My research investigates the design and efectiveness of Edu-Metaverses, i.e., persistent, multi-user, virtual learning environments leveraging Extended Reality technologies, with the aim of addressing both learners' and educators' perspectives to develop evidence-based guidelines for efective Edu-Metaverses. My first user study, conducted within an Edu-Metaverse I developed, reveals that collaborative learning significantly improves the acquisition and retention of concise cultural heritage knowledge compared to individual exploration. Other achieved results showed how locomotion techniques influence learning outcomes, showing that teleportationbased methods enhance vocabulary learning while minimizing simulator sickness. From a technical standpoint, we optimized path planning algorithms to support scalable navigation in complex virtual environments, laying the groundwork for future integration of autonomous agents and multi-user interactions inside Edu-Metaverses. My research also identified future research areas related to embodiment, adaptive content delivery, and feedback mechanisms for both educators and learners, along with broader possibilities in AI-enhanced pedagogical tools. Building on this, I propose a draft framework to guide educators, developers, and institutions in creating scalable Edu-Metaverses to improve knowledge acquisition, retention, and engagement across diverse educational contexts.
The rapid evolution of Extended Reality (XR) technologies has expanded educational possibilities [
While XR technologies’ efects and their positive impact on learning are well-established in the
literature [
My contribution is a structured, dual perspective (learners’ and educators’) framework that extrapolates lessons learned from user studies to guide efective construction of Edu-Metaverse learning environments that promote improved knowledge acquisition and retention across diverse learning contexts and scenarios. To this purpose, my work seeks to establish evidence-based design principles that leverage the unique possibilities of Edu-Metaverses while acknowledging their distinct characteristics and similarities to XR applications.
The Metaverse has gained attention in recent years (see Mystakidis [16]), with a growing interest in the
industry, especially after Facebook’s 2021 rebranding as Meta [15, 17]. Although the idea first appeared
in science fiction (see Neal Stephenson’s 1992 book [ 18]), it has gradually shaped into a more tangible
concept. As a result, researchers are looking into the Metaverse as a promising tool in many fields. In
education specifically, this has led to the Edu-Metaverse paradigm [
The Edu-Metaverse is opening up new possibilities for education by taking learning beyond traditional
classrooms and linear teaching methods [
To realize its potential, researchers have developed multiple frameworks. Wang et al. [
Beyond more theoretical frameworks, recent research has produced concrete Edu-Metaverse
applications and domain-specific systems [
However, existing frameworks acknowledge significant limitations in bridging theory and practice
[
To address the lack of integrated pedagogical and technical models (see Peña-Rios and Wu [25]) in
immersive learning, I propose a dual-perspective Edu-Metaverse framework, considering both XR
capabilities [
From the learners’ perspective, the framework is rooted in constructivist learning theory [22], which
emphasizes active, contextualized knowledge construction through social interaction, and incorporates
its application within Edu-Metaverse environments [
From the educators’ perspective, the framework helps instructional innovation through tools
such as AI-assisted content creation and AI-based topic modeling [31, 32], enabling faster preparation
of immersive lessons while supporting consistency and alignment with curricular goals. It supports
learning performance evaluation and instructional control (e.g., pacing and sequencing) to optimize
the efectiveness of learning sessions. Although Edu-Metaverses are often conceptualized as primarily
learner-centric environments [
This framework combines established theories on XR, learning science, and educational technologies with practical insights, drawn from our user studies, to support the design of efective Edu-Metaverse experiences, thus avoiding focus on specific hardware implementations and possibly adapting to evolving XR technologies. The goal is to ensure that Edu-Metaverses enhance educational outcomes by improving knowledge gain, retention, and transfer of learning, both formal and informal (see Ainsworth and Eaton [33]), while fostering collaboration, creativity, and critical thinking.
XR Technologies
Virtual Reality TRADITIONAL
LEARNING
My eforts to develop an evidence-based Edu-Metaverse framework began by focusing on a core subset of this broader research agenda. Specifically, I have explored several key components related to the learner experience. Three studies have been completed and are ready for submission.
In my first study I developed an immersive VR Edu-Metaverse reconstructing a North-East Italy rural area to evaluate how collaborative learning impacts cultural heritage knowledge acquisition. I investigated whether collaborative learning boosts content retention compared to individual learning and for which content type. Results show that collaborative learning significantly benefits concise content, where users activated more objects, while discursive content showed no diference. Social presence was higher in collaborative conditions, while no statistically significant diference was found for the sense of presence. These findings indicate that immersive collaboration improves learning when information is concise and easily shareable between peers, highlighting the need to organize knowledge into manageable chunks inside Edu-Metaverses in informal learning scenarios.
The second study examined how locomotion techniques afect cultural heritage learning in immersive VR. Native Italian speakers learned dialect vocabulary about rural objects from the Natisone Valleys in a virtual environment populated with autonomous agents simulating a collaborative experience. Three conditions were tested: teleportation, arm-swing, and “lerport” (teleportation for self with continuous movement visualization for others). Teleportation significantly improved vocabulary retention, reduced simulator sickness compared to arm-swing, and enhanced presence over lerport. No significant diferences were found in social presence, satisfaction, or usability, supporting teleportation as an efective technique for cultural learning in immersive VR and potentially inside Edu-Metaverses.
The third study developed a framework for generating VE, extracting navigation meshes, and comparing pathfinding algorithms across 1.5 million instances. Results indicate that Best-First Search is a resource-eficient alternative to A*, while BFS balances path quality and computational speed. These ifndings could enhance agent navigation (especially on resource-constrained standalone XR devices) in scalable Edu-Metaverses, supporting our proposed concept of “bootstrapping the Metaverse”.
Ongoing research focuses on expanding and refining the proposed Edu-Metaverse framework, which will be validated through user studies across diverse educational contexts.
A key ongoing efort deploys an implementation of Edu-Metaverse based on the proposed framework to compare immersive VR classrooms with traditional lecture in class, using low-cost head-mounted displays in formal education. This comparison assesses Edu-Metaverse’s impact on engagement, knowledge retention, and instructional efectiveness. In parallel, I investigate how embodiment (self and others) afects learning, focusing on avatar representation, non-verbal communication, and interactions. Additional work centers on designing and validating AI-assisted tools for educators, enabling content generation, adaptive sequencing, and instructional feedback within immersive VEs.
Future research will propose the concept of “bootstrapping the Metaverse”, integrating AI-driven agents with real Edu-Metaverse users to ensure persistent interaction even in scarcely populated EduMetaverses or individual learning contexts. Additionally, I aim to develop adaptive learning tools for educators to provide educational content based on behavioral and performance data. I will also conduct user studies to evaluate the efectiveness of traditional learning approaches and diferent XR technologies (e.g., virtual vs. augmented reality), comparing traditional methods against XR technologies as well as the two XR technologies, one against the other, within Edu-Metaverses.
My doctoral research proposes an Edu-Metaverse framework grounded in empirical evidence from multiple user studies. The results so far demonstrate that collaborative learning enhances knowledge acquisition for concise content, that teleportation-based locomotion optimizes vocabulary learning while maintaining user comfort, and that eficient pathfinding algorithms efectively support scalable multi-agent environments.
The proposed framework bridges technical capabilities with pedagogical needs by addressing learner and educator perspectives. It emphasizes that successful Edu-Metaverses require careful orchestration of multiple interdependent elements.
My doctoral thesis will apply a revised version of the proposed framework, grounded in my research results, to define guidelines and design principles for building efective Edu-Metaverses.
During the preparation of this work, the author used ChatGPT-4o and Grammarly for grammar and spelling checks, writing style improvements, and paraphrasing and rewording, in line with the GenAI Usage Taxonomy at https://ceur-ws.org/GenAI/Taxonomy.html. After using these tools, the author reviewed and edited the content as needed and takes full responsibility for the publication’s content.