Examining Student and AI Generated Personalized Analogies in Introductory Physics

Amogh Sirnoorkar, Winter Allen, Syed Furqan Abbas Hashmi, N. Sanjay Rebello · 2025

Read Examining Student and AI Generated Personalized Analogies in Introductory Physics by Amogh Sirnoorkar, Winter Allen, Syed Furqan Abbas Hashmi, N. Sanjay Rebello (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-04290.txt.

Introduction

Amogh Sirnoorkar, Winter Allen, Syed Furqan Abbas Hashmi, and N Sanjay Rebello examine how introductory physics students use analogies, comparing self-made ones with AI generated ones. Analogies link abstract ideas such as circuits to familiar systems such as plumbing, and self-generated versions are thought to teach better than handed-down ones, though prompting them at scale is hard.

Around eight hundred student responses were analysed for spontaneous analogy use and for comparisons built in preferred everyday contexts. Students explaining a potential curve in simple language seldom used analogies on their own, unlike model responses. Themes of ease and difficulty in generating them round out the results for large courses.

Catalog record

CREATORAmogh Sirnoorkar, Winter Allen, Syed Furqan Abbas Hashmi, N. Sanjay Rebello
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-04290

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Examining Student and AI Generated Personalized Analogies in Introductory Physics Preprint: APS/123-QED Amogh Sirnoorkar, Winter Allen, Syed Furqan Abbas Hashmi, N. Sanjay Rebello Abstract Comparing abstract concepts (such as electric circuits) with familiar ideas (plumbing systems) through analogies is central to practice and communication of physics. Contemporary research highlights self-generated analogies to better facilitate students’ learning than the taught ones. “Spontaneous” and “self-generated” analogies represent the two ways through which students construct personalized analogies. However, facilitating them, particularly in large enrollment courses remains a challenge, and recent developments in generative artificial intelligence (AI) promise potential to address this issue. In this qualitative study, we analyze around 800 student responses in exploring the extent to which students spontaneously leverage analogies while explaining Morse potential curve in a language suitable for second graders and self-generate analogies in their preferred everyday contexts. We also compare the student-generated spontaneous analogies with AI-generated ones prompted by students. Lastly, we explore the themes associated with students’ perceived ease and difficulty in generating analogies across both cases. Results highlight that unlike AI responses, student-generated spontaneous explanations seldom employ analogies. However, when explicitly asked to explain the behavior of the curve in terms of their everyday contexts, students employ diverse analogical contexts. A combination of disciplinary knowledge, agency to generate customized explanations, and personal attributes tend to influence students’ perceived ease in generating explanations across the two cases. Implications of these results on the potential of AI to facilitate students’ personalized analogical reasoning, and the role of analogies in making students notice gaps in their understanding are discussed. I Introduction Making sense of an abstract concept (e.g., electric current) by comparing it to a familiar idea (water flow) is a common practice in physics. Such comparisons facilitate integration of new information into the knowledge system by bridging unknown ideas with familiar ones Podolefsky and Finkelstein 2007 . Consequently, comparisons are extensively employed by teachers and informal educators in communicating complex ideas to students and non-disciplinary audiences. Comparisons also form a key component of physicists’ practices towards generating new knowledge Verostek et al. 2022 . Given their extensive use in learning environments, several studies have demonstrated the effectiveness of comparisons in enhancing students’ content understanding in physics Serbin and Wawro 2024 ; Brown and Clement 1989 . Comparisons can be categorized into similarities and analogies Gentner and Maravilla 2017 . Similarity statements such as the “ electronic structure of Hydrogen atom is like that of Deuterium ” map both the object features (such as the number and orbital trajectory of electrons) and the underlying abstract relational features (electrostatic force binding the nucleus and electrons). On the other hand, analogies such as “ Hydrogen atom is like our planetary system ” predominantly map relational features (centripetal forces) as compared to object features (color of the central, massive entity). Analogies are thus sophisticated comparisons rather than similarity statements, as they establish associations between two dissimilar systems based on the underlying non-trivial, higher order, and abstract relations. Theoretically, an analogy is a mapping between a familiar idea known as the “Base” and a relatively unknown idea, the “Target”. In the above analogy, while the planetary system represents the Base, the atomic model acts as the Target Gentner 1983 . Studies in physics education have highlighted several benefits and caveats on leveraging analogies for students’ learning. To begin with, analogies can be generative as they aid students in drawing novel inferences about the Target domain by extending ideas from the Base domain Podolefsky and Finkelstein 2006 . This can lead to students’ enhanced content understanding and better problem-solving outcomes. However, self-generated analogies tend to better facilitate students’ learning than the taught ones. This can be due to various reasons including students’ inability to see the analogy’s relevance in the given context Brown 1994 or its limitations in explaining a phenomenon Glynn 1989 . However, facilitating students’ generation of personalized analogies present challenges (particularly in large enrollment courses) such as attending to individual student’s varied lived experiences and consequently their preferred analogical contexts. Recent advancements in Generative Artificial Intelligence (AI) have promised a huge potential in addressing this issue. AI platforms such as ChatGPT are large language models trained on huge datasets which enable them to produce human-like and personalized responses (albeit sometimes incorrect) based on user inputs. Given their capabilities, recent studies in physics education have explored their utility in generating and solving physics assessments Kortemeyer 2023 ; Sirnoorkar et al. 2024 , or providing personalized feedback Wan and Chen 2024 ; Sirnoorkar and Rebello 2025 . However, studies focused on comparing student- and AI-generated analogies based on students’ preferences in physics education remain scarce. Furthermore, efforts on facilitating students’ personalized analogies about abstract college-level physics concepts remain relatively underexplored. Also, research on students’ perceived affordances and constraints in generating personalized analogies remains relatively limited. We address the aforementioned gaps in the literature by qualitatively investigating how introductory students leverage analogies when asked to explain the Morse potential curve in a language suitable for second graders. We expected that when asked to do so, students would resort to making spontaneous comparisons with second graders’ lived experiences to effectively communicate a complex idea to such a non-disciplinary audience. We then contrast student-generated descriptions of these concepts with the AI-generated ones (also generated and reported by students). Furthermore, we analyze students’ self-generated analogies, while they described the behavior of the curve in terms of their preferred everyday contexts. Lastly, we investigate students’ perceived affordances and constraints in coming up with spontaneous and self-generated analogies. This manuscript is structured as follows: In the next two sections, we present the background on personalized analogies and articulate our research questions. In Sections IV and V we detail theoretical and design frameworks guiding our study. We then discuss the study’s context along with methods of our data collection and analysis in Section VI . In Section VII , we present our results and then discuss them in Section VIII . Lastly, in Section IX , we provide the concluding remarks along with the study’s implications, limitations, and future work. II Background Analogies establish associations between a familiar domain ( Base ) and a relatively unfamiliar one ( Target ). Constructing such mappings constitutes a central epistemic practice in physics, as they enable learners to make sense of novel and abstract ideas. Prior research has consistently underscored their role in supporting students’ conceptual understanding and problem-solving in physics Gentner and Gentner 2014 . However, several challenges accompany the effective use of analogies in learning. Firstly, students may not always employ analogies as a reasoning tool Gick and Holyoak 1980 , and when they do, there is a risk of drawing unintended inferences Zook 1991 . For example, when analogizing the atomic model to a planetary system, students might incorrectly infer that electrons, like planets, attract each other. Brown Brown 1994 observed that analogies that seem intuitively appropriate to experts may not resonate similarly with students. Thus, learners may not always appreciate the limits or scope of analogies Glynn 1989 . Brown and Clement Brown and Clement 1989 argue that for productive reasoning with analogies, students must participate in an epistemologically dynamic environment rather than simply accommodating a provided analogy. To

Read the complete text