Determinants of medical crowdfunding information sharing attitudes through integrated source credibility and technology acceptance models

Abstract

Medical crowdfunding has emerged as an important alternative for financing healthcare expenses, enabling individuals to raise funds through donation-based campaigns. While social media facilitates campaign outreach, concerns regarding privacy, misinformation, and trust hinder information-sharing behaviors, limiting campaign effectiveness. This study contributes to this understanding by integrating Source Credibility Theory and the Technology Acceptance Model to examine how expertise, trustworthiness, perceived usefulness, and perceived ease of use influence attitudes toward sharing medical crowdfunding information. A cross-sectional survey was conducted among 345 young social media users in China, with data analyzed using PLS-SEM. The findings reveal that expertise does not directly impact sharing attitudes but influences them through perceived usefulness, whereas trustworthiness exerts both significant direct and indirect effects on sharing attitudes. Additionally, perceived usefulness and perceived ease of use positively predict sharing attitudes, highlighting the importance of practical value and platform usability in driving user engagement. The findings advance understanding of credibility-technology interactions in medical crowdfunding information sharing, though the youth-focused sample limits generalizability. This integrated approach offers promising directions for optimizing campaign engagement strategies through balanced credibility signaling and enhanced platform usability.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-48330-8.

Introduction

Medical crowdfunding has emerged as a significant alternative for financing healthcare costs, providing individuals with online platforms to raise funds for medical expenses through donation-based campaigns¹. This digital approach expands the potential donor base beyond traditional networks to include both personal connections and altruistic strangers. A notable feature of these platforms is their interactive design, which permits fundraisers to craft personalized narratives incorporating multimedia elements that may elicit engagement and potentially encourage financial contributions². Despite these inherent platform characteristics, the success of medical crowdfunding campaigns depends not only on the platform design itself but also significantly on the strategic dissemination of campaigns through broader social networks.

The effectiveness of medical crowdfunding campaigns is often enhanced when information is shared through social media networks. These platforms expand campaign reach beyond immediate social circles, increasing the likelihood of attracting a larger donor base³. Research indicates that social information increased donations by 17% in crowdfunding contexts, particularly during early campaign stages⁴. However, despite the critical importance of information sharing for medical crowdfunding, many patients and campaign initiators face significant challenges in the dissemination process.

Several significant barriers impede effective dissemination of medical crowdfunding campaigns. First, privacy concerns and reluctance to disclose sensitive information present substantial challenges⁵, directly affecting the willingness of patients and their advocates to create detailed, compelling campaigns. Second, this reluctance can result in incomplete information that fails to resonate with potential donors, resulting in insufficient funds to meet urgent healthcare needs. Finally, the proliferation of misinformation on social media has eroded trust in online charity appeals, as individuals fear the consequences of sharing potentially misleading information, including diminished personal credibility and reputational harm⁶. This interplay of barriers significantly constrains the reach and effectiveness of medical crowdfunding, ultimately affecting healthcare access for those most in need.

Despite the critical importance of information sharing for medical crowdfunding success, existing research has primarily examined content characteristics and social dimensions of sharing behavior in isolation^7–9. However, a comprehensive theoretical framework that simultaneously incorporates both information credibility assessment and technology acceptance factors in explaining medical crowdfunding sharing behavior remains absent from the literature. This theoretical gap is particularly significant because sharing decisions in medical crowdfunding contexts involve a dual evaluation process where potential sharers must simultaneously assess campaign legitimacy and evaluate platform functionality.

Given the complexity of this dual evaluation process, we propose an integrated theoretical approach to address this gap. This framework captures both the informational and technological dimensions of medical crowdfunding sharing behavior. We integrate Source Credibility Theory (SCT) and Technology Acceptance Model (TAM) as complementary theoretical frameworks. SCT emphasizes source expertise and trustworthiness in shaping audience perceptions¹⁰, providing a foundation for understanding how individuals evaluate medical crowdfunding campaign credibility. Concurrently, TAM highlights the impact of perceived usefulness and ease of use on technology adoption¹¹, directly relating to user interaction with crowdfunding platforms. This integration creates a novel theoretical synthesis that transcends single-theory limitations by recognizing that sharing decisions in medical crowdfunding contexts are influenced simultaneously by both perceived campaign information credibility and platform technological characteristics.

To validate our proposed framework, we conducted a cross-sectional survey among young social media users in China and analyzed data using Partial Least Squares Structural Equation Modeling (PLS-SEM). The empirical results not only support our integrated model but also contribute to the medical crowdfunding literature in three ways. First, we demonstrate how informational and technological factors interact to shape sharing attitudes, extending beyond what SCT or TAM could explain independently. Second, we identify specific design considerations for platform developers, including perceived expertise, trustworthiness, usefulness, and ease-of-use. Third, we provide evidence-based insights for campaign creators to develop content that balances credibility and shareability, potentially addressing the gap between patient needs and fundraising outcomes.

The remainder of this paper is structured as follows: Section “Literature review and theoretical background” reviews the theoretical foundations of SCT and TAM, emphasizing their relevance to medical crowdfunding. Section “Research model and hypotheses” presents our research framework and hypotheses. Section “Methodology” details our methodology, including measurement tools and data collection methods. Sections “Data analysis” and “Discussion and implications” present our empirical findings and discuss both the interpretations of key relationships in our model and their theoretical and practical implications. Finally, Section “Conclusions” concludes with key contributions, limitations, and future research directions.

Literature review and theoretical background

Medical crowdfunding

Medical crowdfunding enables individuals to raise funds for medical expenses through online platforms^12,13. These campaigns leverage social media to reach broader audiences, with success significantly influenced by information sharing on social networks^14,15. Campaign visibility through repeated sharing directly impacts donation volume³. However, misinformation and fraudulent activities have raised credibility concerns, undermining platform trust¹⁶. Therefore, individuals harbor negative attitudes toward sharing information due to concerns about potential repercussions of disseminating misleading content, including damage to their personal credibility and reputation⁶.

Research reveals both emotional and rational factors drive medical crowdfunding information sharing, creating a complex motivational framework that determines sharing behavior. Affective frameworks highlight emotional drivers, with Tellis, et al.¹⁷ finding that emotional appeals in narratives generate higher sharing rates than informational content. This emotional-cognitive distinction echoes broader patterns in viral content research, where Berger and Milkman¹⁸ established that content evoking high-arousal positive emotions (awe, excitement) was shared up to 30% more frequently than neutral informational content. Within this dual framework, cognitive assessment particularly manifests in users’ trust evaluations, which are a prerequisite for sharing, facilitated through content quality, initiator credibility, and platform reputation⁶. These trust mechanisms work alongside attribution perceptions, emotional responses, and social capital to collectively shape sharing decisions⁷, ultimately functioning within social ecosystems where sharer credibility and connection strength significantly influence information dissemination⁸.

Keshavarz¹⁹ demonstrated that perceived credibility of social media information is influenced by multiple factors including participant characteristics, information source types, and content categories. Extending this information sharing to the medical crowdfunding domain, our investigation examines how cognitive determinants shape information sharing attitudes, particularly focusing on source reliability and technology acceptance dimensions. This research contributes to literature by analyzing how these factors collectively influence users’ propensity to disseminate medical crowdfunding campaigns, offering both theoretical insights into cognitive mechanisms underlying sharing behaviors and practical implications for designing healthcare fundraising campaigns that effectively optimize information dissemination.

Source credibility theory

Source Credibility Theory (SCT) posits that message effectiveness is shaped by source credibility, primarily determined by expertise and trustworthiness¹⁰. Expertise refers to the perceived knowledge and competence of the source, while trustworthiness reflects the belief that the source is honest and unbiased²⁰. Messages from highly credible sources are more persuasive and elicit stronger cognitive and attitudinal responses^21,22.

In medical crowdfunding, trust in information sources plays a pivotal role. Since campaigns are typically initiated by individuals rather than organizations⁷, their perceived legitimacy depends heavily on authenticity and transparency. Endorsements from medical professionals or reputable charities enhance credibility²³, while compelling narratives and clear campaign descriptions attract more support²⁴.

Technology acceptance model (TAM)

The Technology Acceptance Model (TAM) explains technology adoption through perceived usefulness and perceived ease of use¹¹. Perceived usefulness refers to the extent a system enhances user performance, while perceived ease of use reflects the effort required to operate it²⁵. TAM has extensively explored how people view different technologies and what influences their acceptance or rejection^25,26.

Perceived usefulness and perceived ease of use significantly influence attitudes, as users more likely adopt and share information through technologies they find beneficial^27,28. Users more likely view a platform as beneficial if they have faith in its validity, security protocols, and transparency²⁹. Regarding medical crowdfunding, users’ perceptions of the platform’s usefulness for promoting and supporting fundraising initiatives are influenced by their faith in the campaign.

Integrating SCT and TAM

The integration of SCT and TAM offers a complementary framework for examining medical crowdfunding sharing behavior. SCT addresses source credibility dimensions through expertise and trustworthiness—potentially important considerations in contexts where fraudulent campaigns have been reported—while TAM encompasses technological aspects through perceived usefulness and ease of use.

The relationship between source credibility and perceived usefulness operates through a cognitive processing mechanism. When potential sharers perceive information as originating from an expert source, they are more likely to believe that online public engagement will effectively enhance the perceived benefits for recipients³⁰. This occurs because expertise signals high-quality information, leading individuals to trust that the campaign represents a legitimate need worthy of support³¹. Similarly, when campaign sources are perceived as trustworthy, potential sharers develop confidence that their sharing behavior will not result in negative consequences, which increases their perception of sharing’s instrumental value³².

Prior research has established theoretical connections between credibility and adoption models. Pavlou and Gefen³³ demonstrated that institution-based trust mechanisms effectively mitigate perceived risk and facilitate adoption in online marketplaces, while Uche, et al.³⁴ integrated trust constructs with technology acceptance variables to develop a comprehensive model of consumer trust in electronic commerce. In healthcare contexts specifically, Lanseng and Andreassen³⁵ combined service quality dimensions (including credibility) with TAM to explain patients’ adoption of self-service health technologies, establishing that credibility perceptions significantly influence both perceived usefulness and usage intentions.

These integrative approaches align with findings suggesting source credibility influences perceived usefulness³⁶, while perceived usefulness mediates between information presentation and user attitude³⁷. In the donation-based crowdfunding domain, Um³⁸ found that perceived usefulness, ease of use, credibility significantly influenced potential backers’ willingness to engage with crowdfunding platforms. By building on established theoretical integrations, SCT-TAM integration provides a more comprehensive understanding of the psychological mechanisms through which credibility assessments shape technology acceptance in prosocial contexts.

Research model and hypotheses

Building on theoretical integration, our study addresses TAM’s limitations by incorporating SCT to examine medical crowdfunding sharing attitudes. While TAM lacks consideration of contextual factors^39,40, SCT integration enhances its explanatory power by accounting for credibility’s critical role. In our framework, SCT contributes expertise and trustworthiness as dimensions influencing credibility perceptions, which interact with perceived usefulness (a TAM construct) mediating the relationship between credibility and sharing attitudes, as illustrated in Fig. 1.

Fig. 1.

Research model.

The effect of expertise and trustworthiness on attitude

Expertise plays a critical role in enhancing users’ confidence in the information they encounter, primarily through credibility assessment mechanisms⁴¹. In general, sources perceived as knowledgeable or authoritative are considered more credible than peers or unfamiliar individuals on social media platforms^42,43. This distinction becomes particularly important in medical crowdfunding contexts, where credibility can significantly influence user engagement. Empirical evidence suggests that campaign success is often linked to indicators of perceived expertise, such as the patient’s occupation, the clarity of monetary needs, and the length and structure of narrative content⁴⁴. Well-structured descriptions that include highly credible communicators reduce cognitive load by simplifying information processing and reinforcing positive attitudes^45,46.

Moreover, trustworthy cues—such as endorsements from medical professionals, affiliations with reputable institutions, or detailed and verifiable narratives—serve as indicators that the information is credible and worthy of dissemination^47,48. These elements reduce users’ perceived uncertainty and enhance the perceived integrity and reliability of the message, thereby strengthening their willingness to engage with and share such content^49,50. Even when the sharer is not the original source, trust in the communicator can significantly increase the likelihood of dissemination, as credibility can be transferred through perceived sincerity and accuracy⁸. In medical crowdfunding specifically, both cognitive- and affective-based trust have been shown to influence donation willingness, suggesting that perceived trustworthiness plays a central role in shaping supportive behaviors⁵¹. Furthermore, emotionally compelling narratives—especially those accompanied by third-party testimonials or endorsements—can heighten perceived credibility and attract more supporters to the campaign⁵². Summarizing the above reasoning, this study proposes:

H1

Expertise is positively associated with attitudes toward sharing medical crowdfunding information.

H2

Trustworthiness is positively associated with attitudes toward sharing medical crowdfunding information.

The effect of perceived usefulness and perceived ease of use on attitude

Perceived usefulness is a key determinant in shaping users’ attitudes toward sharing medical crowdfunding information. When individuals perceive shared content as helpful, relevant, and capable of delivering tangible benefits, they are more likely to develop favorable attitudes toward disseminating such information. User-generated content that is clear, emotionally resonant, and action-oriented tends to enhance the perceived value of medical crowdfunding campaigns, thereby increasing users’ willingness to engage with and share them³⁷. In online health communities, this effect is particularly pronounced, as users often prioritize content they deem meaningful and beneficial, leading to more active and sustained participation^28,53. Furthermore, studies have shown that perceived usefulness, especially when accompanied by empathetic elements such as personal stories or expressions of gratitude, fosters stronger emotional connections and more positive attitudes toward donation behaviors⁵⁴. Thus, the perceived utility of crowdfunding information plays a pivotal role in shaping supportive attitudes toward information sharing.

The usability of medical crowdfunding platforms also significantly influences attitudes toward sharing campaign information. In the sharing economy, ease of use enhances user engagement and facilitates information exchange⁵⁵. User-friendly platforms promote trust and accessibility, thereby increasing the likelihood of technology-based systems⁵⁶. Perceived ease of use significantly influences users’ attitudes toward information sharing, as intuitive and accessible social media platforms foster greater engagement in such activities⁵⁷. Moreover, in medical crowdfunding contexts, perceived ease of use becomes crucial as it reduces the technical barriers that might otherwise prevent urgent information sharing. Intuitive platform interfaces further minimize cognitive load by eliminating the need to struggle with complex navigation or sharing procedures⁵⁸. This can ultimately lead to more positive attitudes toward disseminating medical funding appeals. Summarizing the above reasoning, this study proposes:

H3

Perceived usefulness is positively associated with attitudes toward sharing medical crowdfunding information.

H4

Perceived ease of use is positively associated with attitudes toward sharing medical crowdfunding information.

Mediating effects of perceived usefulness

Expertise and trustworthiness represent fundamental dimensions of credibility that significantly influence how users evaluate online information utility. Users perceive high-quality information as more useful because it demonstrates relevance, accuracy, and applicability to their specific needs⁵⁹. Expert assessments also function as credibility indicators, enhancing perceived information usefulness particularly where authority is valued. Information quality directly impacts perceived usefulness in information processing path⁶⁰. In online health information sharing, illustrated articles with more detailed health information is more frequently perceived as useful by readers, ultimately shaping users’ recognition⁶¹. When users encounter expert-verified or professionally endorsed medical crowdfunding campaigns, they are more likely to perceive the information as useful for themselves and for others in their network. This perceived usefulness, in turn, positively shapes their attitudes toward sharing such campaigns.

User assessment of information usefulness is predominantly influenced by their perception of credibility and trustworthiness. In health-focused online communities, the credibility of user-generated content is often strengthened through mechanisms such as detailed user profiles, community rating systems, and expert moderation, all of which collectively enhance trust and elevate the perceived value of the information³². During crises like the COVID-19 pandemic, the importance of information trustworthiness and quality became especially pronounced, significantly affecting online engagement patterns. Users tend to prioritize and actively engage with content that is perceived as both credible and beneficial, highlighting trust as a critical antecedent to perceived information usefulness and subsequent public participation online⁶². In the context of medical crowdfunding, campaigns that are transparent, well-documented, and supported by verifiable evidence are generally evaluated as more reliable and valuable by potential donors and sharers⁵². Thus, trustworthiness acts as a foundational factor that shapes how useful users find the information, which in turn drives their attitudes to share medical crowdfunding messages.

H5

Expertise indirectly influences attitudes toward sharing medical crowdfunding information through perceived usefulness.

H6

Trustworthiness indirectly influences attitudes toward sharing medical crowdfunding information through perceived usefulness.

Methodology

Data collection and sampling strategy

This study targeted young social media users as the primary research population. This demographic was selected because young users typically employ social media as their predominant communication medium, demonstrate considerable proficiency in digital content creation, and regularly participate in online activism⁶³. Their established digital presence and active sharing practices render them significant for understanding factors that may enhance campaign visibility and potentially influence fundraising outcomes in medical crowdfunding contexts.

Data collection occurred between September and October 2024 following approval from the Ethics Committee for Research Involving Human Subjects (JKEUPM). We employed a cross-sectional survey approach with two convenience sampling strategies: (1) targeted invitations in specific online communities (QQ groups and WeChat groups), with initial participants encouraged to share the survey with eligible peers; and (2) voluntary participation through invitation links posted on Weibo and WeChat Moments, which facilitated further distribution as users shared these links within their networks.

The survey was distributed via the Wenjuanxing platform, with participants offered a 5 RMB incentive upon completion. To ensure data quality, we implemented several quality control measures: (1) attention check questions (e.g., “To show you are reading carefully, please select ‘somewhat agree’ for this item”) were embedded throughout the survey to identify careless respondents, (2) IP address tracking was used to prevent duplicate responses from the same participant, and (3) completion time monitoring flagged submissions completed in unreasonably short durations (average completion time was approximately 4 min, with responses completed in less than 2 min flagged for review). From 436 initial responses, we excluded 91 invalid submissions (incomplete questionnaires, straight-lining patterns, or unreasonably fast completion times), yielding 345 valid responses (79.1% valid response rate).

We acknowledge that convenience sampling may introduce selection bias, potentially overrepresenting digitally active individuals while underrepresenting those with lower online engagement. Additionally, the youth-focused sample limits generalizability across age demographics. These limitations should be addressed in future research through more diverse sampling methods and broader demographic representation.

Measurement instrument

We adapted validated scales from prior research to construct our measurement tool. Expertise and trustworthiness measures were based on frameworks by Ayeh, et al.⁶⁴ and Qiu, et al.⁶⁵. Perceived usefulness and perceived ease of use were adapted from Qiu, et al.⁶⁵ and Yin, et al.⁶⁶, while attitude items derived from Ayeh, et al.⁶⁴. All items employed a six-point Likert scale (1 = “strongly disagree” to 6 = “strongly agree”).

Two communication specialists reviewed the survey items to ensure content validity. Following their feedback, several semantically redundant or ambiguous items were removed, and minor wording adjustments were made. A pilot test with 50 participants was then conducted to assess face validity. Based on the pilot results, no substantial additional revisions were necessary as all items demonstrated clear comprehension among participants. Internal consistency measured via Cronbach’s alpha with this pilot sample ranged from 0.723 to 0.775, exceeding the 0.70 threshold⁶⁷ and indicating satisfactory reliability. Detailed questionnaire items are provided in Appendix A. Raw survey data are available in Supplementary File.

Data screening and validation

We conducted descriptive statistical analyses using SPSS to characterize demographic attributes and primary variables. For hypothesis testing, we employed PLS-SEM via SmartPLS 4.0, selected for its robustness with complex models and emphasis on variance explanation and predictive relevance⁶⁸.

Normality assessment

Our data presented a statistical paradox regarding normality. While the skewness (− 0.265 to 0.009) and kurtosis (− 0.572 to − 0.254) values for all five study variables fell within conventionally acceptable ranges ± 2 suggesting normal distribution⁶⁹ shown as Table 1, the Kolmogorov–Smirnov test yielded significant results (p < 0.05) indicating departures from normality⁷⁰. This discrepancy can be attributed to the K-S test’s high sensitivity to sample size, where even minor deviations in large datasets can produce significant p-values⁷⁰. Following methodological recommendations for handling such conflicting normality indicators, we proceeded with PLS-SEM analysis using SmartPLS 4.0 software. This methodological selection is justified because PLS-SEM does not necessitate strict adherence to normality assumptions⁶⁸. PLS-SEM offers robust parameter estimation through its distribution-free iterative algorithm, making it appropriate for our research context where the primary objective was predictive relevance rather than theory confirmation⁷¹. Moreover, PLS-SEM’s capability to handle complex structural models with both reflective and formative constructs further supported our analytical approach⁷².

Table 1.

Skewness and kurtosis values of study variables.

Variable	Skewness	Kurtosis
Expertise	− 0.265	− 0.572
Trustworthiness	0.009	− 0.465
Perceived usefulness	− 0.154	− 0.510
Perceived ease of use	− 0.186	− 0.472
Attitude	− 0.142	− 0.254

Common method bias assessment

To minimize the potential impact of common method bias (CMB) in this cross-sectional study, we implemented both procedural and statistical remedies. Procedurally, all measurement items were adapted from validated scales in existing literature and underwent pre-testing to ensure reliability and clarity. Questionnaire wording was carefully refined to be concise, specific, and unambiguous, reducing the risk of misinterpretation. Respondents were explicitly assured of anonymity and confidentiality, with these assurances prominently displayed on the survey interface.

Statistically, we conducted two tests to assess CMB. Harman’s single-factor test revealed the first factor accounted for 44.82% of total variance, below the 50% threshold, suggesting CMB was not significant⁷³. Additionally, the full collinearity test yielded Variance Inflation Factor values for all constructs ranging from 1.38 to 1.84, well below the recommended 3.3 threshold⁷⁴. These results provide evidence that CMB did not substantially affect our findings.

Data analysis

In accordance with the methods suggested by Agnew, et al.⁷⁵ a two-step structural equation modeling (SEM) approach was employed to examine the data. This method entailed evaluating the relationships in the structural model and the quality of the measurement model one after the other. SEM was selected for its ability to analyze linear equations involving both observable and latent variables while accommodating complex dependencies within the research framework⁷⁶. Using SmartPLS, we first established measurement model validity and reliability before proceeding to examine the structural pathways between constructs.

Demographic results

The participants’ demographic and socioeconomic details are displayed in Table 2 (n = 345). The sample is predominantly male, composed mainly of young adults aged 20–29. Most participants have attained junior college or undergraduate education, while income levels are concentrated in the lower to middle ranges, with the majority earning below 4000 RMB. This profile reflects a youthful, moderately educated, and economically diverse population.

Table 2.

Participants’ sociodemographic attributes (n = 345).

Item	Content	Frequency	Percentage (%)
Gender	Male	201	58.26
	Female	144	41.74
Age	15–19	62	17.97
	20–29	283	82.03
Education	Below secondary vocational school	54	15.65
	Junior college	130	37.68
	Undergraduate	139	40.29
	Postgraduate	22	6.38
Income	Below 2000 RMB	57	16.52
	2001 to 4000 RMB	115	33.33
	4001 to 6000 RMB	95	27.54
	6001 to 8000 RMB	48	13.91
	Above 8000 RMB	30	8.70

Measurement models

Prior to finalizing the measurement model, we conducted a thorough assessment of all questionnaire items. Following established methodological guidelines⁷⁷, items were removed based on two criteria: (1) factor loadings below the recommended threshold of 0.6, and (2) negative impact on construct reliability and validity. This process resulted in the elimination of eight items across the constructs of Expertise, Trustworthiness, Perceived Usefulness, Perceived Ease of Use, and Attitude. The removal of these problematic items significantly improved the internal consistency of constructs, with Cronbach’s alpha values ranging from 0.723 to 0.775, all exceeding the recommended threshold of 0.7.

As shown in Table 3, all remaining measurement indicators met the recommended thresholds: factor loadings exceeded 0.6, AVE values were above 0.5, and CR values surpassed 0.7^68,77, demonstrating the constructs’ reliability and convergent validity. Additionally, Table 3 shows the outer model’s variance inflation factor (VIF) values, all of which are below 3.3⁷⁸, indicating that there are no collinearity problems. Furthermore, to establish discriminant validity, we examined the heterotrait-monotrait ratio (HTMT) values for all constructs, reported in Table 4, with all scores below the 0.9 cutoff⁷⁹, confirming the discriminant validity of our measurement model.

Table 3.

Convergent validity, composite reliability and collinearity assessments.

Item	Loading	CR	AVE	VIF
Expertise		0.868	0.687
Exp1	0.843			1.846
Exp2	0.818			1.779
Exp3	0.826			1.400
Trustworthiness		0.844	0.643
Trus1	0.768			1.398
Trus2	0.809			1.382
Trus3	0.828			1.521
Perceived usefulness		0.859	0.670
PerUse1	0.763			1.409
PerUse2	0.835			1.575
PerUse3	0.855			1.630
Perceived ease of use		0.856	0.665
PerEas1	0.814			1.480
PerEas2	0.853			1.643
PerEas3	0.778			1.431
Attitude			0.861	0.674
Att1	0.839			1.606
Att2	0.796			1.466
Att3	0.827			1.535

Table 4.

Discriminant validity assessment.

	Exp	Trus	PerUse	PerEas	Att
Exp
Trus	0.819
PerUse	0.795	0.818
PerEas	0.793	0.738	0.855
Att	0.702	0.754	0.849	0.875

Exp = Expertise; Trus = Trustworthiness; PerUse = Perceived Usefulness; PerEas = Perceived ease of use; Att = Attitude;

Assessment of the structural model

After validating the measurement model, we proceeded to assess the structural model to examine the hypothesized relationships among the constructs. Figure 2 presents a visual representation of our structural model, displaying the path coefficients between constructs and the R² values for endogenous variables. The structural model assessment consisted of evaluating the model fit, testing the hypothesized direct relationships, analyzing the mediating effects, and examining the explained variance and predictive relevance of the model.

Fig. 2.

Structural model assessment results.

Model fit evaluation

The SRMR value measures the discrepancy between the observed and predicted correlation matrices. In this study, the SRMR values were 0.071 for the saturated model and 0.080 for the estimated model, both of which are at or below the recommended threshold of 0.08. These results suggest an acceptable model fit, indicating that the structural model sufficiently explains the relationships among the constructs⁸⁰.

Direct path analysis

All of the inner model’s VIF values fall below 3.3, indicating that there are no collinearity issues and that the model is reliable⁷⁴. Table 5 presents the direct path analysis results examining the influence of source credibility (Expertise, Trustworthiness) and technology acceptance factors (Perceived Usefulness, Perceived Ease of Use) on attitude toward sharing medical crowdfunding information.

Table 5.

Direct path analysis results.

Relationship	β	p	VIF	f2	Hypothesis	Decision
Exp→Att	0.056	0.275	2.064	0.003	H1	Not supported
Trus→Att	0.163	0.001	1.899	0.031	H2	Supported
PerUse→Att	0.291	0.000	2.155	0.086	H3	Supported
PerEas→Att	0.351	0.000	1.978	0.136	H4	Supported

Exp = Expertise; Trus = Trustworthiness; PerUse = Perceived usefulness; PerEas = Perceived ease of use; Att = Attitude; VIF = Variance inflation factor; β = standardized path coefficient; p = significance level; f² = effect size measure (proportion of variance explained); According to Cohen⁸¹, f² values of 0.02, 0.15, and 0.35 represent small, medium, and large effect sizes, respectively.

The results indicate that expertise does not have a significant direct effect on attitudes (β = 0.056, p = 0.275, f² = 0.004), suggesting that expertise alone may not directly shape attitudes toward sharing behavior and rejecting H1. The f² value of 0.004 indicates an effect size below the threshold for a small effect, suggesting that expertise contributes negligibly to explaining variance in sharing attitudes when other factors are present. In contrast, trustworthiness exhibits a significant positive effect on attitudes (β = 0.163, p = 0.001, f² = 0.032), supporting H2. This f² value represents a small effect size, representing a practically meaningful though small contribution to understanding sharing behavior.

Perceived usefulness and perceived ease of use both significantly predict attitudes toward sharing behavior, supporting H3 (β = 0.291, p < 0.001, f² = 0.098) and H4 (β = 0.351, p < 0.001, f² = 0.147), respectively. The f² values reveal that perceived usefulness has a small effect, accounting for nearly 10% of the variance and indicating substantial practical importance for platform design. Perceived ease of use demonstrates a medium effect size, which signifies a significant contribution to explaining sharing attitudes, accounting for approximately 14.7% of the variance and establishing it as the most practically important predictor for encouraging medical crowdfunding information sharing.

Mediation analysis

To test for mediation effects, we employed the bootstrapping procedure in SmartPLS with 10,000 resamples. This approach enables robust estimation of direct, indirect, and total effects, along with their significance levels^68,82. Additionally, we calculated the Variance Accounted For (VAF) to determine the proportion of the indirect effect relative to the total effect. We followed Zhao, et al.⁸² classification to identify the type of mediation present in our model. Table 6 presents the results of our mediation analysis, showing the direct, indirect, and total effects for the hypothesized mediation relationships.

Table 6.

Results of mediation analysis with perceived usefulness as mediator.

Relationship	Direct effect		Indirect effect			Total Effect		VAF	Mediation type
	β	p	β	p	95% CI	β	p	%
Exp→Att	0.056	0.275	0.114	<0.001	[0.070, 0.163]	0.169	0.001	67.5	Indirect-only mediation
Trus→Att	0.163	0.001	0.107	<0.001	[0.066, 0.159]	0.270	<0.001	39.6	Complementary mediation

Exp = Expertise; Trus = Trustworthiness; Att = Attitude; VAF = Variance accounted for (indirect effect/total effect).

The results reveal two distinct mediation patterns in our model. First, the relationship between expertise and attitude demonstrates an indirect-only mediation pattern⁸². The direct effect of expertise on attitude was non-significant (β = 0.056, p = 0.275), while the indirect effect through perceived usefulness was significant (β = 0.114, p < 0.001, 95% CI [0.070, 0.163]). The total effect was also significant (β = 0.169, p = 0.001). The VAF value was calculated as 67.5% (0.114/0.169), indicating that 67.5% of the total effect of expertise on attitude is explained by the indirect effect through perceived usefulness. The bootstrap bias-corrected confidence interval for the indirect effect does not include zero, confirming the significance of the mediation effect. This finding suggests that expertise influences attitude primarily through perceived usefulness.

For the relationship between trustworthiness and attitude, both the direct effect (β = 0.163, p = 0.001) and the indirect effect through perceived usefulness (β = 0.107, p < 0.001, 95% CI [0.066, 0.159]) were significant. The total effect was also significant (β = 0.270, p < 0.001). The VAF value was calculated as 39.6% (0.107/0.270), indicating that 39.6% of the total effect of trustworthiness on attitude is explained by the indirect effect through perceived usefulness. This pattern represents what Zhao, et al.⁸² call “complementary mediation”, suggesting that trustworthiness influences attitude both directly and indirectly through perceived usefulness, highlighting the dual pathways through which trustworthiness impacts attitude formation. The bootstrap bias-corrected confidence interval for the indirect effect confirms the robustness of this mediation relationship.

Explained variance (R²) and predictive relevance (Q² predict)

The R-square (R²) values, which show the percentage of variation accounted for by independent variables, and Q² predict, which evaluates the predictive accuracy of the model, are shown in Table 7. The R² values for the endogenous constructs—0.543 for attitude (Att) and 0.464 for perceived usefulness (PerUse)—exceed the widely accepted threshold of 0.2⁸³, demonstrating strong in-sample explanatory power. Additionally, the Q²predict values, calculated using PLSpredict with a tenfold cross-validation approach, are 0.489 for attitude and 0.458 for perceived usefulness. As both values are well above zero, the model exhibits substantial out-of-sample predictive relevance⁸⁴. These results validate the model’s robustness in both explaining and predicting the target constructs, highlighting its reliability and applicability within the study’s context.

Table 7.

In-sample model fit and out-of-sample predictive power.

	R-square	Q2 predict
Attitude	0.543	0.489
Perceived usefulness	0.464	0.458

Discussion and implications

Discussion

This section examines the key findings of our study on information sharing attitudes. We analyze how expertise and trustworthiness influence sharing attitudes, explore the role of perceived usefulness and usability as motivating factors, and investigate how perceived usefulness mediates between expertise/trustworthiness and attitudes.

Expertise and trustworthiness in information sharing attitudes

This study finds that expertise alone does not directly influence sharing attitudes in medical crowdfunding contexts. The non-significant direct relationship between expertise and sharing attitudes can be explained through cognitive processing mechanisms. Specialized terminology and concepts may create cognitive overload for average social media users unfamiliar with healthcare jargon⁸⁵. This aligns with dual-process theories of persuasion⁸⁶, which suggest that in contexts where detailed processing is challenging, peripheral cues such as trustworthiness and emotional resonance become more influential than central cues like expertise. Furthermore, this finding may reflect eHealth literacy disparities among social media users, as individuals with limited health literacy may struggle to evaluate expertise indicators in health information⁸⁷. This might lead them to rely on simpler heuristics rather than systematic processing of expert content.

In contrast, our findings affirm the critical role of trustworthiness in directly shaping attitudes toward sharing medical crowdfunding information. Consistent with previous research^47,48, our results demonstrate that trustworthiness serves as a significant predictor of attitudes toward sharing behavior, where perceived sincerity directly influences campaign amplification. While Zhang, et al.⁵¹ identified trust as contributing to donation willingness, our research extends this by revealing how trustworthiness influences the pre-donation phase of information sharing, effectively serving as a gatekeeper for campaign visibility.

Perceived usefulness and usability in information sharing attitudes

This study’s findings also reinforce the significance of perceived usefulness and perceived ease of use in shaping individuals’ attitudes toward sharing medical crowdfunding information. Campaigns that effectively communicate clear, impactful, and actionable messages enhance their perceived value, increasing engagement³⁷. The positive relationship between perceived usefulness and sharing attitudes aligns with existing literature, demonstrating that users are more likely to share content when they perceive it as meaningful and beneficial²⁷. This pattern is also observed in other healthcare-related online communities, where users who recognize tangible benefits tend to engage more actively in information exchange^28,53. Additionally, our findings resonate with Chen, et al.⁵⁴, who highlight that perceived usefulness significantly influences donation attitudes in medical crowdfunding.

Furthermore, perceived ease of use plays a pivotal role in motivating individuals to share medical crowdfunding campaigns. These results align with prior research emphasizing the role of usability in facilitating engagement within the sharing economy⁵⁵. When medical crowdfunding platforms are intuitive and user-friendly, individuals experience fewer barriers to interaction, leading to increased willingness to disseminate campaign information. This observation is consistent with Elkaseh, et al.⁵⁷, who argue that user-friendly digital platforms foster participation, enhancing the likelihood of information sharing.

Perceived usefulness as a mediator of expertise and trustworthiness effects

Expertise operates through indirect pathways rather than direct influence on sharing attitudes. While expertise does not directly influence sharing attitudes, it remains relevant by operating through the mediating mechanism of perceived usefulness. This suggests a more nuanced role than previously recognized, challenging assumptions about direct expertise effects in information sharing contexts. The expertise-usefulness pathway also reveals a cognitive evaluation mechanism in medical crowdfunding. Building on Machdar⁵⁹ finding that high-quality information is perceived as useful due to its relevance and applicability. Our results show that expertise embedded in campaign content enhances perceived usefulness, which subsequently shapes positive sharing attitudes. In medical crowdfunding specifically, expertise signals influence sharing through this cognitive pathway where perceived usefulness serves as the critical mediator.

Trustworthiness demonstrates dual pathways of influence on sharing attitudes. Unlike expertise, trustworthiness not only directly affects sharing attitudes but also indirectly influences them through perceived usefulness. This dual mechanism suggests that trustworthiness operates both as an immediate attitude driver and as a precursor to usefulness evaluations. The trustworthiness-usefulness mediation pathway extends existing theoretical understanding. This aligns with Hajli, et al.³² observation that in health-focused communities, credibility enhances both trust and perceived value, while extending beyond their general framework to specify the mediating role of usefulness.

These mediation patterns have important implications for campaign design and platform development. Our results also extend Zhao and Mao⁵² by demonstrating that transparent campaign evaluation involves a specific cognitive pathway, where perceived usefulness mediates the relationship between trustworthiness and sharing attitudes in medical crowdfunding contexts. This finding suggests that both expertise indicators and trustworthiness signals should be designed to enhance users’ perceptions of sharing utility rather than merely displaying credibility markers.

Theoretical contributions

This study contributes to theory by proposing an integrated framework that combines elements from SCT and TAM in the medical crowdfunding context. Our research identifies specific determinants influencing attitudes toward sharing medical crowdfunding information, offering insights into digital health communication behaviors. The integration of SCT and TAM provides a theoretical perspective that considers both content credibility and technological usability factors, illuminating how these elements interact to shape sharing decisions. Unlike previous SCT-TAM integrations that primarily focused on technology adoption or e-commerce trust (e.g., Pavlou and Gefen³³ ; Uche, et al.³⁴), our model specifically addresses prosocial information sharing behavior in health contexts, where credibility assessment becomes critical due to potential misinformation risks. Our findings suggest that source credibility dimensions and technological features work through distinct but complementary mechanisms, with perceived usefulness serving as a mediating factor between credibility assessments and sharing attitudes. It offers a foundation for understanding the complex interplay between trust considerations and user experience in health-related information sharing contexts.

Practical contributions

Our findings yield actionable insights for medical crowdfunding stakeholders across multiple dimensions. First, campaign initiators should translate complex terminology into accessible language and incorporating explanatory visual elements, thereby balancing expertise with comprehensibility. Second, implementation of verification badges for validated campaigns, standardized financial disclosure templates, systematic progress updates, and third-party testimonials could enhance trustworthiness in the crowdfunding ecosystem. Third, campaign creators and platforms should collaboratively demonstrate the tangible value of sharing campaigns by clearly illustrating how donations translate to specific interventions, quantifying impact through visualization features, and documenting actual medical benefits achieved, consequently reinforcing the perception of usefulness that motivates future information sharing behaviors. Fourth, platforms can improve user experience by simplifying sharing processes through one-click multi-platform distribution mechanisms, thus optimizing perceived platform usability and expanding campaign reach.

Conclusions

Our analysis examines the factors influencing attitudes toward sharing medical crowdfunding information on social media platforms. The findings suggest that trustworthiness, perceived usefulness, and perceived ease of use directly shape sharing attitudes, while trustworthiness additionally exerts indirect effects, and expertise influences attitudes only through the mediating role of perceived usefulness. By integrating the Technology Acceptance Model with Source Credibility Theory, this empirically validated relationships not only advance our theoretical understanding of information-sharing behavior in digital healthcare contexts but also establishes a foundation for examining how trust mechanisms in emerging AI-facilitated platforms may significantly reshape technology acceptance processes across diverse digital ecosystems.

This study has several limitations that should be considered when interpreting the results. The exploratory nature of our hypotheses reflects the limited empirical research in this specific domain. Our focus on attitudinal rather than behavioral measures creates a gap that future research should address. The framework primarily emphasizes cognitive factors while underrepresenting emotional and social dimensions. Additionally, the cross-sectional design limits causal inferences, and the absence of control variables may obscure potential confounding effects. Furthermore, our convenience sampling approach affects external validity by restricting generalizability to broader populations with diverse technology adoption levels and privacy concerns, while sample homogeneity may influence structural path sensitivity through range restriction and potentially inflated correlations between constructs. Future research would benefit from: (1) longitudinal designs tracking actual sharing behaviors, (2) inclusion of emotional and social variables, (3) experimental approaches testing causal relationships, (4) testing whether demographic factors such as gender and education moderate relationships or create subgroup differences, and (5) probability sampling methods across diverse demographic groups to validate structural relationship robustness and enhance generalizability.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Y.C. and S.K. made substantial contributions to the conception and design of the study, data acquisition, analysis, and interpretation. Both authors were involved in drafting the manuscript and making critical revisions for important intellectual content. All reviewed and approved the submitted version, as well as any significantly revised versions incorporating substantial modifications to their contributions.

Data availability

The data used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Ethics Committee for Research Involving Human Subjects (JKEUPM) under approval reference number JKEUPM-2024-302. All participants voluntarily provided informed consent prior to participation. Given the anonymous and minimal-risk nature of the survey, only participants’ self-consent was obtained.