public administration assumes that LGBTQIA+ communities will reflect their heterosexual counterparts. The resulting deficient outreach to vulnerable communities has worsened health disparities, as seen in the 2022 Mpox epidemic, in which gay and bisexual men experienced the highest infection rates. This research applied queer theory to examine social media efforts to eliminate Mpox spread. This study critically assessed 591 social media messages from influencers, non‐profits, and government for queer‐focused and heteronormative discourse. Findings revealed that most governmental and non‐profit accounts failed to augment content for the at‐risk audience. Gay and bisexual men had to navigate heteronormative content to locate resources critical to their well‐being. Conversely, influencers acted as a public health system through repeated messaging and by supplementing governmental guidance with steps tailored for a queer audience. Influencers' value came from leveraging their lived experiences to challenge discriminatory societal norms, highlighting the crucial role of culturally relevant communication.
- 16 March 2025
- Published by Elsevier in Patient Education and Counseling
- Vol. 136, 108752
Assessing virtual patients for empathy training in healthcare: A scoping review
- 16 March 2025
- Published by Elsevier in Patient Education and Counseling
- Vol. 136, 108752
Assessing virtual patients for empathy training in healthcare: A scoping review
- Open Access
- 7 March 2025
- Published by JMIR Publications Inc. in JMIR Cancer
- Vol. 11, e73616
Virtual Health Assistants in Preventive Cancer Care Communication: Systematic Review (Preprint)
Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Virtual health assistants (VHAs), interactive digital programs that emulate human communication, are being increasingly used in healthcare to improve patient education, care, and to reduce burden on healthcare providers. VHAs have the potential to promote cancer equity through facilitating patient engagement, providing round-the-clock access to information, and reducing language barriers. However, it is unclear to what extent audience-centeredness is being considered in the development of cancer-related applications. Objective: This systematic review identifies and synthesizes strategies used to make VHA-based cancer prevention and screening interventions audience-centered. Methods: Following PRISMA 2020 guidelines, we searched four databases (PubMed, Embase, Web of Science, EBSCOhost) for peer-reviewed studies on VHA interventions promoting cancer screening (January 2022). Included studies focused on adult populations in primary care settings, with interventions emphasizing interactivity and immediacy (key VHA features). Excluded were studies on cancer treatment, non-interactive decision aids, or technical VHA development. Screening, data extraction, and quality assessment (Mixed Methods Appraisal Tool) were performed independently by multiple reviewers. Thematic synthesis was used to analyze audience-centered strategies. Results: Of 1,055 records screened, 17 studies met inclusion criteria. Most (n=11) targeted colorectal cancer, with others addressing prostate, breast, cervical, or lung cancer. Sixteen studies were U.S.-based; one focused on Uganda. Key strategies for audience-centered design included: 1) Demographic Concordance: Race/gender alignment between VHA and users (e.g., African American participants interacting with Black-coded avatars); 2) User Feedback: Iterative testing via interviews, think-aloud protocols, or pilot studies to refine interventions; 3) Pre-Intervention Needs Assessment: Identifying cultural, linguistic, or literacy barriers (e.g., myths about screening in Ugandan communities); 4) Theoretical Frameworks: Health Belief Model (most common), MAIN model, or tailored messaging theories guided design; 5) Information Customization: Culturally adapted content (e.g., Spanish-language interfaces, narratives addressing racial disparities); and 6) Feature Customization: Adjusting VHA appearance (e.g., animations, fonts) based on user preferences. Notably, 7/17 studies focused on racially minoritized groups (e.g., African Americans, Hispanic farmworkers), addressing systemic barriers like mistrust in healthcare. However, gaps persisted in intersectional tailoring (e.g., rurality, income) and non-English languages (only 2/17 studies). Recruitment methods influenced diversity; community-based strategies yielded more representative samples than online-only approaches. Conclusions: The systematic review identified the audience-centered development practices currently being used for VHA-based interventions in preventive cancer care. The majority of the studies included processes to diversify and segment the intended audience, focused on medically underrepresented population groups, and implemented strategies to be culturally sensitive to the population of interest. Yet, opportunities remain to address multidimensional inequities (e.g., rural access, low literacy). Future interventions should integrate intersectional frameworks, expand language diversity, and measure social presence to enhance engagement. This review provides a roadmap for developing equity-focused eHealth tools in cancer prevention.
- 7 March 2025
- Published by JMIR Publications Inc. in JMIR Cancer
- Vol. 11, e73616
Open AccessVirtual Health Assistants in Preventive Cancer Care Communication: Systematic Review (Preprint)
Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet #Preprint #PeerReviewMe: Warning: This is a unreviewed preprint. Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn. Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period. Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author). Background: Virtual health assistants (VHAs), interactive digital programs that emulate human communication, are being increasingly used in healthcare to improve patient education, care, and to reduce burden on healthcare providers. VHAs have the potential to promote cancer equity through facilitating patient engagement, providing round-the-clock access to information, and reducing language barriers. However, it is unclear to what extent audience-centeredness is being considered in the development of cancer-related applications. Objective: This systematic review identifies and synthesizes strategies used to make VHA-based cancer prevention and screening interventions audience-centered. Methods: Following PRISMA 2020 guidelines, we searched four databases (PubMed, Embase, Web of Science, EBSCOhost) for peer-reviewed studies on VHA interventions promoting cancer screening (January 2022). Included studies focused on adult populations in primary care settings, with interventions emphasizing interactivity and immediacy (key VHA features). Excluded were studies on cancer treatment, non-interactive decision aids, or technical VHA development. Screening, data extraction, and quality assessment (Mixed Methods Appraisal Tool) were performed independently by multiple reviewers. Thematic synthesis was used to analyze audience-centered strategies. Results: Of 1,055 records screened, 17 studies met inclusion criteria. Most (n=11) targeted colorectal cancer, with others addressing prostate, breast, cervical, or lung cancer. Sixteen studies were U.S.-based; one focused on Uganda. Key strategies for audience-centered design included: 1) Demographic Concordance: Race/gender alignment between VHA and users (e.g., African American participants interacting with Black-coded avatars); 2) User Feedback: Iterative testing via interviews, think-aloud protocols, or pilot studies to refine interventions; 3) Pre-Intervention Needs Assessment: Identifying cultural, linguistic, or literacy barriers (e.g., myths about screening in Ugandan communities); 4) Theoretical Frameworks: Health Belief Model (most common), MAIN model, or tailored messaging theories guided design; 5) Information Customization: Culturally adapted content (e.g., Spanish-language interfaces, narratives addressing racial disparities); and 6) Feature Customization: Adjusting VHA appearance (e.g., animations, fonts) based on user preferences. Notably, 7/17 studies focused on racially minoritized groups (e.g., African Americans, Hispanic farmworkers), addressing systemic barriers like mistrust in healthcare. However, gaps persisted in intersectional tailoring (e.g., rurality, income) and non-English languages (only 2/17 studies). Recruitment methods influenced diversity; community-based strategies yielded more representative samples than online-only approaches. Conclusions: The systematic review identified the audience-centered development practices currently being used for VHA-based interventions in preventive cancer care. The majority of the studies included processes to diversify and segment the intended audience, focused on medically underrepresented population groups, and implemented strategies to be culturally sensitive to the population of interest. Yet, opportunities remain to address multidimensional inequities (e.g., rural access, low literacy). Future interventions should integrate intersectional frameworks, expand language diversity, and measure social presence to enhance engagement. This review provides a roadmap for developing equity-focused eHealth tools in cancer prevention.
- 27 November 2024
- Published by Taylor & Francis in The Social Science Journal
- Vol. 62 (3) , 688-703
What do people say about Replika, an AI chatbot, on social media? Investigating diverse perspectives on the implications of Replika through a topic modeling analysis
- 27 November 2024
- Published by Taylor & Francis in The Social Science Journal
- Vol. 62 (3) , 688-703
What do people say about Replika, an AI chatbot, on social media? Investigating diverse perspectives on the implications of Replika through a topic modeling analysis
- Open Access
- 20 May 2024
- Published by MDPI AG in Healthcare
- Vol. 12 (10) , 1049
Comparison of Financial Hardship and Healthcare Utilizations Associated with Cancer in the United States Medicare Programs during the COVID-19 Pandemic
Background: In the United States, Medicare beneficiaries diagnosed with cancer often face significant financial challenges due to the expensive nature of cancer treatments and increased cost-sharing responsibilities. However, there is limited knowledge regarding the financial hardships and healthcare utilizations faced by those enrolled in Medicare Advantage (MA) compared to those in traditional fee-for-service Medicare (TM) during the COVID-19 pandemic. Our study aims to investigate the subjective financial hardships experienced by individuals enrolled in TM and MA and to determine whether these two Medicare programs exhibit differences in healthcare utilization during the pandemic. Methods: We utilized data from the 2020–2022 National Health Interview Survey (NHIS), focusing on nationally representative samples of cancer survivors aged 65 or older. Financial hardship was categorized into three distinct groups: material (e.g., problems with medical bills), psychological (e.g., worry about paying), and behavioral (e.g., delayed care due to cost). Healthcare utilization included wellness visits (preventive care), emergency care services, hospitalizations, and telehealth. We used survey design-adjusted analysis to compare the study outcomes between MA and TM. Results: Among a weighted sample of 4.4 million Medicare beneficiaries with cancer (mean age: 74.9), 76% were enrolled in MA plans. Cancer survivors with a college degree (59.3% vs. 49.8%) and high family income (38.2% vs. 31.1%) were more likely to enroll in MA plans. There were no significant differences in any material, psychological, or behavioral financial hardship domains between beneficiaries with MA and TM plans except forgone counseling due to cost. For healthcare utilization measures, cancer survivors in MA were more likely than those in TM to have flu vaccination (77.2% vs. 70.1%) and experience lower hospitalizations (16.0% vs. 20.0%). However, there were no differences in other health service utilizations between MA and TM. Conclusion: While no significant differences were observed in any materialized, psychological, or behavioral financial hardships, older cancer survivors enrolled in MA plans were more likely to receive vaccinations and lower hospitalization rates during COVID-19. Although other preventive or primary care visits (i.e., wellness visits) were higher, their difference did not reach statistical significance. As MA grows in popularity, it is essential to consistently monitor and evaluate the performance and outcomes of Medicare plans for cancer survivors as we navigate the post-pandemic landscape.
- 20 May 2024
- Published by MDPI AG in Healthcare
- Vol. 12 (10) , 1049
Open AccessComparison of Financial Hardship and Healthcare Utilizations Associated with Cancer in the United States Medicare Programs during the COVID-19 Pandemic
Background: In the United States, Medicare beneficiaries diagnosed with cancer often face significant financial challenges due to the expensive nature of cancer treatments and increased cost-sharing responsibilities. However, there is limited knowledge regarding the financial hardships and healthcare utilizations faced by those enrolled in Medicare Advantage (MA) compared to those in traditional fee-for-service Medicare (TM) during the COVID-19 pandemic. Our study aims to investigate the subjective financial hardships experienced by individuals enrolled in TM and MA and to determine whether these two Medicare programs exhibit differences in healthcare utilization during the pandemic. Methods: We utilized data from the 2020–2022 National Health Interview Survey (NHIS), focusing on nationally representative samples of cancer survivors aged 65 or older. Financial hardship was categorized into three distinct groups: material (e.g., problems with medical bills), psychological (e.g., worry about paying), and behavioral (e.g., delayed care due to cost). Healthcare utilization included wellness visits (preventive care), emergency care services, hospitalizations, and telehealth. We used survey design-adjusted analysis to compare the study outcomes between MA and TM. Results: Among a weighted sample of 4.4 million Medicare beneficiaries with cancer (mean age: 74.9), 76% were enrolled in MA plans. Cancer survivors with a college degree (59.3% vs. 49.8%) and high family income (38.2% vs. 31.1%) were more likely to enroll in MA plans. There were no significant differences in any material, psychological, or behavioral financial hardship domains between beneficiaries with MA and TM plans except forgone counseling due to cost. For healthcare utilization measures, cancer survivors in MA were more likely than those in TM to have flu vaccination (77.2% vs. 70.1%) and experience lower hospitalizations (16.0% vs. 20.0%). However, there were no differences in other health service utilizations between MA and TM. Conclusion: While no significant differences were observed in any materialized, psychological, or behavioral financial hardships, older cancer survivors enrolled in MA plans were more likely to receive vaccinations and lower hospitalization rates during COVID-19. Although other preventive or primary care visits (i.e., wellness visits) were higher, their difference did not reach statistical significance. As MA grows in popularity, it is essential to consistently monitor and evaluate the performance and outcomes of Medicare plans for cancer survivors as we navigate the post-pandemic landscape.
- 28 November 2023
- Published by SAGE Publications in American Journal of Health Promotion
- Vol. 38 (3) , 394-401
Understanding the Design of Fear Appeals by Applying the Extended Parallel Process Model: A Qualitative Analysis of COVID-19 Public Service Announcements
Purpose: Identify how early COVID-19 public health messages incorporated in the tenets of the extended parallel process model (EPPM). Setting: YouTube videos developed by governmental departments, medical institutions, news organizations, and non-profit organizations in the United States were aggregated. Method: This qualitative study conducted a keyword search to identify public service announcements (PSAs). The sample was further refined after searching PSAs that contained fear appeals. A thematic analysis was performed by using the constant comparative method. Sample: A total of forty-three videos was included in the final analysis. Result: Two themes emerged regarding messages aimed at arousing the perceived severity of threat. These themes include emphasizing the consequences of being infected and utilizing personal narratives. Perceived susceptibility of threat was aroused by emphasizing that some groups have higher risks than others. Two themes emerged around arousing perceived response efficacy: (1) the authority of professionals; and (2) altruism and personal responsibility. One way was identified to arouse perceived self-efficacy, which is informing the protective measures. Conclusion: Multiple strategies were used in PSAs about COVID-19 to arouse fear during the early stages of the pandemic. The utilization of self-efficacy was oversimplified, by not providing details about the rationale for the recommended behavior.
- 28 November 2023
- Published by SAGE Publications in American Journal of Health Promotion
- Vol. 38 (3) , 394-401
Understanding the Design of Fear Appeals by Applying the Extended Parallel Process Model: A Qualitative Analysis of COVID-19 Public Service Announcements
Purpose: Identify how early COVID-19 public health messages incorporated in the tenets of the extended parallel process model (EPPM). Setting: YouTube videos developed by governmental departments, medical institutions, news organizations, and non-profit organizations in the United States were aggregated. Method: This qualitative study conducted a keyword search to identify public service announcements (PSAs). The sample was further refined after searching PSAs that contained fear appeals. A thematic analysis was performed by using the constant comparative method. Sample: A total of forty-three videos was included in the final analysis. Result: Two themes emerged regarding messages aimed at arousing the perceived severity of threat. These themes include emphasizing the consequences of being infected and utilizing personal narratives. Perceived susceptibility of threat was aroused by emphasizing that some groups have higher risks than others. Two themes emerged around arousing perceived response efficacy: (1) the authority of professionals; and (2) altruism and personal responsibility. One way was identified to arouse perceived self-efficacy, which is informing the protective measures. Conclusion: Multiple strategies were used in PSAs about COVID-19 to arouse fear during the early stages of the pandemic. The utilization of self-efficacy was oversimplified, by not providing details about the rationale for the recommended behavior.
- 21 November 2023
- Published by Edward Elgar Publishing
- p. 337-353
The Media Are Social Actors paradigm and beyond: theory, evidence, and future research
The computers are social actors paradigm was proposed about three decades ago. Recently, scholars have attempted to extend this classic theoretical framework to further enhance its explanatory and predictive power. As one of the attempts, the media are social actors (MASA) paradigm was proposed to indicate the different power of social cues in evoking users’ medium-as-social-actor presence and social responses to a range of emerging technologies, including social robots and smartphones. The MASA paradigm further included nine propositions that scholars can rely on to derive research questions. Based on these propositions, experiments and meta-analyses have been conducted to test and refine the paradigm. This chapter summarizes the findings of these studies and compares the empirical evidence with the original MASA paradigm. These findings further point to some of the future research directions on human-AI interaction.
- 21 November 2023
- Published by Edward Elgar Publishing
- p. 337-353
The Media Are Social Actors paradigm and beyond: theory, evidence, and future research
The computers are social actors paradigm was proposed about three decades ago. Recently, scholars have attempted to extend this classic theoretical framework to further enhance its explanatory and predictive power. As one of the attempts, the media are social actors (MASA) paradigm was proposed to indicate the different power of social cues in evoking users’ medium-as-social-actor presence and social responses to a range of emerging technologies, including social robots and smartphones. The MASA paradigm further included nine propositions that scholars can rely on to derive research questions. Based on these propositions, experiments and meta-analyses have been conducted to test and refine the paradigm. This chapter summarizes the findings of these studies and compares the empirical evidence with the original MASA paradigm. These findings further point to some of the future research directions on human-AI interaction.