Abstract
Digital competence among adolescents increasingly requires more than basic operational digital skills; it also involves ethical judgement, psychosocial well-being, safety, responsible use of artificial intelligence (AI), and a supportive implementation ecosystem. This study reports the second-round Fuzzy Delphi Method (FDM) consensus analysis used to refine the Digital Education Competency Framework for Insan Sejahtera (i-Digital Sejahtera), a holistic digital education competency framework for adolescents. Eleven experts evaluated 79 proposed items organised into 19 themes and five clusters using a five-point Likert scale. Ratings were converted to triangular fuzzy numbers and analysed using the threshold value (d), expert-consensus percentage, and defuzzification score (A). The pre-specified acceptance rules were d ≤ 0.20, expert consensus ≥ 80%, and A ≥ 0.50. Overall, 76 of 79 items (96.2%) met all three criteria. Four clusters achieved complete acceptance: Foundation & Value Formation (18/18), Digital Safety & Law (9/9), Future Literacy & Competency (17/17), and Ecosystem & Implementation (9/9). The Psychosocial & Well-being cluster achieved high overall acceptance (23/26), but three items within the Risk theme which are Reward vs. Punishment, Motivation and Behavioural Control, did not reach the 80% consensus criterion (72.727%), despite acceptable d and A values. The highest defuzzification score (A = 0.800) was observed for cross-stakeholder collaboration, psycho-educational programmes and infrastructure strengthening. The findings provide expert-consensus evidence for a multidimensional adolescent digital competency framework that integrates values, well-being, safety, future-oriented literacy and ecosystem support.