The ability to reflect is a lifelong learning skill necessary for success within and beyond the classroom. This project seeks to advance people's skills around reflection through using AI-supported prompts to deepen their critical thinking, using problem/puzzle solving as the concrete learning task. Furthermore, the team aims to advance knowledge of how recent advances in AI can be used to foster reflective learning "just-in-time" and "on-the-job" in different interaction contexts: while working behind a computer, in virtual spaces, or physical spaces. The methods are intended to be adaptable to a number of contexts, including education (digital learning environments), the workplace (digital training), gamified simulations (such as escape rooms or virtual scenarios for training), and self-understanding in everyday life (e.g., personal health or fitness apps). Through this work, the project will make significant advances around supporting reflection, with the potential to develop self-guided educational experiences that may be especially useful for populations underserved by existing educational opportunities.<br/><br/>The project team will create a study environment in the form of an escape room that poses problem-solving tasks, using three different interaction contexts: a 3D digital screen-based version, a virtual reality simulation, and an augmented reality physical format. The team will build and validate an AI-empowered reflective learning tool that seeks to deepen and accelerate learning with progressively more complex and expansive capabilities, including using large language models to give feedback and prompts for reflection, data-driven detection of learners' behaviors and emotions, and data-driven detection of the learner's environment. The researchers will run in-person and online user studies to examine when and how reflection should take place across the three virtuality contexts, with the goals of optimizing the timing and style of reflection to deepen learners' understanding of problems, accelerate their ability to find solutions to problems, and increase their engagement in the learning environment while reflecting. The result of this project is the creation of a flexible AI-empowered reflective learning tool that can be implemented in many forms of modern technologically enhanced learning and training environments.<br/><br/>This project is funded by the Research on Innovative Technologies for Enhanced Learning (RITEL) program that supports early-stage exploratory research in emerging technologies for teaching and learning.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.