TRANSFORMATIONAL LEADERSHIP IN AI-MEDIATED WORK: DRIVING INNOVATION AND LEARNING IN EDUCATION

Measurement Development
The survey instrument comprised three sections. The first introduced the study, explained its purpose, assured respondents of anonymity, and sought informed consent. The second collected demographic information. The third contained 30 measurement items representing the five constructs of this study, measured on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).
Transformational leadership was measured using eight items adapted from (Dai et al., 2013), grounded in the Multifactor Leadership Questionnaire (Bass & Avolio, 1994). The items capture the extent to which leaders inspire, motivate, intellectually stimulate, and provide individualized support.
Innovative work behavior was measured with nine items from (Janssen, 2000), capturing idea generation, promotion, and implementation. Items assess the extent to which employees create, advocate, and implement new solutions to improve organizational performance.
Motivation to learn was measured with four items adapted from (Afsar & Umrani, 2020). Example items include: “I am motivated to learn the skills emphasized in my job” and “I will try to learn as much as I can from my job.”
Innovation climate was measured using items from(Ostroff et al.), which was also utilized in (Kang et al., 2016), reflecting organizational openness to change and support for innovation. Example items include: “In general, my organization rewards achievement in R&D and technological innovation” and “My organization publicly recognizes those who are innovative”.
AI-enabled job complexity was assessed with three items adapted from (Verma & Singh, 2022). Example items include: “AIS allows me to handle several tasks simultaneously” and “AIS makes my job relatively easier.”

Sampling and Data Collection
A simple random sampling method was employed to give each qualified employee in the target institutions an equal chance of selection. The study population consisted of staff in four universities in Vietnam. Principals and managers were contacted in advance to explain the study’s objectives and to guarantee confidentiality. With their consent, 800 questionnaires were distributed in May 2025.
A total of 365 responses were returned, yielding a response rate of 45.6%. After removing 65 incomplete or careless responses, 300 valid surveys were retained for analysis. This exceeds the minimum sample size recommended by (Rosce, 1975) for behavioral studies and provides sufficient statistical power for structural modeling. Importantly, the sample included highly experienced professionals, with nearly half (47.7%) having 10–15 years of experience, ensuring responses reflected deep institutional knowledge.