This study proposes and validates the Cognitive-Emotional Technology Acceptance Model (CETAM), a framework that integrates cognitive, emotional and behavioral factors to explain the adoption of banking chatbots.
An online survey was conducted with 231 banking chatbot users across six Latin American countries. A multi-method SEM approach was employed: partial least squares-structural equation modeling (PLS-SEM) (SmartPLS 4.0) for predictive evaluation (R2, Q2, f2, 5000-resample bootstrap) and covariance-based (CB-SEM) [Jeffreys's Amazing Statistics Program/Maximum Likelihood Robust (JASP/MLR)] for confirmatory fit assessment [Comparative Fit Index (CFI), Tucker–Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), Standardized Root Mean Square Residual (SRMR)]. Psychometric validation confirmed reliability (a, CR > 0.70), convergent validity (average variance extracted (AVE) >0.50) and discriminant validity (heterotrait-monotrait (HTMT) <0.85, Fornell–Larcker). Common method bias/variance (CMB/CMV) was assessed via Harman's single-factor test (first factor: 49.15% < 50% threshold; Bozionelos and Simmering, 2022). Mediation effects were tested using bias-corrected bootstrap confidence intervals.
An online survey was conducted with 231 users of banking chatbots in six Latin American countries. The study employed a multi-method SEM approach: PLS-SEM (SmartPLS 4.0) for predictive evaluation (R2, Q2, f2, bootstrap routes with 5,000 resamples) and CB-SEM (JASP/MLR) for confirmatory theoretical testing and overall fit assessment (CFI, TLI, RMSEA, SRMR). Rigorous psychometric validation ensured reliability (a, CR > 0.70), convergent validity (AVE >0.50) and discriminant validity (HTMT <0.85, Fornell-Larcker criterion). CMB/CMV were assessed via Harman's single-factor test, with the first unrotated factor accounting for 49.15% of total variance, below the recommended 50% threshold. Mediation effects were examined using bootstrap with bias correction and confidence intervals.
While the CETAM demonstrates strong explanatory power for banking chatbots, its generalizability to other financial technologies requires further testing. The model may be more applicable to high-risk financial decisions than to routine transactions. Future research should incorporate more differentiated emotional states beyond trust, employ experimental designs to map cognitive-emotional causal pathways and test the model with balanced, multinational samples to examine actual technology use within the cognitive-emotional-behavioral process.
The CETAM provides banking institutions with a structured implementation framework that prioritizes interventions according to their psychological impact. Banking institutions should prioritize building trust and omnichannel access when implementing chatbots, as these factors show the greatest psychological impact. Structural safeguards should follow, while traditional usability features are secondary. This suggests a strategic reorientation: first establish emotional trust and contextual reliability, and then introduce functional benefits within this secure framework. Marketing and onboarding messages should reflect the CETAM cognitive-emotional sequence – addressing security and accessibility concerns before highlighting usability – to align with customers' psychological adoption process.
This study presents CETAM, an integrated framework that systematically combines cognitive, emotional and behavioral factors to explain the adoption of banking chatbots. Through validation using a multi-method SEM approach, it demonstrates the primacy of trust as an emotional factor preceding chatbot usage intentions. The CETAM's theoretical innovation lies not merely in adding emotional constructs to existing models but in reconceptualizing the acceptance process itself as a sequential cognitive-emotional cascade, coherent with Lazarus's (1991) proposition that cognitive appraisals serve as the foundation for emotional and behavioral responses.
