https://arxiv.org/api/UcdXmQUWLejae0RAzzkpN5HlJHE 2026-06-10T22:35:57Z 4597 315 15 http://arxiv.org/abs/2504.09720v3 2025-10-12T13:03:46Z

This study explores NotebookLM, a Google Gemini - powered AI platform that integrates Retrieval-Augmented Generation (RAG) as a Socratic tutor for physics education. In this implementation, NotebookLM was configured to support students in solving conceptually oriented physics problems through a guided, questioning-based dialogue. When deployed as a collaborative tutor, the system restricts student interaction to a chat-only interface, promoting controlled and guided engagement. By grounding its responses in teacher-provided source documents, the AI tutor helps mitigate one of the major shortcomings of standard Large Language Models - hallucinations - thereby ensuring more traceable and reliable answers. This work details the methodological design of the tutor, including the iterative development of a pedagogical "Training Manual", and presents preliminary qualitative observations from demonstrations with pre-service and in-service teachers. These observations highlight both the promising potential of the tool and key pedagogical challenges, such as managing user motivation. While limitations remain, this work offers a promising and replicable model for educators seeking to implement grounded AI tutors in their own teaching contexts.

2025-04-13T21:01:21Z 9 pages, 5 figures. Revised version accepted for publication in The Physics Educator (World Scientific, in press) Eugenio Tufino http://arxiv.org/abs/2510.09971v1 2025-10-11T03:04:22Z

Liquid nitrogen (LN2) is a long-time favorite for physics demonstrations, with a large repertoire of crowd-pleasing experiments that are cornerstones in outreach efforts. While R1 universities usually have a ready LN2 supply for their Physics, Chemistry, and Biology departments, K-12 and two-year college teachers often have to go to specialty suppliers to obtain LN2, and also need access to expensive storage equipment. Dry ice (solid CO2) is available at many super markets, and as such its suitability as a substitute for LN2 was explored, with the results discussed below. At 77K (-196°C) LN2 is considerably colder than dry ice at 195K (-78.5°C), however some demonstrations are still viable.

2025-10-11T03:04:22Z 7 pages, 8 figures David Marasco Paolo Marasco http://arxiv.org/abs/2510.09725v1 2025-10-10T09:26:07Z

The switching losses of power transistors are generally measured using the so-called double pulse method. Measuring the opposition of two switching cells is a complementary method that is more accurate but indirect. However, implementing this method can be more complex and requires calibration steps and comprehensive control, with the added issue of thermal management. In this context, we proposed to address this topic through open and collaborative science, first in the form of a two-day hackathon, followed by monthly open sessions. More than 20 participants contributed to the two-day hackathon, followed by monthly sessions for those wishing to continue working together. This enabled us to set up an automated bench, in open science, including the generation of switching commands, the configuration and control of measuring instruments, and the hardware part. Here we present and share our work and this open approach.

2025-10-10T09:26:07Z Paper in french, presented at the french national electrical engineering conference SGE 2025 Nicolas Rouger Luiz Villa Matthieu Masson Pauline Kergus Joseph Kemdeg Lorenzo Leijnen Jean Alinei Adrien Colomb Ayoub Farah-Hassan Arnauld Biganzoli http://arxiv.org/abs/2510.08261v1 2025-10-09T14:19:32Z

We examine more than a decade of quota policy at Unesp, analyzing Physics, Biology, and Pedagogy as representative programs of distinct assessment styles. Quotas show little impact in Physics, where the admission barrier is low, and in Pedagogy, where high pass rates make it difficult to differentiate students, but they reveal systematic differences in Biology. Focusing the analysis on Calculus I - an introductory course in Physics and other Science, Technology, Engineering, and Mathematics (STEM) programs - for which much larger statistics are available, a clear hierarchy emerges: students admitted through open competition perform best, those from public schools achieve intermediate results, and students from racial quotas perform worst. When students are divided directly by the admittance exam grade, the performance difference is even clearer. Statistical analysis also shows that, contrary to expectation, the probability of passing decreases as the number of attempts increases, indicating that initial educational gaps are difficult to overcome within higher education.

2025-10-09T14:19:32Z 20 pages, 11 figures Ricardo D. Matheus Elmer M. Gennaro Marcelo T. Yamashita http://arxiv.org/abs/2410.18531v2 2025-10-09T07:40:05Z

Although conceptual assessment tests are commonly administered at the beginning and end of a semester, this pre-post approach has inherent limitations. Specifically, education researchers and instructors have limited ability to observe the progression of student conceptual understanding throughout the course. Furthermore, instructors are limited in the usefulness of the feedback they can give to the students involved. To address these challenges, we propose an alternative approach that leverages computerized adaptive testing (CAT) and increasing the frequency of CAT-based assessments during the course, while reducing the test length per administration, thus keeping or decreasing the total number of test items administered throughout the course. The feasibility of this idea depends on how far the test length per administration can be reduced without compromising the test accuracy and precision. Specifically, the overall test length is desired to be shorter than when the full assessment is administered as a pretest and subsequent post-test. To achieve this goal, we developed a CAT algorithm that we call Chain-CAT. This algorithm sequentially links the results of each CAT administration using collateral information. We developed the Chain-CAT algorithm using the items of the Force Concept Inventory (FCI) and analyzed the efficiency by numerical simulations. We found that collateral information significantly improved the test efficiency, and the overall test length could be shorter than the pre-post method. Without constraints for item balancing and exposure control, simulation results indicated that the efficiency of Chain-CAT is comparable to that of the pre-post method even if the length of each CAT administration is only 5 items and the CAT is administered 9 times throughout the semester. (To continue, see text.)

2024-10-24T08:24:47Z We dedicate this work to the memory of Professor Masaaki Taniguchi, in deep appreciation of his contributions to physics education and his lasting impact on our lives Physical Review Physics Education Research 21, 020139 (2025) Jun-ichiro Yasuda Michael M. Hull Naohiro Mae Kentaro Kojima 10.1103/x2rq-z8lm http://arxiv.org/abs/2510.08634v1 2025-10-08T18:53:29Z

Artificial Intelligence and especially Large Language Models (LLM), such as ChatGPT has revolutionized the way educators work. The results we get from LLMs depend on how we ask them to help us. The process and the technique behind an effective input is called prompt engineering. The aim of this study is to investigate whether science educators in secondary education improve their attitude toward ChatGPT as a learning assistant after appropriate training in prompt engineering. The results of the pilot study presented in this paper show an improvement in the previously mentioned teachers perceptions.

2025-10-08T18:53:29Z Dimitrios Gousopoulos http://arxiv.org/abs/2510.07154v1 2025-10-08T15:53:00Z

We describe a teaching-lab experiment that applies basic optical spectroscopy to examine the physics of semiconductor diode lasers. By using a low-power visible laser, this experiment is suitable for use in an open lab environment, where students assemble the spectrometer optics themselves as part of the project. A small holographic grating disperses light onto a camera driving a large display monitor, providing a high-resolution, real-time look at the laser output spectrum. Observing this spectrum as a function of injection current, diode temperature, and optical feedback reveals several features of the laser internal optical resonator and semiconductor properties. While teaching about the intrinsically quantum phenomenon of laser physics, this experiment also provides a hands-on experience that gives students a practical familiarity working with lasers, optics, and imaging technologies.

2025-10-08T15:53:00Z Kenneth G. Libbrecht http://arxiv.org/abs/2505.19099v8 2025-10-06T16:16:33Z

We present SeePhys, a large-scale multimodal benchmark for LLM reasoning grounded in physics questions ranging from middle school to PhD qualifying exams. The benchmark covers 7 fundamental domains spanning the physics discipline, incorporating 21 categories of highly heterogeneous diagrams. In contrast to prior works where visual elements mainly serve auxiliary purposes, our benchmark features a substantial proportion of vision-essential problems (75%) that mandate visual information extraction for correct solutions. Through extensive evaluation, we observe that even the most advanced visual reasoning models (e.g., Gemini-2.5-pro and o4-mini) achieve sub-60% accuracy on our benchmark. These results reveal fundamental challenges in current large language models' visual understanding capabilities, particularly in: (i) establishing rigorous coupling between diagram interpretation and physics reasoning, and (ii) overcoming their persistent reliance on textual cues as cognitive shortcuts.

2025-05-25T11:28:34Z 46 pages Kun Xiang Heng Li Terry Jingchen Zhang Yinya Huang Zirong Liu Peixin Qu Jixi He Jiaqi Chen Yu-Jie Yuan Jianhua Han Hang Xu Hanhui Li Mrinmaya Sachan Xiaodan Liang http://arxiv.org/abs/2510.03493v1 2025-10-03T20:17:52Z

Writing is a critical skill for modern science, enabling collaboration, scientific discourse, public outreach, and more. Accordingly, it is important to consider how physicists and astronomers are trained to write. This study aims to understand the landscape of science writing education, specifically in physics and astronomy, in higher education in the United States. An online survey probing various aspects of their writing training in both undergraduate and graduate school was administered to 515 participants who have obtained training in physics and/or astronomy, or related fields, at the level equal to or beyond upper-division undergraduate study. Humanities and writing requirement courses appear to have a key role in general writing education, while laboratory courses and feedback from mentors are the dominant modes of science writing education in undergraduate and graduate school respectively. There is substantial variation in the quality of writing education in physics and astronomy, often dependent on the student's institution and/or mentor. Some participants also report that their success in disciplinary writing was a result of a solid foundation from K-12 education and/or self-direction towards resources; such reliance on past experiences and student background may contribute to inequality in the field. Many participants also stated a clear desire for more structured writing training to be available in the field. We provide suggestions for how to implement such training to meet the needs of the community identified in the survey.

2025-10-03T20:17:52Z Accepted to Physical Review PER Briley L. Lewis http://arxiv.org/abs/2007.09480v3 2025-10-03T13:52:54Z

We present a critical analysis of the classical approaches to energy subjects, based on the work-energy theorem and the conservation of mechanical energy proposed in the courses of the first years of tertiary education. We show how these approaches present a series of inconsistencies and errors that are a source of conceptual difficulties among students. We then analyze a modern treatment of mechanical courses based on the results of research in physics education over the last 40 years. We place special emphasis on the principle of conservation of energy as one of the fundamental principles of nature, prioritizing the concepts of system, surrounding, and energy transfer and transformation.

2020-07-18T17:20:36Z Avances en Enseñnza de la Física, vol 2, 2020 Alvaro Suarez Daniel Baccino Arturo C. Marti Martin Monteiro http://arxiv.org/abs/2510.01467v1 2025-10-01T21:19:12Z

The rapid emergence of generative artificial intelligence (AI) and related technologies has the potential to dramatically influence higher education, raising questions about the roles of institutions, educators, and students in a technology-rich future. While existing discourse often emphasizes either the promise and peril of AI or its immediate implementation, this paper advances a third path: a principled framework for guiding the use of AI in teaching and learning. Drawing on decades of scholarship in the learning sciences and uses of technology in education, I articulate a set of principles that connect broad our educational goalsto actionable practices. These principles clarify the respective roles of educators, learners, and technologies in shaping curricula, designing instruction, assessing learning, and cultivating community. The piece illustrates how a principled approach enables higher education to harness new tools while preserving its fundamental mission: advancing meaningful learning, supporting democratic societies, and preparing students for dynamic futures. Ultimately, this framework seeks to ensure that AI augments rather than displaces human capacities, aligning technology use with enduring educational values and goals.

2025-10-01T21:19:12Z Noah D. Finkelstein http://arxiv.org/abs/2510.01124v1 2025-10-01T17:11:27Z

A growing number of introductory physics instructors are implementing active learning methods in their classrooms, and they are modifying the methods to fit their local instructional contexts. However, we lack a detailed framework for describing the range of what these instructor adaptations of active learning methods look like in practice. Existing studies apply structured protocols to classroom observations and report descriptive statistics, but this approach overlooks the complex nature of instruction. In this study, we apply network analysis to classroom observations to define a typology of active learning that considers the temporal and interactional nature of instructional practices. We use video data from 30 instructors at 27 institutions who implemented one of the following named active learning methods in their introductory physics or astronomy course: Investigative Science Learning Environment (ISLE), Peer Instruction, Tutorials, and Student-Centered Active Learning Environment with Upside-down Pedagogies (SCALE-UP). We identify five types of active learning instruction: clicker lecture, dialogic clicker lecture, dialogic lecture with short groupwork activities, short groupwork activities, and long groupwork activities. We find no significant relationship between these instruction types and the named active learning methods; instead, implementations of each of the four methods are spread across different instruction types. This result prompts a shift in the way we think and talk about active learning: the names of developed active learning methods may not actually reflect the specific activities that happen during instruction. We also find that student conceptual learning does not vary across the identified instruction types, suggesting that instructors may be flexible when modifying these methods without sacrificing effectiveness.

2025-10-01T17:11:27Z Meagan Sundstrom Justin Gambrell Colin Green Adrienne L. Traxler Eric Brewe 10.1103/ylsr-968q http://arxiv.org/abs/2510.08584v1 2025-09-30T22:28:29Z

CERN's new flagship education and outreach centre, Science Gateway, opened its doors to the public in autumn 2023. Through a combination of immersive scenography with interactive exhibits and real scientific objects, its permanent exhibitions address the organisation's particle physics research and how this knowledge applies to other scientific fields and the visitor's everyday life. While in Discover CERN visitors learn about the accelerator and detector technologies at CERN, Our Universe covers the evolution of the Universe and the role that fundamental physics plays in it, as well as open questions in modern physics. Their underlying basic quantum mechanical principles can eventually be explored in Quantum World. Firstly, we will give a short overview of the exhibitions' highlights. Next, we delve into how various aspects of accessibility were incorporated into the exhibition development process to target the broadest audience possible and to inspire the next generation of scientists. We give an account on how this topic was approached and how inclusive exhibits and infrastructure were realised. By combining long-term visitor statistics and preliminary results of a recent visitor survey, we obtain insights about the actual audience and the visitors' perceived accessibility of the exhibitions with respect to their educational backgrounds.

2025-09-30T22:28:29Z 8 pages, 4 figures, published in PoS (ICRC2025) 1259; 39th International Cosmic Ray Conference (ICRC2025), Geneva 15-24 July 2025 Tamara Caldas Cifuentes Jemma Harris Patricia Verheyden 10.22323/1.501.1259 http://arxiv.org/abs/2510.00150v1 2025-09-30T18:21:04Z

As the field of quantum information science and engineering (QISE) continues its rapid growth, there are increasing concerns about the workforce demands and the necessity of preparing students for quantum-related careers. Given the interdisciplinary nature of the field, it is necessary to offer diverse educational opportunities to ensure that students are well-prepared for careers emerging from the second quantum revolution. In this paper, we present our findings from a qualitative study involving semi-structured interviews with university QISE educators, who have taken on the challenges and opportunities in developing QISE courses and curricula for undergraduate and graduate students from different academic backgrounds. Our findings focus on common themes across undergraduate and graduate QISE education, as well as advising and mentoring students to prepare them for research in the field. The interviewees discussed the various strategies they had implemented, such as incorporating hands-on lab activities, integration of Python coding with Qiskit, and including project-based learning experiences. Furthermore, their reflections on mentorship and advising students emphasized the importance of recognizing students' prior preparation, providing targeted resources, and supportive learning environments. Our findings are meant to provide guidance for educators looking to implement effective QISE educational strategies that address the changing landscape of quantum information revolution.

2025-09-30T18:21:04Z Eur. J. Phys. 46 055709 (2025) Fargol Seifollahi Chandralekha Singh 10.1088/1361-6404/ae0200 http://arxiv.org/abs/2509.25769v1 2025-09-30T04:30:15Z

Understanding concepts and practices of measurement uncertainty is a core competency of physicists and engineers, and many physics lab courses aim to have students learn these ideas. However, there is strong evidence that these goals are often not met. To address the challenge of improving students' proficiency with measurement uncertainty concepts and practices, we designed and developed a new PhET simulation, Projectile Data Lab (PDL), featuring statistical noise and measurement tools in the context of projectile motion. We integrated this simulation into the Common Online Data Analysis Platform (CODAP), creating an instructional platform for collecting and analyzing data from the simulation, and designed three simulation-based instructional activities for instructors to use in their lab courses. We describe the pedagogical design of the new simulation, the PDL+CODAP instructional platform, and the associated instructional activities. We highlight how the targeted learning goals guided the pedagogical design, as well as how these three instructional tools (the simulation, the PDL+CODAP platform, and the lab activity) work together and leverage the affordances of each to scaffold learning. The goal of this work is to provide a model of how noise-enhanced simulations and activities can be designed to enhance student learning of measurement uncertainty.

2025-09-30T04:30:15Z 18 pages, 4 figures Qiaoyi Liu Matthew Blackman Gayle Geschwind Catherine Carter Katherine K. Perkins H. J. Lewandowski