An analysis of consumer's trusting beliefs towards the use of e-commerce platforms

The study addresses how consumers’ perceptions of security-related dimensions in e-commerce (privacy awareness, information confidentiality, prevention of unauthorized secondary data usage, and information integrity) shape trusting beliefs, and how these trusting beliefs drive behavioral intention to use e-commerce platforms. Motivated by rising online fraud, unethical practices, and data misuse, the research aims to clarify how security mechanisms influence consumer trust and adoption, grounded in social contract theory (SCT) and economic theory (ET). The paper sets three objectives: (RO1) examine links between consumers’ security perceptions and behavioral intention to use e-commerce platforms; (RO2) assess the mediating effect of trusting beliefs between security measures and behavioral intention; (RO3) test moderating effects of gender, age, and frequency of e-commerce use. Using SEM with a sample of Indian online shoppers, the study highlights the importance of strengthening privacy protections, transparency, and consumer education to foster trust and adoption.

Trust formation in e-commerce is multifaceted, involving disposition to trust, institutional trust, and interpersonal trust. Prior work shows that privacy protection, readability and clarity of privacy policies, and information quality contribute to consumer trust, while data breaches and unethical behaviors undermine it. SCT emphasizes moral and ethical norms (privacy, integrity, confidentiality) and hyper-norms (policies) to build trust, suggesting that transparent privacy guidelines and data usage policies enhance trust and usage. ET underscores transparency, legal protections, and enforceable contracts to safeguard consumer interests and bolster trust in online transactions. Exploitation theory highlights consumers’ comparative weakness versus firms in online contexts due to information asymmetries, reinforcing the need for protective frameworks. Theoretical gaps include: (1) trust often treated as a single antecedent alongside other factors, with few studies elaborating multiple trust-related dimensions in e-commerce; (2) predominance of descriptive or fragmented studies; (3) limited empirical work positioning trust as a central construct with multiple privacy/security antecedents; and (4) scant analysis of demographic moderators (gender, age, usage frequency). Building on Suh and Han (2003), the present study extends by incorporating awareness of privacy protection and secondary data usage concerns within an Indian context and tests moderation by demographics and frequency of use.

Design: Positivist, cross-sectional study employing covariance-based structural equation modeling (CB-SEM) for confirmatory theory testing. CB-SEM (AMOS 26.0) was chosen over PLS-SEM due to model fit assessment and suitability for confirmatory research. An action research strategy is described as informing a problem-solving orientation. Measures: All constructs measured with established multi-item scales on a 7-point Likert scale (1=strongly disagree, 7=strongly agree) without deliberate rewording to preserve cognitive validity. - Awareness about privacy protection: 3 items from Malhotra et al. (2004), validated by Okazaki et al. (2020) (AVE=0.708, CR=0.878). - Perceived information confidentiality: 4 items from Smith et al. (1996)/Malhotra et al. (2004), validated by Okazaki et al. (2020), Vimalkumar et al. (2021) (AVE=0.776, CR=0.932). - Preventing unauthorized secondary data usage (PUSDU): 4 items from Smith et al. (1996)/Malhotra et al. (2004), validated by Okazaki et al. (2020), Vimalkumar et al. (2021) (AVE=0.874, CR=0.965). - Information integrity: 3 items from Smith et al. (1996) (AVE=0.809, CR=0.927). - Trusting beliefs: 5 items from Jarvenpaa et al. (1999), Malhotra et al. (2004) (AVE=0.760, CR=0.940). - Behavioral intention: 4 items from Lee et al. (2017) (AVE=0.840, CR=0.955). Sampling and data collection: Purposive plus snowball sampling targeted online shoppers across India via online survey (Google Forms) distributed by email, WhatsApp, and social media. Data were collected August–September 2021. After cleaning, 780 valid responses remained. Sample profile: 50.1% male; age: <25 (56.7%), 25–34 (36.4%), 35–44 (2.9%), 45–54 (3.2%), 55–64 (0.89%); education mainly graduates (59.7%) and postgraduates (29.5%); internet experience >3 years (82.4%). Common method bias: Procedural remedies applied; statistically, Harman’s single-factor test indicated 36.882% variance (below 40% threshold), suggesting CMB unlikely. Analysis: CFA (AMOS 26.0) confirmed measurement properties: χ²=602.645, df=215, χ²/df=2.803, NFI=0.955, IFI=0.964, CFI=0.963, RMSEA=0.066; all loadings 0.731–0.938 (p<0.01). Convergent validity (AVE≥0.708) and reliability (CR≥0.878) satisfied; discriminant validity met (Fornell-Larcker). Structural model fit: χ²=585.317, df=203, χ²/df=2.883, NFI=0.975, RFI=0.968, IFI=0.983, TLI=0.971, CFI=0.982, RMSEA=0.049. R²: Trust=0.740; Behavioral intention=0.799. Mediation tested via indirect effects in AMOS. Moderation tested by multi-group invariance across gender (male n=391; female n=389), age (low n=726; high n=54), and frequency of use (low n=437; high n=343) using chi-square difference tests.

- Direct effects (all significant and positive): Awareness→Trust (β=0.048, t=2.024, p<0.05); Confidentiality→Trust (β=0.473, t=15.747, p<0.01); PUSDU→Trust (β=0.060, t=2.588, p<0.01); Information Integrity→Trust (β=0.714, t=19.675, p<0.01). Trust→Behavioral Intention (β=0.713, t=17.442, p<0.01). - Variance explained: Trust R²=0.740; Behavioral Intention R²=0.799. - Indirect (mediated) effects via Trust: Confidentiality→Trust→BI: 0.337 (p<0.01); Integrity→Trust→BI: 0.509 (p<0.01). Indirect paths for Awareness and PUSDU to BI were not significant. - Total effects on BI: Trust (β=0.713, p<0.01) largest, followed by Integrity (β=0.509, p<0.01), and Confidentiality (β=0.337, p<0.01); Awareness (β≈-0.034, ns) and PUSDU (β≈0.043, ns) showed no significant total effect on BI. - Moderation: No significant differences by gender (H6a–H6e not supported) or age group (H7a–H7e not supported). Frequency of use significantly moderated two paths: Confidentiality→Trust (Δχ²[1]=15.716, p<0.01) and PUSDU→Trust (Δχ²[1]=14.200, p<0.01) (H8b and H8c supported); other moderated paths by frequency not supported. - Measurement model and structural model exhibited good fit indices (e.g., CFI=0.963 and 0.982 respectively; RMSEA=0.066 and 0.049). Overall, information integrity and information confidentiality are the strongest antecedents of trusting beliefs, and trusting beliefs robustly predict behavioral intention to use e-commerce platforms.

Findings confirm that enhancing key security dimensions—information integrity, confidentiality, and clear policies preventing unauthorized secondary data usage—builds consumers’ trusting beliefs, which in turn substantially increase intention to use e-commerce platforms. Information integrity emerged as the strongest antecedent, underscoring the role of accurate, reliable, and authentic information practices in reducing uncertainty and risk. Confidentiality significantly shaped trust, aligning with SCT and ET tenets that emphasize moral obligations and enforceable protections. While awareness of privacy protection positively influenced trust, its overall contribution to intention was limited without the trust pathway. The frequency-of-use moderation indicates that more experienced or frequent users weigh confidentiality and secondary data usage policies differently in forming trust, suggesting user experience shapes sensitivity to specific security facets. Practically, results advocate for transparent, enforceable privacy frameworks, active communication of security measures, and user education to strengthen trust and drive adoption. The study extends theory by integrating SCT and ET in a comprehensive trust-centric model explaining substantial variance in behavioral intention.

The study demonstrates that consumers’ trusting beliefs mediate the effects of multiple security-related perceptions on their intention to use e-commerce platforms. Information integrity and information confidentiality are the most influential drivers of trust, and trust strongly predicts behavioral intention. While demographic factors (gender, age) did not moderate the relationships, frequency of platform use significantly moderated the effects of confidentiality and prevention of unauthorized secondary data usage on trust. Contributions include a multi-dimensional trust framework grounded in social contract and economic theories, robust empirical support via CB-SEM, and practical guidance for e-commerce firms to bolster trust through transparent, secure data practices. Future research should expand theoretical lenses beyond SCT/ET, diversify samples and contexts, adopt longitudinal designs to assess causality and persistence (e.g., repeat use/return intention), and examine additional factors such as UI design, engagement, and personalization on continued usage.

- Theoretical scope: Model grounded primarily in social contract theory and economic theory; future studies should incorporate alternative theoretical constructs for comparison. - Method: Self-administered online survey may introduce self-response bias; although procedural/statistical controls were applied, generalizability should be interpreted cautiously. - Sampling: Purposive and snowball sampling in India may yield sampling bias; random sampling and broader demographic/sociographic diversity are recommended. - Design: Cross-sectional data limits causal inference; longitudinal designs are encouraged to assess temporal effects and persistence behaviors. - Outcome focus: Study focuses on behavioral intention; future work should examine continued usage/return intention and additional antecedents (e.g., engagement, UI, personalization).