The Influence of Assisted Digital Trust on Elderly Purchasing: The Case of Thai Digital Financial Services

Abstract

Aim/Purpose: Despite 84.3% smartphone ownership and 62.1% Internet usage among elderly Thais aged 60 or over, only 17% shopped online in 2024 (NSO, 2024). This striking disconnect between technological access and transactional behavior warrants systematic investigation. This study examined whether digital financial experience, particularly helper-mediated e-banking and e-payment usage, catalyzed shifts from digital hesitation to confident online purchasing among Thailand's elderly population. Understanding this progression matters critically as Thailand's Silver Digital Economy expands and aging populations worldwide face similar digital inclusion challenges.

Introduction/Background: Thailand's aging population represents growing economic importance yet has demonstrated minimal e-commerce participation despite high digital connectivity. Traditional technology acceptance models emphasize cognitive factors but have proved inadequate for explaining elderly adoption patterns where emotional barriers dominate rational assessment. Digital financial confidence among elderly Thais is substantially below that of younger cohorts even after controlling for education, income, and urban residence (NSO, 2024). Fear of irreversible financial mistakes have emerged as the primary psychological obstacle, intensified by a dramatic increase in online fraud complaints in recent years (ETDA, 2023). Trust formation occurs through interpersonal channels rather than direct system evaluation. While the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT) focus on cognitive evaluations such as perceived usefulness and ease of use, they failed to adequately address emotional vulnerability, relational trust transfer, or experiential learning requirements characteristic of elderly digital adoption. These gaps have necessitated new theoretical frameworks capable of explaining how vulnerable populations overcome fear-based barriers through assisted experiences.

Methodology: In this study, a sequential explanatory mixed-methods design was employed. The quantitative phase analyzed Thailand's Information and Communication Technology (ICT) Household Survey 2024 Q2 microdata (NSO, 2024),elderly respondents representing a weighted population of approximately 14 million (response rate 87.3%). Weighted binary logistic regression using Stata 17 tested whether digital financial experience predicted online purchasing probability while controlling for demographic, socioeconomic, geographic, and digital engagement factors across four nested models, examining the moderating influences of social media engagement frequency, educational attainment, and urban-rural residence. For the qualitative investigation, 24 semi-structured interviews were conducted across six Thai regions using purposeful maximum variation sampling. Trained research assistants conducted 50 to 70–minute Thai language interviews exploring digital platform usage, first financial technology experiences, and trust development mechanisms. Reflexive thematic analysis was employed following Braun and Clarke's (2019) six-phase approach, achieving an inter-rater reliability with Cohen's κ = .82. Integration occurred through joint displays juxtaposing statistical results with qualitative themes.

Findings: Digital financial experience dramatically dominated all other predictors. E-banking increased purchasing probability by 16.7 percent while e-payment added 12.4 percent, substantially exceeding tertiary education (+8.3%), social media frequency (+5.2%), and urban residence (+4.1%). Model performance was excellent (Area Under the Receiver Operating Characteristic Curve = .84, Hosmer-Lemeshow χ² = 12.4, p = .13). All four hypotheses received empirical support. H₁ was strongly supported: e-banking (OR = 2.76), and e-payment (OR = 2.41) significantly predicted purchasing. H₂ was supported: social media moderated e-banking effects (+5.0% among frequent users). H₃ received the strongest moderation support: tertiary education amplified e-banking effects by 7.2 percent. H₄ received partial support: urban residence moderated effects modestly (+4.3%). Qualitative analysis revealed four interconnected mechanisms: a) fear of irreversible financial errors, b) primacy of relational trust through human intermediaries, c) experiential proof through assisted first transactions, and d) self-reinforcing confidence loops through repeated successes. These findings revealed that experiential trust formation through assisted digital finance, rather than demographic advantage or technology access, was the primary driver of elderly online purchasing behavior, with the effects of digital finance nearly double those of tertiary education.

Contribution: This research proposes the Assisted Digital Trust Mechanism (ADTM), a four-stage theoretical model that integrates emotional vulnerability barriers, relational trust transfer processes, and experiential learning as fundamental mechanisms for adoption. ADTM stages comprise: (a) Digital Exposure where transaction barriers feel insurmountable; (b) Assisted Digital Finance Experience where trusted figures guide high-stakes financial transactions; (c) Trust Transfer where confidence shifts from helpers toward technological systems; and (d) Confidence Reinforcement where repeated successes enable autonomous e-commerce participation. ADTM extends the TAM and UTAUT by foregrounding emotional vulnerability, relational trust transfer, and experiential learning as dimensions that cognitive-focused models largely neglect. The findings challenge assumptions that socioeconomic uplift or classroom training alone closes digital divides, enabling more effective intervention design prioritizing guided transactional experience.

Recommendations: Platform providers should implement elder-safe design features that include prominent security assurances, confirmation dialogs, and accessible human support. Financial institutions should offer systematically assisted onboarding programs where trained staff guide elderly customers through initial digital transactions using real money, recognizing that experiential mediation proves more effective than classroom instruction. Policymakers should prioritize community-embedded programs in culturally familiar settings, providing one-on-one guidance with actual financial transactions. Core ADTM mechanisms operate universally across demographic groups, though oldest-old and least-educated segments require longer support timelines.

Research Limitations: Several limitations warrant consideration. Cross-sectional design prevented definitive causal inference regarding temporal sequences. The survey lacked direct measures of helper presence or emotional states during adoption. Qualitative sampling (n = 24) may have missed experiences of marginalized subgroups. Thai cultural traditions emphasizing family interdependence may have amplified trust transfer mechanisms relative to individualistic societies, limiting direct transferability to Western contexts.

Future Research: Four research directions emerge. Longitudinal studies would establish causal sequences and identify trigger events precipitating trust transfer. Experimental interventions would validate whether guided practice outperforms classroom training. Cross-cultural research would establish ADTM boundaries and distinguish culture-specific from universal mechanisms. Domain extension testing ADTM for telemedicine, e-government, and fintech adoption would demonstrate generalizability beyond e-commerce contexts.