This study aims to clarify the associations between task complexity and Artificial Intelligence (AI) dependency among university students and to examine the roles of cognitive load, future anxiety, and task motivation in these associations. A structured questionnaire survey was conducted with 442 college students, collecting self-reported data on task complexity, AI dependency, cognitive load, future anxiety, and task motivation. Confirmatory factor analysis was used to assess the adequacy of the proposed framework, and structural equation modeling was then employed to examine the associations among the variables. Task complexity was significantly and positively associated with AI dependency. Further analyses indicated that students facing more complex tasks tended to report higher cognitive load and stronger future anxiety, which were linked to greater AI dependency. In contrast, students with higher task motivation were less likely to over-rely on AI even when dealing with complex tasks. No significant gender differences were found in AI dependency or related psychological variables. This study advances a task-focused perspective on AI dependency by integrating Cognitive Load Theory and Self-Determination Theory and by specifying cognitive load, future anxiety, and task motivation as correlates linking task complexity with AI dependency. Practically, the results suggest actionable directions for course and assessment design (e.g., calibrating task complexity, scaffolding task decomposition and staged feedback, and supporting motivation and emotion management) to help students balance AI use with autonomous learning.
Li et al. (Thu,) studied this question.