<![CDATA[The electrification of two-wheelers (E2Ws) is increasingly recognised as a high-impact decarbonisation strategy in emerging Asian markets, where motorcycles are the dominant mode of transport. However, the literature remains fragmented, and behavioural studies consistently report a persistent gap between stated adoption intentions and actual purchase behaviour. This study provides a robust empirical proxy to address these limitations through a two-phase hybrid approach: a Scoping Review and Bibliometric Analysis (ScoRBA) of 278 Scopus-indexed articles, synthesised using the PAGER framework, followed by empirical validation via a multi-model machine learning approach applied to a stated-preference dataset of 6,040 respondents from Solo, Indonesia. Bibliometric mapping identified three core socio-technical research clusters. The empirical analysis revealed a finding that departs substantially from prevailing assumptions: Perceived E-bike Quality, not financial incentives or operational costs, emerged as the dominant predictor of adoption. While initially identified via a baseline Decision Tree, this dominance was robustly validated across advanced ensemble algorithms and confirmed via SHAP analysis (Mean |SHAP Value| = 17.88%). Correctly situated within the Technology Perception dimension of the Technology Acceptance Model (TAM), this variable's dominance implies a sequential cognitive architecture: technology credibility must be established before economic evaluation becomes relevant. Consequently, policymakers in similar motorcycle-dominated transitional markets should prioritise quality certification and demonstration programmes before deploying purchase subsidies at scale.]]>
