Browsing by Author "Flores, Pablo"

Objectives. This article-based master’s thesis explores the use of ChatGPT in educational settings. Previous research on human-LLM remains scarce, and most literature addressing the use of GPT in educational settings is theoretical and lacking empirical evidence. The new technological developments, however, urge for a deeper understanding of its novel dynamics for the development of efficient and safe AI-systems. Consequently, our study aims explore the use of a novel guided interaction design for modeling users’ information foraging behavior when navigating GPT-generated content and the role of Computational Thinking Skills in shaping such behavior. Methods. Conducted with nine educational researchers in a doctoral-level AIEd course, our research used editable prompt templates and keywords to structure the prompt crafting process. We modeled and analyzed participants’ interactions with ChatGPT in terms of exploration (to generate and explore various information landscapes) and exploitation (to delve deeper into a specific landscape). Additionally, we conducted the Computational Thinking Scale survey. We employed descriptive statistics to describe participants’ foraging behavior, and network analysis to explore the relationship between foraging behavior and Computational Thinking Skills. Results and conclusions. Our results suggested that Algorithmic Thinking and Creativity might encourage exploitation behavior, leaning more on AI-generated information rather than pre-defined design elements. Furthermore, including participants' interests in the interaction design seemed to foster a shared conceptual space in prompt construction. This approach encouraged the use and combination of diverse interests for content creation, as opposed to relying solely on individuals' interests. Our findings also suggested that exploitation prompts are predominantly driven by GPT-generated content. While this seems to add value to AI-generated content, it raises concerns about potential overreliance, especially in educational settings. The article, entitled ‘Exploring the Use of GPT-4 when Generating Personalized Case Scenarios for Higher Education.’, follows the guidelines of the International Journal of Artificial Intelligence in Education.