Mackenna Schiavo, Todd Loughead, Krista Monroe-Chandler, Mason Sheppard

Athlete leadership plays a critical role in fostering effective team functioning in sport, with research consistently demonstrating positive associations between athlete leadership, team cohesion, and performance outcomes (Loughead et al., 2016). As interest in athlete leadership continues to grow, there is increasing demand for measures capable of capturing its multidimensional and context-sensitive nature. Recent advances in artificial intelligence (AI), particularly large language models (LLMs), provide new opportunities to support theory-driven scale development through automated and linguistically informed item generation (De Duro et al., 2025). Frameworks such as AI-GENIE (Automatic Item Generation and Validation via Network-Integrated Evaluation) integrate generative AI with network psychometric methods to enhance the efficiency and conceptual rigor of measurement development. The present study applies the AI-GENIE framework to generate questionnaire items assessing athlete leadership. Using theoretically grounded prompts, AI-generated items will be evaluated using network analytic techniques, including Exploratory Graph Analysis, Bootstrap Exploratory Graph Analysis, Unique Variable Analysis, and Entropy Fit Indices, to examine dimensional structure, stability, redundancy, and conceptual coherence. To establish factorial validity, the dimensional structure derived from the AI-GENIE framework will be tested in a sample of approximately 300 intercollegiate athletes using confirmatory factor analysis, thereby evaluating the extent to which the AI-generated items reproduce a theoretically coherent latent structure using human data. Through these analyses, this study aims to situate the AI-generated questionnaire within the existing athlete leadership literature and contribute to theory-driven advances in sport psychology measurement.