Abstract
Expanding social insurance coverage to informal sector workers remains a critical challenge in emerging economies. This study advances theoretical understanding by proposing and empirically validating an innovative integrated framework that synthesises the Theory of Planned Behaviour (TPB) and Self-Determination Theory (SDT), addressing the limitations of single-theory approaches that inadequately explain the intention-behaviour gap in voluntary social insurance (VSI) participation. Whilst TPB elucidates cognitive determinants of intention formation, SDT enriches the model by capturing motivational quality underpinning sustained behaviour, representing a novel theoretical integration previously unexplored in informal sector financial protection contexts. Through structural equation modelling analysis of survey data from 800 informal workers across Vietnam’s six regions, this research empirically confirms traditional TPB relationships. It establishes intrinsic motivation as a direct predictor and crucial mediator between risk perception and intention—a previously unidentified psychological mechanism. Income demonstrates a paradoxical dual role, negatively affecting direct participation whilst positively moderating the intention-behaviour relationship, revealing how material constraints interact with psychological readiness in complex ways. The integrated model exhibits substantial explanatory power for both intention (R2 = .639) and participation (R2 = .564), significantly exceeding single-theory applications and substantially extending current theoretical understanding of voluntary financial behaviours. These findings demonstrate how the integration of cognitive and motivational perspectives creates synergistic theoretical advancement, offering valuable implications for policymakers seeking to bridge the alarming gap between Vietnam’s ambitious social protection targets (60% labour force participation by 2030) and current empirical reality (3.92% VSI participation), whilst providing insights for similar emerging economies confronting informal sector coverage challenges.
Plain language summary
Most informal workers in Vietnam lack social insurance, leaving them financially vulnerable in old age. Participation rates remain low despite government efforts to provide voluntary social insurance (VSI). This study investigates why people join or avoid VSI programs by looking at thinking patterns and motivational factors.
We found that people are more likely to join VSI when they have positive attitudes about it, feel social support from others, and believe they can manage the enrolment process and payments. Importantly, people’s inner motivation—feeling that participation aligns with their values—strongly influences their intention to join and actual participation. When people perceive risks to their future security, this awareness increases their motivation to participate.
Interestingly, higher income doesn’t always mean greater participation. While wealthier informal workers may have alternative protection options, they can better act on their intentions to participate when they form them.
These findings suggest that policymakers should focus on providing information about VSI benefits and fostering personal motivation by connecting VSI to people’s values and life goals. Programs should be designed to support people’s sense of choice rather than simply pressuring them to join while also addressing income-related barriers to participation
Keywords
Introduction
Vietnam’s social protection system faces a dual challenge of systemic proportions that extends beyond traditional policy implementation difficulties. Whilst the government has set ambitious targets through Resolution No. 28-NQ/TW to achieve 60% labour force participation in social insurance by 2030, a stark reality emerges when examining current coverage rates. Despite significant policy support and financial investment, voluntary social insurance (VSI) participation among informal workers remains disappointingly low at merely 1.83 million individuals by the end of 2023, representing only 3.92% of the working-age population (Vietnam Social Security, 2024). This dramatic gap between policy aspirations and empirical reality becomes even more concerning when contextualised within Vietnam’s informal sector, which comprises approximately 33.3 million workers—64.9% of the total employed workforce—who remain largely excluded from mandatory social insurance schemes (General Statistics Office of Vietnam, 2023). This challenge reflects broader regional patterns, as recent evidence across ASEAN demonstrates that whilst financial inclusion initiatives have successfully promoted green growth and economic integration, they have not correspondingly addressed social protection coverage gaps among informal workers (Putra & Oktora, 2024). The regional experience suggests that mechanisms facilitating financial access for economic development operate through different pathways than those required for sustained social protection engagement.
The magnitude of this challenge is further amplified by concurrent trends within the formal sector, where an alarming pattern of early withdrawal from compulsory social insurance has emerged. Recent analysis reveals that over 960,000 workers opted for lump-sum withdrawals in 2021 alone—a 9% annual increase trend driven by precarious living conditions, institutional distrust, and immediate financial pressures (T. P. Nguyen, 2023). This parallel phenomenon suggests that the barriers to sustained social protection engagement transcend simple access issues, pointing to deeper psychological and structural mechanisms that current policy frameworks have not adequately addressed.
Previous research examining VSI participation has predominantly employed single theoretical frameworks, most commonly the Theory of Planned Behaviour (TPB), to identify determinants of enrolment intention (C. C. Nguyen et al., 2024; Vu, 2025). Whilst these studies have provided valuable insights into cognitive factors shaping intentions, they have insufficiently addressed the fundamental theoretical gap regarding motivational quality in voluntary financial behaviours—specifically, the distinction between acting from external pressure versus personal volition that determines sustained behavioural engagement. This theoretical limitation may partially explain the persistent intention-behaviour gap observed in practice, where expressed intentions to participate frequently fail to translate into actual, sustained enrolment.
Recent advances in behavioural theory integration, particularly the successful synthesis of TPB and Self-Determination Theory (SDT) in health contexts (Hagger & Chatzisarantis, 2009), have demonstrated enhanced explanatory power for understanding complex voluntary behaviours. However, this theoretical integration remains entirely unexplored in voluntary financial protection contexts, despite the unique characteristics of such decisions: immediate costs with delayed benefits, long-term commitment requirements, and operation within resource-constrained environments. Furthermore, whilst financial constraints are frequently cited as barriers to VSI participation, their potential role in moderating the intention-behaviour relationship—and the psychological mechanisms through which risk perceptions translate into sustained motivation—requires systematic empirical investigation.
This study addresses these substantive research gaps by developing and empirically validating the first integrated TPB-SDT framework to understand VSI participation among informal workers in emerging economy contexts. The research makes four primary contributions: (1) establishes the theoretical foundation for TPB-SDT integration in voluntary financial behaviours, demonstrating how motivational quality interacts with cognitive determinants; (2) empirically validates intrinsic motivation as both a direct predictor and critical mediator in VSI decision-making processes; (3) elucidates the complex dual role of income as both barrier and facilitator in intention-behaviour translation; and (4) provides evidence-based insights for policy interventions targeting the 96% coverage gap between current participation and national social security objectives.
By integrating these complementary theoretical perspectives within Vietnam’s informal sector context, this study advances theoretical understanding of voluntary financial behaviours whilst offering practical guidance for addressing one of emerging economies’ most pressing social protection challenges. The findings have broader implications for understanding voluntary financial decision-making in resource-constrained environments and may inform social protection expansion efforts across similar developing economy contexts.
Theoretical Framework and Hypotheses Development
Theory of Planned Behaviour: Foundations and Theoretical Limitations
The Theory of Planned Behaviour (TPB), conceptualised by Ajzen (1991), posits that behavioural intention—the immediate cognitive antecedent of behaviour—is determined by three psychological constructs: attitude toward the behaviour, subjective norms (perceived social pressure), and perceived behavioural control (perceived capability to execute the behaviour). Extensive empirical evidence supports TPB’s applicability to financial decision-making, with meta-analytic studies demonstrating that attitudinal factors consistently predict intentions across financial inclusion programmes (Musa et al., 2024), whilst subjective norms exhibit particular potency in collectivist cultural contexts where social conformity influences individual financial decisions (Hien et al., 2023). Perceived behavioural control is especially critical for resource-intensive financial behaviours, consistently predicting investment decisions and insurance participation across emerging economies (Khalid et al., 2022; Raut, 2020).
However, TPB exhibits fundamental limitations when applied to voluntary long-term financial behaviours such as VSI participation. First, the theory’s cognitive orientation emphasises intention formation mechanisms whilst providing insufficient insight into motivational dynamics sustaining behavioural engagement over time—particularly problematic for VSI participation requiring sustained commitment despite income volatility. Second, TPB’s unitary conceptualisation of intention fails to differentiate qualitatively different intentions based on their motivational foundations, implicitly assuming equivalent translation likelihood regardless of underlying psychological sources (Deci & Ryan, 2008). Third, TPB’s volitional control emphasis may inadequately capture the interplay between psychological readiness and material constraints characterising financial decision-making in resource-constrained environments.
Within VSI contexts, these limitations have been empirically observed. Vu’s (2025) comprehensive examination of Vietnamese informal workers demonstrated significant TPB effects (R2 = .412) yet revealed substantial unexplained variance and notable intention-behaviour gaps that traditional constructs could not address, suggesting additional theoretical mechanisms are required.
Based on an extensive theoretical foundation and empirical evidence, we propose:
Extended Cognitive-Contextual Determinants
Beyond core TPB constructs, research has identified additional factors enhancing explanatory power in emerging economy contexts where institutional and informational factors play crucial roles.
Risk perception—individuals’ subjective assessment of potential adverse outcomes and personal vulnerability—emerges as a fundamental insurance-seeking determinant. The theoretical rationale stems from protection motivation theory, positing that individuals engage in protective behaviours when perceiving significant threats and believing protective actions can effectively mitigate such threats. Recent evidence demonstrates that individuals more vulnerable to old-age poverty and health expenses exhibit significantly stronger insurance protection intentions (Parvathi & Paul, 2024). Vietnam’s demographic transition and informal sector income volatility amplify workers’ vulnerability awareness, potentially motivating VSI participation.
Trust in institutional frameworks represents a critical determinant, particularly in emerging economies where credibility may be questioned. Trust encompasses confidence in the government’s ability to honour long-term commitments, system administration transparency, and investment management competence. Vietnam’s context makes trust especially relevant, with concurrent early withdrawals from compulsory social insurance—exceeding 960,000 individuals in 2021—reflecting underlying credibility concerns likely extending to voluntary programmes (T. P. Nguyen, 2023).
Financial literacy—knowledge of financial concepts and products—enables informed decision-making by reducing information barriers. Extensive evidence demonstrates that higher financial literacy levels enhance intentions and actual adoption rates across emerging economies (Bayakhmetova et al., 2023; Hakim et al., 2023). VSI’s voluntary nature places greater responsibility on individuals to understand complex concepts, making literacy particularly relevant for intention formation and behavioural enactment.
Perceived benefits—subjective evaluation of expected advantages—constitute a critical determinant extending TPB’s attitudinal component by focusing on instrumental outcome expectations. Cross-national research provides substantial support, with comprehensive studies in Asian contexts demonstrating that benefit perceptions significantly predict participation intentions, often with effect sizes exceeding general attitudinal measures (Shen et al., 2023).
Consistent with TPB’s central proposition and extensive meta-analytic evidence, intention represents the most proximal cognitive determinant of behaviour, with robust relationships demonstrated across health, environmental, and financial decision contexts (Close et al., 2018).
Self-Determination Theory and Voluntary Financial Behaviour
Self-Determination Theory (SDT), developed by Deci and Ryan (2000), addresses TPB’s motivational quality limitations by distinguishing between autonomous motivation (acting from personal values and volition) and controlled motivation (acting from external pressure or compulsion). This distinction proves crucial for voluntary financial behaviours requiring sustained commitment, where autonomous motivation facilitates temporal preference for future security over immediate consumption—particularly relevant given VSI’s immediate costs with delayed benefits.
Extensive empirical evidence supports autonomous motivation’s superior predictive power for sustained behavioural engagement. Longitudinal studies consistently demonstrate that intrinsically motivated individuals adhere significantly more to long-term commitments than those with controlled motivation profiles (Steinhorst & Klöckner, 2017; Teixeira et al., 2012). For VSI participation, intrinsic motivation’s relevance stems from several mechanisms: facilitating prioritisation of future security, enabling resistance to periodic discontinuation temptations due to income volatility, and providing motivational sources when external validation is lacking.
The selection of intrinsic motivation as the core SDT construct is theoretically justified on three grounds: VSI participation in Vietnam lacks strong social norms activating controlled motivation; long-term pension planning aligns with intrinsic motivation’s characteristic persistence; and preliminary research indicated successful VSI participants describe motivation in autonomous rather than externally controlled terms.
However, SDT exhibits limitations when applied independently to financial contexts, providing less theoretical guidance regarding cognitive decision factors and potentially underestimating instrumental cost-benefit reasoning that characteristically dominates financial decisions (Gagné & Deci, 2005). Additionally, SDT frameworks typically lack mechanisms for incorporating structural constraints that significantly influence financial behaviours in economically diverse populations.
Theoretical Integration of TPB and SDT
The integration of TPB and SDT addresses each framework’s limitations whilst leveraging complementary strengths, proving particularly valuable for voluntary financial behaviours requiring cognitive decision-making and sustained motivational commitment. Previous TPB-SDT integration attempts have primarily occurred within health contexts. Hagger and Chatzisarantis (2009) demonstrated that autonomous motivation significantly influenced TPB constructs, enhancing model explanatory power by 15% to 20% compared to TPB-only approaches.
However, financial behaviours exhibit unique characteristics, distinguishing them from health behaviours, where integration has been validated. Financial protection decisions involve immediate costs with delayed probabilistic benefits, creating temporal asymmetry requiring cognitive evaluation and sustained motivation. Additionally, financial behaviours in informal contexts face substantial material constraints, potentially moderating psychological factor-behaviour relationships.
Table 1 provides a systematic comparison of these theories and illustrates how their integration addresses the limitations of each framework.
Comparative Analysis of TPB, SDT, and Their Integration.
Our proposed integration advances beyond previous approaches through three theoretical innovations. First, intrinsic motivation operates through dual pathways: as a direct predictor of intentions and behaviour, and as a mediator between cognitive factors and intentions. Second, we incorporate income as a moderator of intention-behaviour relationships, acknowledging that psychological readiness may be insufficient when structural barriers exist. Third, we propose specific mediating mechanisms through which risk perception influences motivation, thereby influencing intentions.
The theoretical foundation for intrinsic motivation mediating risk perception-intention relationships stems from SDT’s propositions regarding value internalisation. When individuals recognise financial vulnerabilities associated with ageing, this awareness may trigger internalisation processes whereby financial protection becomes integrated into personal goal structures rather than remaining external requirements (Ryan & Deci, 2000). Research supports this mechanism, demonstrating that risk consciousness positively influences intrinsic motivation when individuals perceive control over protective actions (Munuera et al., 2023).
Income as Moderator: Addressing the Resource Paradox
Material constraints can fundamentally alter relationships between psychological readiness and actual behaviour. Income represents a crucial constraint among informal workers experiencing low, irregular earnings that create direct participation barriers and indirect decision-making effects.
The relationship between income and voluntary insurance participation exhibits theoretical complexity extending beyond simple resource availability, creating the “income paradox.” From a resource perspective, higher income should facilitate participation by reducing the contribution burden and providing financial flexibility. However, an alternative perspective suggests higher income may reduce VSI necessity by providing access to superior alternative protection mechanisms—private savings, investments, or commercial insurance offering greater flexibility and potentially higher returns.
Recent empirical evidence supports both perspectives simultaneously, demonstrating that higher income is often associated with lower voluntary public insurance participation rates and strengthens intention-behaviour relationships when positive intentions exist (Vu, 2025). This moderation effect operates through several pathways: higher income reduces opportunity costs, provides greater financial stability, and enhances perceived behavioural control by increasing confidence in maintaining consistent contributions.
Beyond moderation effects, the alternative access hypothesis suggests higher-income informal workers may directly reduce VSI valuation due to access to superior alternatives, perceiving defined benefits as inadequate relative to lifestyle maintenance needs or restrictive compared to alternative financial strategies.
Integrated Theoretical Framework
The proposed integrated TPB-SDT framework synthesises cognitive, motivational, and structural determinants through a theoretically coherent model addressing single-theory limitations whilst leveraging complementary strengths. This integration advances theoretical understanding through several innovations: explicitly connecting cognitive decision formation processes with motivational quality sustaining behavioural engagement; proposing specific mediating mechanisms through which cognitive factors influence behaviour via motivational pathways; and incorporating structural constraints through moderation effects, acknowledging that psychological readiness may be insufficient when material barriers exist.
The comprehensive hypothesis structure reflects these theoretical integrations, encompassing core TPB relationships (H1–H3), extended cognitive-contextual factors (H4–H9), SDT integration including mediation effects (H10–-H12), and material constraint integration (H13–H14). This framework represents the first systematic attempt to synthesise TPB and SDT specifically for voluntary financial behaviour contexts, incorporating cognitive extensions and structural constraints characterising real-world decision-making environments (Figure 1).
Integrated theoretical framework for VSI participation.
The framework’s theoretical sophistication lies not merely in comprehensiveness, but in the specification of theoretically motivated relationships addressing fundamental questions about psychological and structural determinants of sustained voluntary financial protection behaviour. By combining cognitive decision-making mechanisms with motivational quality dynamics and acknowledging material constraints, this integrated approach provides enhanced explanatory power for understanding the complex phenomenon of voluntary social insurance participation among informal workers in emerging economy contexts.
Methodology
Research Design and Sampling Strategy
This study employed a cross-sectional, quantitative research design utilising a structured questionnaire to examine determinants of voluntary social insurance participation among informal workers in Vietnam. A multi-stage stratified random sampling approach was implemented to ensure representativeness across Vietnam’s diverse geographical and socioeconomic landscape, addressing concerns regarding the heterogeneity of the informal sector.
The selection of a sample size of 800 informal workers was based on four robust theoretical foundations, ensuring research strength and reliability. First, the sample size substantially exceeds the minimum theoretical requirement of 384, as determined by Cochran’s formula, providing over double the required size and significantly increasing estimation precision whilst reducing sampling error. Second, the sample adequately meets requirements for complex statistical analyses, particularly Partial Least Squares Structural Equation Modelling (PLS-SEM), which demands sufficient sample sizes for stable estimations when models include multiple mediating and moderating relationships. Third, post-hoc power analysis using G*Power software confirmed statistical power exceeding 0.99 with α = .05, substantially above the recommended 0.80 threshold, ensuring adequate capacity to detect essential effects whilst minimising Type II error risk. Fourth, the multi-stage stratified sampling method ensures high representativeness and generalisability through scientific stratification across six geographical regions and urban/rural categories within each region, with random participant selection at the final stage reducing selection bias and enhancing sample heterogeneity, reflecting the informal sector.
Initially, six regions were stratified (Northern Highlands, Red River Delta, North Central, South Central, Central Highlands, and Mekong Delta), with two provinces systematically selected from each region representing varying urbanisation levels. Within each province, districts were stratified into urban and rural categories, followed by random commune/ward selection using probability proportional to size sampling. Finally, eligible informal workers were identified through local administrative records and randomly selected using systematic sampling with random start points.
However, certain limitations regarding sampling should be acknowledged. First, regarding informal sector heterogeneity, whilst stratified sampling attempted to capture geographical and socioeconomic diversity, Vietnam’s informal economic sector exhibits extreme heterogeneity, encompassing various job types with different stability levels and characteristics. Consequently, certain highly specific or hard-to-reach labour groups may be underrepresented in our sample. Second, regarding proportional sample distribution, whilst samples were stratified by region and urban/rural areas, distribution across specific job types was not implemented according to exact national proportions due to a lack of detailed, updated micro-level data. Therefore, caution should be exercised when generalising results to specific occupational groups within the informal sector.
Scale Development and Instrument Validation
The measurement instrument was developed through a comprehensive five-stage process that adheres to established psychometric protocols whilst ensuring cultural appropriateness and contextual relevance for Vietnam’s informal sector (Churchill, 1979; DeVellis, 2017). This rigorous development process was essential given the need to adapt existing scales developed in different cultural contexts whilst developing new measures for constructs lacking validated instruments in Vietnamese informal sector contexts.
Stage 1: Literature Review and Domain Specification. An extensive systematic literature review was conducted to identify existing validated scales for each theoretical construct included in the integrated model. This review encompassed psychology, economics, health behaviour, and social policy journals to ensure comprehensive coverage of relevant measurement approaches. Selection criteria for constructs with multiple available scales emphasised psychometric quality, cultural appropriateness, and alignment with theoretical definitions adopted in this study.
Stage 2: Initial Item Pool Generation. Based on the literature review, initial item pools were generated for each construct through adaptation of existing measures combined with development of new items contextualised explicitly for Vietnam’s informal sector and VSI programme characteristics. This process involved extensive consultation with Vietnamese social policy experts to ensure that item content accurately reflected local programme features, cultural values, and linguistic preferences.
Stage 3: Expert Review and Content Validation. The preliminary instrument underwent comprehensive expert review by a multidisciplinary panel comprising academic researchers specialising in social protection and behavioural economics, practitioners from Vietnam Social Security with extensive VSI implementation experience, and informal sector representatives with direct programme knowledge. This review process focused on enhancing content validity, ensuring cultural appropriateness, and improving item clarity whilst maintaining theoretical consistency with the underlying constructs.
Stage 4: Cultural Adaptation and Translation. The instrument underwent rigorous cultural adaptation to ensure appropriateness for Vietnamese informal workers whilst maintaining theoretical integrity. For the intrinsic motivation scale, particular attention was devoted to cultural adaptation, given that motivation concepts may vary across cultural contexts. Original SDT-based intrinsic motivation items referencing concepts such as “personal fulfilment” and “individual satisfaction” were carefully modified through extensive consultation with local researchers and pilot interviews to emphasise culturally resonant concepts such as “family responsibility,”“personal security,” and “long-term stability” whilst maintaining consistency with intrinsic motivation’s theoretical definition as autonomous, self-determined engagement.
The translation process followed established forward-backward translation procedures using independent professional translators, followed by reconciliation sessions conducted by a bilingual committee of academic researchers to ensure conceptual equivalence rather than mere linguistic translation. Particular attention was paid to financial terminology, ensuring that concepts such as “contributions,”“benefits,” and “risk” were translated using terms familiar to informal workers whilst maintaining technical precision.
Stage 5: Pilot Testing and Preliminary Validation. The culturally adapted instrument was subjected to comprehensive pilot testing with 80 informal workers (10% of the final sample) selected from areas not included in the main study to avoid contamination effects. Pilot testing facilitated preliminary assessment of reliability, construct validity, and practical questionnaire administration protocols whilst identifying potential comprehension difficulties or cultural sensitivities.
Cognitive interviews were conducted with 12 pilot participants to assess item clarity, comprehension accuracy, and cultural appropriateness through detailed examination of response processes. These interviews revealed several instances where financial terminology required further simplification and identified cultural preferences for concrete examples rather than abstract statements, leading to minor item modifications that enhanced clarity without compromising theoretical content.
Exploratory factor analysis conducted on pilot data established preliminary dimensionality for each construct, with items demonstrating insufficient factor loadings (<0.50) or significant cross-loadings (>0.40) eliminated or refined through iterative testing. The refined instrument achieved satisfactory internal consistency reliability (Cronbach’s α > .70) for all constructs in the pilot sample, providing confidence for complete data collection.
Data Collection Procedures and Quality Assurance
Data collection occurred between November 2023 and March 2024, conducted by 24 trained field researchers. All interviewers received comprehensive 3-day training covering research objectives, sampling procedures, questionnaire administration, ethical considerations, and quality control protocols. Training emphasised building trust with informal workers and included extensive role-playing exercises and supervised practice sessions before independent deployment.
Comprehensive quality control measures included: (1) direct field supervision with 20% of interviews observed by senior researchers; (2) validation callbacks to 15% of respondents within 48 hr to verify participation and response consistency; (3) daily debriefing sessions to address challenges and maintain protocol adherence; (4) systematic questionnaire review for completeness and consistency; and (5) random re-interviews with 5% of participants to assess response stability.
Non-responses were systematically documented, achieving an overall response rate of 88%. Non-response was primarily attributed to unavailability (7%) and refusal to participate (5%), with analysis revealing no systematic biases related to geographic location, occupation type, or demographic characteristics.
Measurement Instruments and Control Variables
All theoretical constructs were measured using multi-item scales with 7-point Likert formats ranging from “strongly disagree” (1) to “strongly agree” (7). Comprehensive control variables were systematically incorporated, including demographic controls (age, gender, education level, monthly income), occupational controls (occupation type, experience years), and experience-related controls (prior insurance experience, VSI familiarity, information exposure). Alternative model specifications were tested with and without controls to ensure robustness of theoretical relationships.
Key construct measurements demonstrated satisfactory reliability:
Attitude toward VSI (ATT): Five items capturing evaluative judgements about VSI participation benefits and value (α = .892)
Subjective Norms (SN): Five items assessing perceived social pressure from family, friends, and community (α = .892)
Perceived Behavioural Control (PBC): Five items measuring perceived capability and resources for VSI participation (α = .893)
Risk Perception (RP): Five items capturing vulnerability to financial and health risks without social protection (α = .888)
Trust in VSI System (TRUST): Five items assessing institutional confidence and system reliability (α = .895)
Financial Literacy (FL): Five items combining objective knowledge and subjective understanding assessments (α = .886)
Perceived Benefits (BEN): Five items evaluating expected advantages from VSI participation (α = .903)
Intrinsic Motivation (IM): Five culturally adapted items emphasising personal satisfaction and autonomous choice (α = .781)
Income and Affordability (INC): Five items measuring objective income levels and affordability perceptions (α = .894)
Intention to Participate (INT): Five items capturing commitment and likelihood of VSI participation (α = .852)
Actual Participation (PART): Five items examining enrolment status and contribution behaviour (α = .825)
Common Method Bias Assessment and Mitigation
Given reliance on self-reported data, systematic attention was devoted to assessing and mitigating common method bias through multiple procedural and statistical remedies. Procedural remedies included: (1) temporal separation of predictor and criterion measures through strategic questionnaire ordering; (2) emphasis on honest responses and anonymity assurance; (3) varied response formats to reduce systematic patterns; and (4) inclusion of reverse-coded items for key constructs.
Statistical remedies encompassed: (1) Harman’s single-factor test revealing the most significant factor explained 24.3% of variance, below the 50% problematic threshold; (2) standard latent factor analysis demonstrating small, non-significant method factor loadings; (3) comprehensive collinearity assessment with all VIF values below 3.0; and (4) correlation analysis revealing theoretically consistent rather than uniformly high relationships, suggesting genuine substantive relationships rather than method artefacts.
Analytical Approach and Model Specification
Data analysis employed Partial Least Squares Structural Equation Modelling (PLS-SEM) using SmartPLS 4.1, selected for its ability to: (1) accommodate complex predictive models with multiple relationships simultaneously; (2) handle both formative and reflective measurement models; (3) demonstrate robustness to non-normal distributions; (4) focus on prediction and theory development; and (5) perform well with the employed sample size.
The analytical approach followed Hair et al.’s (2019) two-stage procedure. Measurement model evaluation encompassed: (1) individual indicator reliability (loadings > 0.70); (2) internal consistency reliability (Cronbach’s α and composite reliability > .70); (3) convergent validity (AVE > 0.50); and (4) discriminant validity (HTMT < 0.85). Structural model assessment involved: (1) collinearity examination (VIF < 3.0); (2) path coefficient estimation via bias-corrected bootstrapping (5,000 resamples); (3) coefficient of determination (R2); (4) effect sizes (Cohen’s f2); (5) predictive relevance (Stone-Geisser’s Q2); and (6) model fit evaluation (SRMR < 0.08, NFI > 0.90).
Mediation analysis followed Zhao et al.’s (2010) contemporary approach, emphasising the significance of the indirect effect. Moderation analysis employed the product indicator approach with interaction term creation between income and intention. Multi-group analysis examined rural-urban differences using permutation tests (5,000 permutations) following MICOM measurement invariance assessment.
Results
Measurement Model Assessment
The measurement model demonstrated exceptional reliability and validity across all theoretical constructs, providing a strong foundation for subsequent structural model analysis (Table 2). Individual indicator reliability was confirmed with all item loadings exceeding the 0.70 threshold, ranging from 0.713 to 0.890. Internal consistency reliability was established through both Cronbach’s alpha (ranging from .781 to .903) and composite reliability (ranging from .851 to .928), with all values exceeding the 0.70 criterion.
Reliability and Validity Assessment of Measurement Model.
Convergent validity was confirmed through AVE values ranging from 0.533 to 0.721, exceeding the 0.50 minimum requirement. Discriminant validity was comprehensively established through the HTMT criterion, with all values substantially below the conservative 0.85 threshold, with the highest ratio being 0.725 between Risk Perception and Intrinsic Motivation, reflecting theoretically expected association.
Structural Model and Hypothesis Testing
The structural model demonstrated exceptional explanatory power and theoretical validity, providing strong empirical support for the integrated TPB-SDT framework whilst revealing essential insights into the psychological and economic mechanisms underlying VSI participation decisions (Table 3 and Figure 2). Model fit assessment revealed excellent overall performance with SRMR = 0.035 and NFI = 0.893, achieving substantial explanatory power: R2 = .639 for intention to participate and R2 = .564 for actual participation.
Path Coefficients and Hypothesis Testing Results.
Structural model results.
All traditional TPB relationships received strong empirical support, confirming the cognitive foundations of VSI intention formation. Attitude toward VSI emerged as the strongest predictor of intention (β = .376, f2 = 0.382), indicating that positive evaluative judgements about VSI benefits constitute the most influential cognitive factor. Perceived behavioural control demonstrated the second-strongest effect (β = .308, f2 = 0.262), whilst subjective norms showed the weakest but still significant effect (β = .242, f2 = 0.162).
All extended cognitive factors demonstrated significant positive effects on intention formation. Perceived benefits showed substantial influence (β = .257, f2 = 0.180), whilst trust in the VSI system demonstrated considerable impact (β = .254, f2 = 0.176), highlighting the critical importance of institutional credibility. Risk perception showed significant influence (β = .230, f2 = 0.093), partially mediated through intrinsic motivation as discussed subsequently.
The integration of SDT constructs provided crucial insights into motivational quality. Intrinsic motivation demonstrated significant direct effects on intention (β = .208, f2 = 0.076) and participation (β = .213, f2 = 0.086), confirming that autonomous, personally meaningful engagement enhances psychological readiness and actual behaviour beyond cognitive determinants alone.
The income findings revealed sophisticated dual relationships illuminating the complex interplay between material constraints and psychological readiness. Income demonstrated a significant adverse direct effect on participation (β = −.119, f2 = 0.032), suggesting that higher-income informal workers may perceive reduced need for state-sponsored insurance due to access to alternative financial protection mechanisms. Simultaneously, income showed a substantial positive moderating effect on the intention-participation relationship (β = .350, f2 = 0.273), indicating that higher income levels significantly facilitate their translation into actual behaviour when positive intentions are formed.
Multi-Group Analysis: Rural Versus Urban Participants
A multi-group analysis was conducted to investigate potential differences between rural and urban informal workers (Table 4). The model demonstrated good fit for both rural (SRMR = 0.044) and urban (SRMR = 0.045) groups. MICOM procedures established partial measurement invariance, confirming that 10 out of 11 constructs were measured equivalently across groups.
Multi-Group Analysis Results.
Two significant differences illuminated important contextual variations in the integrated model’s operation. The effect of trust on intention was substantially more substantial among rural participants compared to urban participants (Δβ = .258, p < .001), suggesting that institutional credibility concerns are more pronounced in rural areas where information asymmetry may be greater, government services may be less accessible, and historical experiences with institutional programmes may differ.
The moderating effect of income on the intention-participation relationship was also significantly more pronounced in rural settings (Δβ = .410, p < .001), indicating that material constraints operate more powerfully in rural contexts where income volatility may be greater, alternative financial services may be less available, and economic vulnerability may be more severe.
Robustness Checks
To ensure the stability and reliability of findings, comprehensive robustness testing was conducted by examining alternative model specifications and analytical approaches (Table 5).
Robustness Analysis—Alternative Model Specifications.
p < .001.
Alternative model specifications demonstrated remarkable consistency in core relationships, with path coefficients maintaining their direction, magnitude, and statistical significance across different model configurations. Removing the income moderator or intrinsic motivation mediator had minimal impact on other relationships, confirming that findings reflect genuine theoretical relationships rather than spurious associations due to model specification choices.
Additional sensitivity analyses included: (1) examination of alternative measurement model specifications; (2) testing with different sample splits; (3) analysis using alternative handling of missing data; and (4) assessment using different bootstrapping procedures. All sensitivity analyses confirmed the stability and robustness of the main findings, providing high confidence in the reliability of the integrated model’s empirical support.
Discussion
Theoretical Implications and Scholarly Positioning
Advancement Beyond Single-Theory Approaches in Financial Protection Psychology
The empirical validation of the integrated TPB-SDT framework represents a significant theoretical advancement that addresses fundamental epistemological limitations inherent in mono-theoretical approaches to understanding voluntary financial protection behaviours. When positioned within the broader scholarly discourse on social insurance participation psychology, our integrated model demonstrates how cognitive determinants and motivational quality operate synergistically rather than independently, revealing emergent properties that neither theoretical perspective could illuminate independently.
The hierarchical pattern of cognitive influence—with attitude demonstrating the most substantial effect (β = .376), followed by perceived behavioural control (β = .308) and subjective norms (β = .242)—provides empirical support for the proposition that VSI participation constitutes a distinctly individualised financial protection strategy characterised by personal cost-benefit evaluation rather than socially constructed conformity behaviour. This finding gains particular theoretical significance when contrasted with Hien et al.’s (2023) application of Behavioural Reasoning Theory, which emphasised cultural factors such as collectivism and long-term orientation as primary determinants of VSI decisions. The relatively modest influence of subjective norms in our national sample suggests that whilst cultural collectivism may provide essential background conditions that shape value formation, the proximal psychological mechanisms underlying actual participation decisions operate primarily through individualised cognitive assessment processes.
Perhaps most significantly, identifying intrinsic motivation as a crucial mediator between risk perception and intention formation (indirect effect = 0.126, accounting for 35.4% of the total impact) constitutes a novel theoretical contribution transcending traditional TPB applications whilst extending SDT into voluntary financial protection contexts. This mediational pathway reveals how external threat recognition transforms into autonomous commitment through the internalisation of financial protection values, demonstrating that risk awareness operates not merely as cognitive information processing but as a catalyst for motivational development that enhances the personal significance and autonomous value of protective behaviours.
Dialectical Integration and Theoretical Synthesis
The superior explanatory power achieved by the integrated framework (R2 = .639 for intention, R2 = .564 for participation) provides compelling evidence for the theoretical synthesis value whilst engaging directly with calls for more sophisticated psychological models in emerging economy financial research (Musa et al., 2024). Compared to Vu’s (2025) foundational TPB application, which achieved R2 = .412 for intention, the enhanced predictive capacity reflects genuine synergistic effects rather than a mere additive combination of theoretical components, suggesting that integration creates emergent understanding that transcends the sum of constituent theoretical parts.
This theoretical advancement addresses a critical lacuna in recent scholarship regarding the intention-behaviour gap in voluntary financial protection contexts. The sophisticated dual relationship between income and VSI participation—demonstrating simultaneous adverse direct effects (β = −.119) and positive moderation effects (β = .350)—reveals how material constraints interact with psychological readiness through competing mechanisms that existing single-theory frameworks inadequately capture. The substitution mechanism whereby higher-income workers perceive reduced VSI necessity due to alternative protection access operates concurrently with an enabling mechanism that facilitates intention-to-behaviour translation when positive psychological dispositions exist, illustrating the theoretical sophistication required to understand voluntary financial behaviours in resource-diverse populations.
Contextualisation Within the Vietnamese Social Protection Research Landscape
Positioning Within Parallel Social Insurance Crises
The theoretical insights generated by our integrated model gain profound significance when positioned within the broader crisis of social insurance sustainability documented by T. P. Nguyen (2023), whose analysis of Vietnam’s social insurance dilemma reveals alarming patterns of withdrawal from mandatory schemes that parallel the low uptake of voluntary programmes examined in our research. The identification of institutional trust as a significant predictor of VSI intention (β = .254) provides a psychological mechanism explanation for the empirical reality that over 960,000 individuals withdrew from mandatory social insurance in 2021, representing a 9% annual increase during 2016 to 2021 that threatens system sustainability.
Our finding that trust effects are substantially stronger in rural contexts (Δβ = .258) resonates powerfully with T. P, Nguyen’s (2023) analysis of the underlying factors driving social insurance withdrawal, particularly worker scepticism about the state’s capacity for responsible fund management and transparency in benefit provision. This convergent evidence suggests that the challenge facing Vietnam’s social protection system extends beyond technical programme design to encompass fundamental questions of institutional credibility and citizen confidence that require a sophisticated understanding of the psychological foundations underlying financial protection engagement.
The complex income moderation effects identified in our model—whereby material constraints simultaneously reduce direct participation likelihood whilst strengthening the conditional probability of intention-to-behaviour translation—provide a theoretical explanation for the “precarious conditions” that T. P. Nguyen (2023) identified as driving workers to treat social insurance contributions as emergency savings rather than long-term protection investments. When informal workers face irregular income patterns and immediate survival pressures, the psychological tension between current consumption needs and future security investments becomes acute, explaining why even positive intentions may fail to translate into sustained behavioural commitment.
Regional Variations and Comparative Analytical Dialogue
The national scope of our research enables systematic comparison with geographically focused studies, particularly C. C. Nguyen et al.’s (2024) investigation of VSI intention in the Mekong Delta region, revealing important insights about how local contextual factors influence the operation of universal psychological mechanisms. Whilst both studies confirm the foundational importance of TPB constructs and risk perception in predicting VSI intentions, our national sample reveals weaker normative influences than the Mekong Delta focus, suggesting that regional variations in social cohesion and community influence patterns systematically affect the operation of psychological mechanisms.
The multi-group analysis revealing more substantial trust effects and income moderation in rural areas provides direct empirical support for the proposition that psychological mechanisms, whilst theoretically universal, demonstrate contextually variant manifestations that reflect local institutional arrangements, economic conditions, and social structures. Rural areas appear characterised by greater institutional uncertainty and more severe resource constraints that amplify the importance of trust and income in VSI decision-making, potentially reflecting limited government service access, greater information asymmetry regarding programme features, and higher income volatility that complicates financial commitment maintenance.
When juxtaposed with the econometric analysis of demographic and structural factors conducted by T. T. H. Nguyen and Pham (2023), our psychological focus reveals how individual-level cognitive and motivational processes may mediate relationships between macro-level determinants and participation behaviour. Education, occupation type, and demographic characteristics identified as significant predictors in their research likely influence VSI participation by shaping the cognitive assessments (attitudes, perceived benefits, risk perceptions) and motivational processes (intrinsic motivation quality) our model identifies as proximal psychological determinants.
Cross-National and Comparative Insights
Asian Context Integration and Theoretical Transferability
The substantial influence of perceived benefits on VSI intention formation (β = .257) receives robust empirical support from Shen et al.’s (2023) research in China, which demonstrated significant effects of benefit perceptions on health insurance participation intentions, providing cross-national validation for the importance of value proposition clarity in voluntary insurance decision-making across Asian emerging economy contexts. This convergent evidence suggests that careful evaluation of programme advantages relative to alternatives may constitute a universal feature of voluntary insurance psychology, particularly in cultural contexts emphasising prudent financial planning and cost-effectiveness assessment.
However, integrating our findings with international research reveals universal psychological mechanisms and context-specific variations illuminating cultural and institutional factors moderating theoretical relationship patterns. The critical importance of institutional trust in our Vietnamese model aligns with research across other emerging economies facing institutional credibility challenges, suggesting that trust effects may be particularly pronounced in governance contexts characterised by limited transparency, variable service quality, or historical experiences creating scepticism about government programme reliability.
The successful integration of TPB and SDT in our Vietnamese context provides a theoretical foundation for similar synthesis efforts across diverse cultural settings. However, the mediating mechanisms and moderating factors may vary with cultural values related to autonomy, individual responsibility, and government roles that differ systematically across societies. Future cross-national research should examine how cultural dimensions such as individualism-collectivism, uncertainty avoidance, and power distance affect the operation of cognitive and motivational mechanisms in voluntary financial protection decisions.
Practical Implications and Evidence-Based Policy Framework
Comprehensive Intervention Strategy Informed by Theoretical Integration
The empirical validation of our integrated model provides a robust foundation for developing evidence-based policy interventions that simultaneously address cognitive, motivational, and structural barriers to VSI participation whilst learning from the broader social insurance crisis context documented by T. P. Nguyen (2023). Given the substantial effect of attitude formation (β = .376) and insights about worker scepticism regarding institutional reliability, communication strategies must transcend traditional information provision to address fundamental concerns about programme value and institutional trustworthiness.
The significant mediating role of intrinsic motivation suggests that effective interventions must foster autonomous engagement rather than external compliance. This requires communication approaches that connect VSI participation with personal values, life goals, and autonomous choice rather than presenting it as a government obligation or social expectation. This approach gains particular importance given T. P. Nguyen’s (2023) analysis of how mandatory scheme failures have created worker cynicism that may extend to voluntary programmes. These interventions rebuild confidence through demonstrated respect for worker autonomy and choice.
The complex income moderation effects revealing enabling and substitution mechanisms require sophisticated policy responses that address material constraints whilst recognising that simple financial accessibility improvements may be insufficient for higher-income workers who perceive reduced programme necessity. For lower-income workers facing implementation barriers despite positive intentions, graduated contribution schedules, emergency withdrawal provisions, and integration with broader financial inclusion programmes could address the “precarious conditions” that force workers to prioritise immediate survival over long-term security.
Trust-Building and System Design Innovations
The substantial influence of institutional trust (β = .254), particularly its stronger rural manifestation, necessitates comprehensive governance improvements that address the underlying institutional credibility challenges that T. P, Nguyen (2023) identified as contributing to social insurance system erosion. Trust-building initiatives must include transparent fund management reporting, independent oversight mechanisms, and clear legal protections that demonstrate responsible stewardship while addressing historical corruption and mismanagement concerns.
The significant influence of perceived behavioural control (β = .308) provides empirical justification for operational reforms, including user-friendly digital platforms, flexible payment mechanisms accommodating irregular income patterns, and one-stop service centres that enhance worker confidence in their ability to navigate programme requirements successfully. These system design improvements must be coupled with rural-specific initiatives recognising more substantial contextual effects, including mobile service provision, community-based information sessions, and partnerships with trusted local organisations that can serve as credible intermediaries.
Limitations and Future Research Directions
Methodological Considerations and Theoretical Development Priorities
The cross-sectional design employed in this research precludes definitive causal inferences regarding temporal sequences of psychological development and behavioural engagement, particularly for the complex mediating relationships identified in the integrated model. Longitudinal research tracking individuals from initial VSI awareness through intention formation, enrolment decisions, and sustained participation maintenance represents a critical priority for understanding how cognitive and motivational factors evolve across different decision-making phases, illuminating the dynamic processes underlying commitment maintenance or withdrawal.
Integrating administrative data with psychological research could provide more robust behavioural measures whilst enabling examination of longer-term participation patterns that reveal the full lifecycle of voluntary social insurance engagement. Such research could examine how programme experiences influence the reinforcement or erosion of initial psychological dispositions, addressing the feedback relationships between behavioural outcomes and subsequent cognitive-motivational states.
Cross-Cultural and Comparative Research Extensions
The successful development of the integrated TPB-SDT framework in the Vietnamese context provides a foundation for systematic cross-cultural research examining how cultural, institutional, and economic factors moderate psychological mechanisms underlying voluntary social protection across diverse emerging economy contexts. Priority research directions include comparative studies with countries sharing similar informal sector challenges whilst differing on key cultural dimensions, investigation of institutional quality effects on trust relationships and perceived behavioural control, and analysis of economic development level effects on income moderation mechanisms.
The theoretical integration demonstrated in this research offers valuable insights for international development organisations designing social protection interventions, suggesting that effective programmes must address psychological and motivational factors alongside technical design considerations. Future research should examine the transferability of integrated psychological frameworks across cultural contexts whilst developing culturally adapted measurement instruments that maintain theoretical consistency for meaningful cross-national comparison.
Conclusion
This study advances theoretical understanding of voluntary social insurance participation by developing and empirically validating an innovative integrated framework that synthesises the Theory of Planned Behaviour and Self-Determination Theory to address critical limitations of single-theory approaches in explaining complex financial protection decisions among informal sector workers. The comprehensive empirical validation through data from 800 informal workers across Vietnam’s six major regions demonstrates that VSI participation emerges from sophisticated interactions between cognitive determinants, motivational quality, and material constraints rather than from any psychological or economic mechanism.
Four distinctive theoretical contributions emerge from this integration that advance understanding beyond existing frameworks. First, the framework explicitly connects the cognitive foundations of intention formation (captured through TPB’s attitude, subjective norms, and perceived behavioural control) with the motivational dynamics underlying sustained engagement (captured through SDT’s intrinsic motivation), providing a comprehensive explanation of both initial participation decisions and ongoing commitment maintenance. Second, identifying intrinsic motivation as a crucial mediator between risk perception and intention reveals novel psychological mechanisms through which external threat awareness transforms into personally meaningful, autonomous commitment to financial protection. Third, incorporating income as a complex moderator demonstrates how material constraints interact with psychological readiness through dual enabling and substitution mechanisms that existing frameworks inadequately capture. Fourth, the substantial explanatory power achieved (R2 = .639 for intention, R2 = .564 for participation) provides compelling empirical evidence for the synergistic value of theoretical integration rather than the mere additive combination of theoretical components.
The practical implications extend beyond academic theory development to provide evidence-based foundation for comprehensive policy interventions that address Vietnam’s alarming coverage gap between ambitious targets (60% labour force participation by 2030) and current reality (3.92% VSI participation). The findings demonstrate that effective VSI expansion requires moving beyond traditional approaches focused solely on financial incentives or information provision toward sophisticated interventions that simultaneously foster positive attitude formation, enhance perceived behavioural control through system design improvements, build institutional trust through transparency and responsive service provision, cultivate autonomous motivation through need-supportive communication approaches, and address income-related barriers through flexible contribution mechanisms that accommodate informal sector income patterns.
The research reveals significant contextual variations that inform intervention design, particularly the substantially stronger effects of institutional trust and income moderation in rural areas than urban contexts. These findings suggest that whilst universal psychological principles apply across diverse informal sector contexts, their practical implementation must be carefully tailored to local conditions that affect the relative importance of psychological and economic factors in decision-making processes.
The international implications extend beyond Vietnam to the broader global challenge of achieving universal social protection coverage in economies characterised by large informal sectors. The integrated theoretical framework demonstrates potential for application across emerging economies with similar coverage challenges. However, successful transferability requires careful attention to cultural values, institutional quality, and economic development factors that may moderate the operation of psychological mechanisms. Countries including Indonesia, the Philippines, India, and parts of Latin America and Sub-Saharan Africa could benefit from adapted applications of this framework whilst contributing to cross-national research that examines universal versus context-specific aspects of voluntary social protection psychology.
The study acknowledges significant limitations that inform future research priorities, particularly the cross-sectional design that limits causal inference and the reliance on self-reported measures that may introduce social desirability effects. Future research should employ longitudinal designs to examine psychological development over time, experimental approaches to test intervention strategies based on theoretical mechanisms identified through this research, and mixed-method approaches that combine quantitative modelling with qualitative exploration of psychological processes underlying financial protection decisions.
As emerging economies worldwide grapple with the fundamental challenge of extending social protection to large informal workforces who remain systematically excluded from formal coverage mechanisms, integrated theoretical frameworks that capture psychological and structural participation determinants offer promising pathways for designing more effective policies and interventions. This research represents an important step toward such a comprehensive understanding, demonstrating how cognitive and motivational perspectives synergistically enhance the explanation of voluntary financial protection whilst providing a validated framework for understanding similar behaviours across diverse economic and cultural contexts.
The findings ultimately demonstrate that achieving universal social protection coverage requires more than technical programme design or financial incentive provision—it demands a sophisticated understanding of the psychological mechanisms that drive individual decision-making combined with systematic attention to the material constraints that affect behavioural implementation. By establishing how cognitive determinants and motivational quality work together to influence both intention formation and the critical intention-behaviour relationship, this study provides valuable insights not only for Vietnam’s specific social protection challenges but for the broader global imperative of ensuring financial security and dignity for all workers, regardless of their employment status or economic circumstances.
The integrated TPB-SDT framework offers a robust theoretical foundation for future research whilst providing immediate practical guidance for policymakers seeking to bridge coverage gaps and enhance social protection effectiveness. As societies continue to evolve and labour markets become increasingly complex, the need for a sophisticated understanding of voluntary financial protection psychology will only grow, making this research contribution both timely and enduring in its relevance for social protection policy development and implementation.
Footnotes
Acknowledgements
The authors would like to express their gratitude to the National Economics University for providing financial support for this research. We also thank the high school administrators who facilitated the distribution of our survey, and all the students who participated in this study. Special appreciation goes to our colleagues in the Business School and Faculty of Planning and Development for their valuable feedback and suggestions throughout the research process.
Ethical Considerations
This study was conducted in accordance with the ethical standards and guidelines set forth by the Committee of Scientific and Training of the National Economics University (NEU), Vietnam. The research proposal was reviewed and approved by the committee to ensure adherence to ethical research practices.
Consent to Participate
All participants in this study provided informed consent prior to their involvement. The consent process included clear information about the study’s purpose, procedures, potential risks and benefits, and the voluntary nature of participation. Participants were informed of their right to withdraw from the study at any time without consequence.
Author Contributions
Quoc Dung Ngo: Conceptualisation; Data curation; Formal analysis; Methodology; Writing—original draft preparation; Writing—review & editing. Quynh Hoa Nguyen: Conceptualisation; Data curation; Investigation; Project administration; Supervision; Writing—original draft preparation; Writing—review & Editing. Both authors have read and agreed to the published version of the manuscript.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the National Economics University, Hanoi, Vietnam.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author, Quynh Hoa Nguyen, upon reasonable request. The data is not publicly available due to privacy and ethical restrictions.