Interactive AI-Assisted Multimodal Learning: A Qualitative Study of Student Experiences Using the NotebookLM Platform
Metaverse · 03.07.2026 · 7 (3)
DOI: 10.54517/m8448 ↗Kurz gefasst
Qualitative Studie in der Zeitschrift Metaverse, online erschienen am 03.07.2026. Verfasst von Aref Abu-Gweder und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This study explores undergraduate students’ learning experiences in an AI-driven multimodal learning environment, focusing on Google’s NotebookLM platform. Grounded in a qualitative inquiry, the research analyzes the experiences of 123 students in Israel—predominantly female (93.5%)—who engaged in inquiry-based tasks utilizing Retrieval-Augmented Generation (RAG), source-based questioning, mind mapping, structured study guides, and podcast-generated summaries. Data analysis was based exclusively on students’ reflective self-reports, without systematic artifact analysis due to limited access to student-generated products. A rigorous thematic analysis was conducted independently by both researchers using a reflexive interpretive approach. The researchers then compared their coding, discussed emerging themes, and reached consensus through iterative dialogue to enhance the credibility and trustworthiness of the findings. This collaborative analytical process resulted in the identification of four primary themes: active multisensory engagement, enhanced conceptual depth through dual-coding pathways, learner autonomy through prompt-based interaction, and persistent challenges related to cognitive overload and content credibility. The findings support a novel Interactive AI-Assisted Constructivism model, which extends traditional constructivist perspectives by positioning artificial intelligence as a co-constructive mediator in knowledge building. Importantly, the study proposes practical pedagogical strategies for reducing hallucination risks and improving the reliability of AI-generated content, including systematic cross-checking of AI responses against uploaded primary sources, citation tracing and source alignment verification, comparison between AI-generated summaries and original texts to identify omissions or distortions, the use of contradiction prompts to test consistency across sources, and instructor-guided peer review of AI outputs. These strategies contribute to the development of critical AI literacy, epistemic responsibility, and evidence-based learning practices. Overall, the study positions NotebookLM as a distinctive educational tool that fosters personalized, reflective, deep, and accountable learning within contemporary higher education.
Zitieren
APA
Abu-Gweder, A., & Solomovich, S. (2026). Interactive AI-Assisted Multimodal Learning: A Qualitative Study of Student Experiences Using the NotebookLM Platform. Metaverse, 7(3). https://doi.org/10.54517/m8448
Vancouver
Abu-Gweder A, Solomovich S. Interactive AI-Assisted Multimodal Learning: A Qualitative Study of Student Experiences Using the NotebookLM Platform. Metaverse. 2026;7(3). doi:10.54517/m8448
BibTeX
@article{AbuGweder2026,
author = {Abu-Gweder, Aref and Solomovich, Shay},
title = {Interactive AI-Assisted Multimodal Learning: A Qualitative Study of Student Experiences Using the NotebookLM Platform},
journal = {Metaverse},
year = {2026},
volume = {7},
number = {3},
doi = {10.54517/m8448}
}
EUVIMED