The present article represents a collaborative effort completed by five co-authors, with my contribution focusing specifically on the literature review and data processing sections. The text below encompasses the content of those segments, while the comprehensive full-text version is attached as a PDF file for your reference.

View the full paper (CN)

Abstract

This study focuses on the contemporary representative Internet new media platform Xiaohongshu APP, conducting a survey targeting college students in Beijing, employing quantitative research methods, integrating information adoption models with social influence theories, and utilizing SPSS for data analysis to examine the impact of online health information on users’ adoption intentions. Through this investigation, the research aims to clarify how dimensions of perceived trust and perceived risk affect users’ adoption willingness and to discern the roles played by information quality and hot level within the context of perceived trust.

Literature Review

Research on User-Generated Content (UGC)

User-Generated Content (UGC) emerged within the framework of Web 2.0, referring to user-driven participation in online platform development where they actively share and display their owned or created content. With the launch of video platform YouTube in 2005, the UGC model swiftly became a focal point in digital content evolution. Scholarly research on both theoretical underpinnings and practical applications of UGC has gradually burgeoned, leading to increased attention from various industries. New media platforms widely adopt the UGC model, relying on users for the integration and management of online resource information. [4]

In its early days, studies on UGC in new media primarily focused on video and blogging platforms. Sungmyung Huh [5] et al. employed the “Uses and Gratifications” theory to investigate users’ satisfaction and motivations behind using UGC platforms, asserting that self-expression constitutes a primary factor driving users to share and create videos on these platforms. Choi et al. [6] and Kim et al. [7], through their studies on video and blog platforms, posited that information interaction is a critical characteristic of UGC, emphasizing that user interfaces need to be designed adaptively to facilitate direct interaction between users and platform activities. Jae [8] also explored the effects of user control, synchronicity, and responsiveness on new media services and user information perception based on information interaction theories.

As mobile Internet became mainstream and the transition from Web2.0 to Web3.0 unfolded, the applicability of the “Uses and Gratifications” theory came under scrutiny. Ichimizu Hoshi [9] analyzed four case studies related to UGC and argued that this theory overlooks the trend of audiences transitioning into communicators in the new media era, attempting to break away from traditional media analysis frameworks and presenting a thought-provoking critical perspective. Overall, UGC research in new media mainly concentrates on three dimensions: platform users, interactive content, and new media technologies.

Health Information Adoption Studies

In 1948, when the World Health Organization was founded, health was defined as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity.”

Broadly speaking, health information encompasses all knowledge, news, data, facts, and materials related to health or illness. With the widespread adoption of the internet and mobile devices, the boundaries of health information dissemination have expanded, barriers have been broken down, and the conventional healthcare environment has progressively transformed into an online space for exchange and sharing. The health information discussed in this study within the context of new media platforms refers to web-based health information broadly relating to people’s physical and mental health, illnesses, nutrition, wellness, and can be roughly categorized into two types: health and medical information and health lifestyle information. [10] In recent years, a series of user information behaviors associated with health information, such as searching, acquisition, evaluation, application, and feedback, have increasingly attracted attention.

Health information adoption refers to the aggregate of all stages of behavior from the acquisition to the utilization of health information by individuals. [11] Closely related concepts include information reception, which typically implies passive receipt of information, whereas information adoption emphasizes the proactive agency of the individual. Current research on health information adoption largely falls into two categories: one focusing on causes, investigating which factors influence individuals’ adoption of health information, and another concentrating on outcomes, exploring the impact of health information adoption on individual health behaviors.

Traditional research on health information adoption tends to lean towards the causal aspect. The classic Information Adoption Model (IAM) was proposed by Sussman and Siegal, who built upon the Technology Acceptance Model (TAM) and the Elaboration Likelihood Model (ELM), suggesting that information adoption behavior is mediated by perceived usefulness, which in turn is influenced by the quality of arguments and the credibility of information sources. [12] Subsequent studies on the influencing factors of health information adoption mostly base their investigations on this model. For instance, Deng Shengli et al., drawing from TAM2 and IAM, incorporated perceived risk into a new model and conducted empirical research on health information adoption among users of Q&A platforms through questionnaire surveys. [13] Additionally, other scholars have delved into alternative factors potentially affecting health information adoption. Mo Xiuting et al., for example, examined the influence of health self-efficacy on health information adoption, finding it could serve as both a moderator and mediator variable. [14] Han Xiaoxiao et al.’s research on health information adoption among urban elderly revealed that individual social capital impacts their health information adoption. [15]

In recent times, the impact of health information adoption on individual health behaviors has also garnered interest from researchers. Some scholars suggest that examining the relationship between changes in health behaviors and health information adoption can provide informational support for modifying individual health behaviors, effectively promoting the change of unhealthy behaviors and enhancing personal health management capabilities. [16] [17] Therefore, research on health information adoption is currently in a stage where both causality and outcome are considered equally important. [18]

Domestic Research on Health Information Adoption on UGC Platforms

Regarding health information adoption on video-based UGC platforms, Peng Sihao et al. [19] studied the communication characteristics of health-related short videos on Douyin (TikTok) and the factors influencing user information behavior. They found that perceived ease of use and perceived usefulness positively affect user information behavior while perceived risk negatively influences it. In the process of disseminating health-related short videos, information quality, source credibility, users’ health concerns, and subjective norms all exert positive effects on information adoption. Li Li et al. [20] established a research framework for the adoption of health-related short video information based on the Information Adoption Theory and the Heuristic-Systematic Model. Using Structural Equation Modeling (SEM), they analyzed the effect paths of antecedent variables on user health information adoption and employed Qualitative Comparative Analysis (QCA) to explore the combined effects of influential conditions on short video users’ health information adoption. SEM results showed that content quality, platform reputation, and source credibility significantly impact perceived usefulness, which in turn affects users’ adoption intention; content quality and perceived usefulness act as chain mediators in the effect of expressiveness on adoption intention. fsQCA analysis revealed that users’ health information adoption decisions on short videos are jointly influenced by systematic and heuristic cues. Wang Xinglan et al. [21] constructed a theoretical model based on the Uses and Gratifications Theory, Health Belief Model, and Elaboration Likelihood Model, discovering that social gratification needs, learning gratification needs, perceived health threats, and perceived health expectations positively influence information adoption behavior, while argument quality, source credibility, and video appeal indirectly affect information adoption.

In the context of specialized online health communities, Liu Zhu [22] introduced perceived trust as a mediating variable and used information quality and information popularity derived from conformity theory to measure users’ perceived trust. Perceived risk was used as an independent variable affecting both perceived trust and user information adoption, with health self-efficacy serving as another independent variable influencing user information adoption. Findings suggested that information quality (source reliability, information accuracy, and timeliness) significantly positively impacts perceived trust; information popularity significantly positively influences perceived trust; perceived risk significantly negatively affects perceived trust but does not significantly influence user information adoption; both perceived trust and health self-efficacy positively and significantly affect user information adoption. Perceived trust partially mediates the relationship between information quality and user information adoption, and fully mediates the relationship between information popularity and user information adoption. Zhu Yunqin et al. [23] utilized the Elaboration Likelihood Model combined with Information Need Theory to examine how high-level central processing (information quality) and low-level peripheral processing (carrier quality and source credibility) in an online health community environment affect users’ perception of information usefulness, subsequently influencing their information search behavior. Their findings indicated that for peripheral processing, users’ perception of usefulness and emotional response are positively and significantly affected by source credibility and carrier quality; for central processing, users’ perception of usefulness and emotional response are positively and significantly affected by information quality. Moreover, the relationships between information quality and source credibility with users’ perception of information usefulness are negatively moderated by the motivational variable of information need.

Research Methodology

Research Hypotheses

Based on the theory of health self-efficacy, the following hypothesis is proposed:

H1: Health self-efficacy among college students positively affects their health information adoption.

Considering that health information quality comprises aspects like accuracy and timeliness [33], and taking into account the strong emotional nature of content expression on Xiaohongshu (Little Red Book), this study adds information emotional intensity to the concept of information quality. Drawing from relevant theories of the Information Adoption Model, the following hypotheses are put forward:

H2: The quality of health information on the Xiaohongshu platform positively influences users’ perceived usefulness. H2a: The emotional intensity of health information on the Xiaohongshu platform positively influences users’ perceived usefulness. H2b: The accuracy of health information on the Xiaohongshu platform positively influences users’ perceived usefulness. H2c: The timeliness of health information on the Xiaohongshu platform positively influences users’ perceived usefulness.

Drawing from theories related to the Information Adoption Model and referencing Chou et al.’s research on source credibility, the following hypotheses are presented:

H3: The credibility of health information sources on the Xiaohongshu platform positively influences users’ perceived usefulness. H3a: The credibility of health information providers on the Xiaohongshu platform positively influences users’ perceived usefulness. H3b: The platform credibility of Xiaohongshu positively influences users’ perceived usefulness.

Grounded in Social Influence Theory, the following hypotheses are advanced:

H4: Social influence of health information on the Xiaohongshu platform positively influences users’ perceived usefulness. H4a: User consensus on health information on the Xiaohongshu platform positively influences users’ perceived usefulness. H4b: The popularity or information热度 of health information on the Xiaohongshu platform positively influences users’ perceived usefulness.

Consistent with the Information Adoption Model, the final hypothesis is:

H5: College students’ perceived usefulness of health information on the Xiaohongshu platform positively influences their health information adoption.

Questionnaire Structure Design

The survey questionnaire consists of two main sections: a Likert scale section and a section collecting basic information about the respondents. The Likert-scale portion is structured as a matrix with 3-6 items designed for each dimension based on the research model and hypotheses. The basic information section primarily gathers data on the gender, educational level, and major of the respondents.

Past studies have shown that medical and social science majors exhibit relatively higher levels of media literacy when processing and accepting information, with no significant statistical differences observed among other majors. Consequently, this study targets college students from medical, social science, and other majors, administering the same standardized questionnaire across three independent surveys.

The research questionnaire is divided into three parts:

1. Introduction: This part outlines the purpose of the survey, assures confidentiality of responses, and expresses gratitude to the participants. It also defines the scope and definition of health information before filtering out participants who have browsed or read health information on Xiaohongshu, focusing primarily on those who have had exposure to such content.
2. Core Content: This section contains the heart of the questionnaire, measuring five constructs within the research model through 37 measurement items:
   • Health Self-Efficacy (5 items)
   • Information Quality (10 items):
     • Information Emotional Intensity (4 items)
     • Information Accuracy (3 items)
     • Information Timeliness (3 items)
   • Source Credibility (7 items):
     • Platform Credibility (3 items)
     • Provider Credibility (4 items)
   • Social Influence (6 items):
     • User Consensus (2 items)
     • Information Popularity (4 items)
   • Perceived Usefulness (4 items)
   • Health Information Adoption (5 items) All measurement items utilize a 5-point Likert scale (1: strongly disagree, 2: disagree, 3: neutral, 4: agree, 5: strongly agree).
3. Personal Information: This part collects demographic details, including gender, academic status, and specific field of study.

Pilot Survey and Questionnaire Revision

The pilot questionnaire consisted of 15 questions, including 11 Likert scale items and 4 personal information questions. A pilot survey was conducted among university students in Beijing, resulting in 123 questionnaires distributed and 73 valid responses, giving an effective return rate of 59.35%. Based on the pilot survey data analysis, several improvements were made:

• Since there was almost no correlation between the respondent’s school and their usage or adoption behaviors, the item asking about the “school attended” was removed. Instead, the “major” question was refined to provide clearer and more precise options.
• Some items in the Likert scale had low factor loadings or lacked specificity, indicating low effectiveness. Such invalid items were eliminated from the final questionnaire.

The revised formal questionnaire comprised 11 matrix Likert scale sets and 3 basic information questions, with a reduction in the number of questions within the matrices and improved clarity and consistency in their phrasing and logical sequence.

Formal Sampling Survey

For the formal survey, three criteria were set for selecting participants: (1) they must have used the Xiaohongshu platform, (2) they must have a certain degree of interest in health information, and (3) they must be currently enrolled college students. Through snowball sampling, eligible participants were asked to recommend others who met these requirements.

Based on 2021 data released by the Ministry of Education on the number of undergraduate students by discipline category, the ratio of students in the medical category to those in four other categories (law, education, management, and economics) was approximately 1:4. This ratio was approximated for the Beijing student population to ensure a consistent sample size ratio in the two comparative studies.

The formal survey began on June 19, 2023, and ended on June 25, 2023, lasting for seven days. A total of 300 questionnaires were distributed, 281 were returned, and 174 were deemed valid, yielding an effective recovery rate of 61.92%.

The data analysis section is omitted here; please refer to the attachment for details.

References(partially)

• [4] 范哲, 朱庆华, 赵宇翔. Web2.0 环境下 UGC 研究述评[J]. 图书情报工作, 2009,53(22): 6-63.
• [5] 성명훈, 이인희. Uses and Gratifications of User-Created Contents: Expressing Self with Self-Produced Video Clips[J]. Korean Journal of Communication & Information, 2007,40(4): 45-79.
• [6] CHOI C, HO-HYUN J. A research on “Avatar blog"as a suggestion for emotion based community in which UCC information is share[J]. Journal ofDigital Design, 2007,7(2): 179-188.
• [7] KIM H, 김희정. A Study on UI Design guideline for efficient UCC(User Created Content) Video Service - Focused on Viewing Interface and Uploading Interface -[J]. Journal of Digital Design, 2007,7(2): 295-304.
• [8] JAE L Y. Effect of Web-services’ Technological Interactivity on User-generated Contents’ Perceived Efficiency and Quality[J]. The Journal of the Korea Contents Association, 2012,12(9): 380-388.
• [9] 池见星. 论新媒体时代传者与受者的身份趋同——用户自创内容(UGC)研究路径探析[J]. 东南学术, 2009(04): 166-168.
• [10] Suri V R, Majid S, Chang Y K, Foo S.Assessing the Influence of Health Literacy on Health Information Behaviors:A Multi-Domain Skills-Based Approach[J].Patient Education&Counseling,2015,99(6):1038-1045.
• [11] 朱庆华,杨梦晴,赵宇翔等.健康信息行为研究：溯源、范畴与展望[J].中国图书馆学报,2022,48(02):94-107.DOI:10.13530/j.cnki.jlis.2022017.
• [12] Sussman S W, Siegal W S. Information influence in organizations: Anintegrated approach to knowledge adoption[J].Information System Research,2003,14(1):47-65
• [13] 邓胜利,管弦.基于问答平台的用户健康信息获取意愿影响因素研究[J].情报科学,2016,34(11):53-59.DOI:10.13833/j.cnki.is.2016.11.011.
• [14] 莫秀婷,邓朝华.健康自我效能对基于社交网站采纳健康信息的影响分析[J].中国卫生统计,2015,32(05):753-757.
• [15] 韩啸,黄剑锋.基于社会资本理论的城市老年人健康信息采纳研究[J].西南交通大学学报(社会科学版),2017,18(03):95-104.
• [16] 李桂玲,曹锦丹,王崇梁,兰雪.信息行为干预对不良健康行为改变进程的影响研究[J].图书情报工作,2017,61(23):108-113.DOI:10.13266/j.issn.0252-3116.2017.23.013.
• [17] 曹锦丹,王崇梁.健康行为改变不同阶段的信息框架效应概念模型研究[J].图书情报工作,2019,63(05):23-31.DOI:10.13266/j.issn.0252-3116.2019.05.003.
• [18] 朱庆华,杨梦晴,赵宇翔等.健康信息行为研究：溯源、范畴与展望[J].中国图书馆学报,2022,48(02):94-107.DOI:10.13530/j.cnki.jlis.2022017.
• [19] 彭思豪. 抖音健康类短视频用户信息行为影响因素研究[D]. 西南交通大学, 2020.
• [20] 李力, 韩平, 张弘, 等. 启发-系统式线索对移动短视频用户健康信息采纳的影响研究——基于 SEM 和 QCA 的混合方法[J]. 农业图书情报学报, 2023(01):73-86.
• [21] 王兴兰, 胡虹, 肖廷超. 大学生健康科普短视频采纳行为的影响因素实证研究[J]. 情报探索, 2023(02):84-89.
• [22] 刘助. 在线健康社区的用户信息采纳影响因素研究[D]. 西安电子科技大学, 2021.
• [23] 朱云琴, 陈渝. 双路径视角下在线健康社区用户健康信息搜寻行为影响因素研究[J]. 图书馆杂志, 2022,41(10):83-96.
• [33] Sun Y L, Hwang H, Hawkins R, et al. Interplay of Negative Emotion and Health Self-Efficacy on the Use of Health Information and Its Outcomes[J]. Communication Research, 2008,35(3):358-381.DOI:
• 10.1177/0093650208315962.
• [34] 胡颢琛. PGC 健康信息特点和医生博主信任机制构建——以小红书平台为例[J]. 青年记者, 2023(02):65-67

完整论文和附图均在附件PDF中

！Download Full Paper